The 4x Keihin CVK carbs (semi-flat slide design) were based on race bikes, and being brand-spanking new tech on the GPz in 1984 they were seriously cutting edge. Although FI was introduced by Kawasaki in 1980 (K1000H) it obviously wasn’t considered refined, powerful or reliable enough at that time.
In 1988/89 the bike was released in Australia with smaller CVK32. No-one knows why but I suspect it was to gain data for the 2nd gen bike release. All other versions use CVK34 BUT there are lots of different Keihin carbs assemblies (+10 for 1st gen) with different top chambers, slides, needles and jets for different markets & models.
Often it was the top cap that limited slide travel resulting in a slight reductions in top end power, but other changes may also be related to improving fuel efficiency.
According to the FACTORY MANUAL the most powerful bikes (max 115hp) were the general Euro-spec A1-A6 bikes. Any location-specific bikes, including the 1984 US & California models (110hp), were slightly detuned, however the top-end performance could clearly be improved with larger jets – up to #142.
Note these were specified for Race Use according to factory PN and only for the 1984 bikes.
Second generation bikes (1990+) are *slightly* less powerful (max 108hp) but interestingly they achieve this with a significantly reduced (50%) airbox intake and a much smaller stock main jet of just #108 vs #138. Later bikes use a matching N67i needle (vs N27a/b) so one assumes the power deliver to 3/4 throttle will be similar, but it’s still a significantly smaller orifice!
For the maths nerds (like me) a #108 jet is 0.916mm² and the #138 is 1.496mm², so the jet could *theoretically* flow 63% more fuel.
So although the early assemblies had *theoretically* more power, given the reduction in intake it’s possible the later assemblies have more scope for improvement. But it’s all weird science to me.
Carburetors are one of those wonderful motoring components that are incredible simple in operation yet incredibly complex at the same time. You can’t just plug it into the PC. So they are a key to improving performance, and as most modern machinery uses FI (especially performance vehicles) Keihin experts (not simply repairers) will logically become rarer & therefore more expensive.
So there is clearly an economic advantage to do the work yourself, especially for the GPz as the final set-up will probably require several cycles of removal, adjustment, refitting and re-measuring.
So I humbly suggest rebuilding carburetors is not the often quoted online *easy* and the final fine-tuning probably best left to experts or someone with hands-on experience.
Click here for the Technical Discussion Page
We all think the GPz900r is hot right?
Well it is!
The GPz900r was a [performance motorcycle so the cooling system is:
designed to run at the optimum operating temp
not use power unless absolutely required (fan)
The system is perfectly designed to be in equilibrium when you are moving, so if your bike is overheating when riding then this is not generally a problem with any of the components, it is something else.
The problem with the design is that without fresh (cooler) air moving through the system heat rises from the headers, and because space limits the size of the electric fan on hot ambient days and slow-moving traffic the fan may not create enough heat exchange through the radiator and the bike will overheat.
This is a flaw by design – even the Factory Manual advises against letting the bike sit at idle for more than 5 min.
Although fundamentally the same, there are lots of small variations between bikes and this is by no means a comprehensive list nor a definitive technical analysis of any listed below. This simply provides a bit of insight and a starting point for your own more detailed research.
Crankcase & Block
Mid 1985 marked the only major engine change for the GPz, the switch is defined by engine numbers. There are two different numbers, I suspect these are used to identify engines originating form the US plant (ZX900AG) and RoW (ZX900AE).
The changes are likely related to the additional bolt detailed in the Cylinder Head section (same engine numbers)
Cylinder Head Cover
Early US bikes had an emissions feature that fed additional air directly into the exhaust to assist with final combustion, this was achieved with two ports on the top of the Cylinder Head Cover [14024-1138]. The RoW cover (including Canada) has a simpler design, hence a different part number [14024-1082]. This part is unchanged right through until 2003.
However just as for the Head, the ported cover appears back in schematics for the A12 (2000). This cannot be a co-incidence and suggests that there is a direct connection between the design of the head and the cover.
[ZX900AE000001 to 031543] – doesn’t match change with other Head engine numbers (30893) [ZX900AG000001 to 004600] – does match change with other component engine numbers (4600)
[14024-1138] – 84/85 – US
[11012-1481] – 84/85 – Cap, Reed Valve
[92150-1147] – 84/85 – Bolt
[92055-1225] – 84/85 – Bolt O-ring
[ZX900AE031544 to 032689] [ZX900AG004601 to 005150]
[14024-1138] – 84/85- US
[11012-1372] – 84/85 – Cap, Reed Valve
[92150-1595] – 84/85 – Bolt
[92055-1314] – 84/85 – Bolt O-ring
So same head, but for some reason 549 bikes had slightly different parts. The subtle differences are in the Cap, Reed Valve, Head Cover bolt & Bolt O-ring. no idea what these differences were.
[14024-1138] – 86 – US
[11012-1481] – 86 – Cap, Reed Valve
[92150-1147] – 86 – Bolt
[92055-1225] – 86 – Bolt O-ring
And then for the A3, they go back to the original parts, so on assumes whatever they were trying didn’t work! And these are the same part for the A12 re-introduction so apply from 2000 to 2003.
US & RoW covers
Unlike modern Exhaust Gas Recirculating (EGR) I do not believe this feature degrades performance but does have rubber valves which can degrade, block and create issues so many users simply block this feature off completely. Kits are available online. However the A1 bikes are generally specified with a much larger main jet, so one would expect these versions to be the ones that actually benefit from this the most. Well the environment will anyway.
As for the confusion with engine numbers, well if I completely ignore any reference to the ZX900AE series it all makes perfect sense with a direct match between heads & covers for US bikes and the re-birthed Malyasian/Japanese ones. And if we are assuming that the emissions port only applies to US ZX900AG engines then we can safely ignore these other engine numbers.
But this is only my in-expert analysis!
Cylinder Head & Oil Pipes
Part number analysis shows there are 4 different heads.
[11002-1153] – 84 & early 1985 – US
[11002-1163] – 84 & early 85 – Canada & RoW
[11008-1177] – late 85 & 86 – US AND 99 to 2003 Japan/Malaysia
[11008-1178] – late 85 to 2003 – Canada & RoW
The most obvious change (based on engine numbers) is the addition of an extra bolt mid 1985. This seems a fairly significant change but curiously there is little online information as to the reasoning for this. Change occurred at:
[ZX900AE000001 to 030894] > [ZX900AE00030895+]
[ZX900AG000001 to 004600] > [ZX900AE004600+]
US 84 to early 85
US late 85-86
As discussed above my instinct says that the unique US head is to match the emissions ports in the cylinder head cover. But……I had read (forum?) that the RoW cover automatically blocks off the emissions vents, and if that is the case then I have no idea why the first US part [11002-1153] would need to be different to [11002-1163] .
However it may be as simple as using the part numbers to identify the manufacturing plant. Thus the Canadian models, like the rest of the world, might be using heads from Japan. Seems a good *alernative* theory!
What is also interesting is the 3rd generation (+A11) bikes look they could have been made with spare US stock from 1986! So it’s likely they are interchangeable, at Apr 2021 these are still listed as available on Impex for US$1,489.
Cylinder Head Oil Pipe
According to MSP the European A2 oil pipes were [32033-1183] and superceded by [32102-1423]. This #1423 part number is used on *all* other sites for *all* versions, so it would seem highly likely that this is the 6mm pipe (upsized from 4mm) to help prevent cam pitting. Oddly the ’86 RoW schematic (from Australian online DB) shows both the 32033 & 32101 part reference, although the PN for both is the later [32033-1423]. Very wierd.
Although US online databases do not list the #1183 part (in fact MSP is the only site to even mention it) it’s probable that they were used intitially as this was a change made *after* the bike was released. So if I had an A1 I would be checking that these had been replaced.
4mm #1183 still available online, I would pass and use the 6mm #1423 pipes!
Although a little buried the GPzzone has replacements. Apr 2021 no international shipping tho….
According to MSP schematics for European A5 bikes (earliest online spares I have found) there is a transmission change again based on engine numbers. These numbers have no direct relationship with the changes to other engine components. I have had a quick look at the schematics & part numbers and cannot see any glaring differences.
One of the strange aspects is this European site listing includes the ZX900AE & ZX900AG engine numbers, as mentioned above my current thinking is this possibly represents US & Japanese production. My Australian bike is AE. However the Kawasaki US site, despite confirming all of the other engine changes, has no record of a transmission change for 84-86. IMPEX also confirms the engine changes for US bikes, but again no record of any transmission changes.
So if ZX900AG does represent US bikes, why is the relevant engine numbers listed here but no-where else?
Again this is really just for interest, obviously if your bikes engine numbers are below the engine number change that’s something extra for you to ponder. But the in-consistency highlights the difficulty in accessing definitive information from online sources, and unfortunately it’s highly unlikely that your local Kawasaki dealer will know any more either.
Airbox , Rubbers, Ducts & CVK32 carbs
See the the Technical Discussion page for comprehensive detail on the many different carb assemblies used. All bikes use CVK34 except for some Australian 87-89 bikes where we also had CVK32.
There are 5 different airboxes.
[11010-1214] – 84 US – probably unique to provide air supply for the cylinder head cover emissions ports.
[11010-1215] – 84 to 89 – Canada & Europe & assume Australia
[11010-1410] – 90 to 99 – RoW
[11010-1765] – 00 to 03 – Japan/Malaysia. Again probably for bikes using the US heads with emissions ports.
All of these use the #1 & #3 Airbox rubbers [14073-1273] & #2 & #4 [14073-1274]
[11010-1765] Again unique probably for bikes using the US heads with emissions ports. Whilst they could possibly have used the original US [11010-1214] airbox this one must be a bit fancier!
[110110-1220] – 1987 to 1989 – Australia (assume the CVK32’s)
This airbox uses the #1 & #3 Airbox rubbers [14073-1224] & #2 & #4 [14073-1225]
In case you didn’t notice a reminder that there are two different shapes of airbox rubbers!
Airbox schematics showing different duct sizes. Note that schematics for 87-89, including those for Australia, only show the full size duct even though there was a smaller one presumably for the CVK32’s.
Airbox Duct Intake
There are also 3 different airbox duct intakes.
[14073-1179] – 86-93 – US & RoW – Full width
[14073-1189] – 87-89 – Australia – 50% for CVK32
[14073-1441] – 90-03 – Everywhere – 50% for all bikes
A likely reason for the new part for 2nd & 3rd gen bikes is that the simple blanking used for Australian CVK32 bikes [14073-1189] probably doesn’t have great air dispersion inside the airbox, although the foam design would also reduce noise, a Kawasaki ‘feature’ noted by some journos at the time. Kawasaki obviously put in a bit more effort on this part for the 2nd gen release.
[14073-1441] shown without the foam
So why do some Australian bikes have CVK32?
Well, in a Covid world I’m beginning to wonder if it wasn’t a clinical trial. Australia is well known as an excellent ‘un-official’ test arena for Japanese machinery, both road & race. And we often don’t get recognition of our early deployment (eg the market release of the GPz in late 1983) because manufacturers want the ‘bells & whistles’ launches to be in Europe or the US. So perhaps the CVK32’s were the original preference for 2nd gen bikes where criteria such as fuel efficiency were more relevant than pure power.
Maybe they were simply cheaper, I don’t know, but it’s easy to see how running these from 87-89 would potentially give Kawasaki some great user and dealer data. The fact that the official 87-89 schematics were not updated with a reduced duct image but post 1990 they are also possibly gives an insight into how Kawasaki were approaching this. Suck it (pun intended) and see.
So for those owners who bought one in the day? Yep, you may well have been a Kawasaki lab rat!
Each fuel tank colour scheme has it’s own part number. This is not a comprehensive list (just Australia) please refer to the online databases for all colour options.
[51002-5122-P9] – 86 – Red/white
[51002-5159-J3] – 86 – Ebony/red
[51002-5219-L8] – 86 – Blue/silver
The major difference is that the deign of the tank changes between 1st (A1-A6) [51002-#] and 2nd (+A7) [51003-#] bikes, visually easy to tell by later bikes *not* having a vacuum line.
1st & 2nd gen tanks.
What is unusual is the schematics still shows the two hose barbs for 2nd gen tanks even though they physically only have one. For most other parts (eg. the starter clutch) the schematics are updated to visually reflect the change where possible. Just not here.
Rear Shock Bolt
Schematics for US-built A1 & US/RoW A2 bikes show 2 different bolt sizes for the top shock mount
[92002-1465] – 10x130mm
[92002-1462] – 12×137.5mm
I have read that this was to prevent cracking in this mount. I’m not familiar with this update, however I assume that the larger bolt must also require a change in some other parts.
Schematics shows both.
By the A3 the smaller bolt is no longer shown on schematics, but this suggests that the smaller bolt my have been used on bikes right up till then.
As well as a coil-over shock Kawasaki finally put grease nipples into the rear suspension link for the +A10 bikes [39007-1283].
Still available at Impex (Apr 2021) for $US133.55.
2nd Gen Handlebars
The second generation bikes had a different top yoke design to 1st gen (A1-A6), one where the handle bar mounting brackets completely cover the top tube. For A7 & A8 they came in two flavours, and I recall reading the integrated version (on my A8) was to stop ‘squeaking’ whilst riding. Apparently not dangerous but disconcerting!
From A9 right through to A16 schematics only show the integrated design available.
Aussie prices for a 1991 (Apr 2021)
The analysis shows you would need to be a it careful ordering these (if available) because the ‘integrated’ handle has exactly the same part number as the just the mount for the component ‘handle’! I’m not sure why you would ever order the component design, as well as potentially squeaking it’s significantly more expensive because you need two parts ($167.45 + $103.60) for each handle.
One of the most obvious (and common) changes to the GPz900R is replacing the OEM 18″ rim with a modern 17″ rim. This provides more options for tyres but arguably modernises the design. A modern tourer like the 2020 1000SX runs a 190/50/17 and even a learner Ninja 600 runs a 160/60/17 rear tyre, still significantly wider than the 1st gen 130/80/18!
An interesting piece of trivia is that in 1984 the tyres were the widest EVER FITTED to a Kawaski motorcycle.
Do wider tyres help handling? Well, although they have a larger contact patch a wider, lower profile tyre typically has a greater profile radius – this means the bike is actually ‘slower’ to tip in & turn. There is also the question of un-sprung weight – less is better.
Physics aside, the simplest option is to replace the standard rim with a ZZR600 rim that lets you run a 160/60/17 tyre – the relatively minor changes required are detailed by MuzWaaMan. Of course you *could* just squeeze a 160/60/18 tyre on the standard rim……..
The next commonly used option is the ZX-10 rim/brake & front offset sprocket allowing up to a 180/60/17 tyre – but more tinkering & parts is required with this one as the swing-arm clearances get tight, even the chain guard gets in the way.
If you want an even wider tyre (see notes on geometry below) you generally to replace the OEM swing arm. There used to be many aftermarket parts (widely used in racing) however I haven’t located any current suppliers. Of course there are many different solutions (even a Honda VFR 800 swingarm & Honda civic rim!) but I’m only linking to sites that give technical detail on the change.
A forum member wrote about the GPz900R: In 93 I replaced it with a zx6 and realized what a pig the 900 was in the twisties. Seriously?
I recently had a forum discussion with a motorcycle shop owner who declared that with the 17″ rear the GPz900R handles ‘like a modern sportsbike’. Now I politely disagreed with her because I simple don’t think the GPz900R can ever handle like many modern bikes. Why?
228kg GPz900r: Wheelbase 1494mm, Rake Angle 29° and Trail 114mm
196kg ZX6 Wheelbase 1399mm, Rake Angle 23.5° and Trail 94mm
You simply cannot compare the GPz900R to a 1990 (or later) ZX6, a fact Kawasaki themselves proved just a year later in 1985. The new GPZ600R (195kg WB: 1430 Trail: 94mm) was more precise & nimble than the GPZ750R (same frame as 900R) and around a racetrack the sharper geometry and lower weight easily balanced the bigger bikes torque & power advantage. The general consensus was the 600 lapped quicker except in the hands of advanced riders where the bikes were evenly matched. This balance is also shown in Australian Production racing where the GPz900R and GSX-750R were also very evenly matched.
In 1988 the GPZ600R was lighter & with even more HP – CVK36carbs (bigger than the GPz900R) and dyno-tuned ignition mapping. Impressive ‘little’ bike!
Since I think the GPz900R is both sporty and a good tourer a fair comparison is the 2020 Ninja 1000SX.described as ‘Are there better sportbikes? Yes. Are there better touring bikes? Of course. However, few do so well at both. ‘
235kg 1000SX Wheelbase 1440mm, Rake Angle 24° and Trail 94mm
It’s clear from these numbers that the GPz simply doesn’t have geometry comparable to a modern sports-tourer, let alone a sports bike. And physics tells us that the GPz900R will require more steering input and more lean angle to corner at the same radius and speed. This doesn’t mean it is a poor handling motorcycle, in fact nothing could be further from the truth. Nor does it mean that it cannot be ridden fast, because it definitely can.
It simply means that it was designed for completely different handling criteria and rider (and public) expectations of 1984. One of the major limiting issues in 1984 was cross-ply tyres. Radials had only *just* entered the market and there development was a significant factor in modern motorcycle geometry.
However the stable & secure nature of the GPz was it’s strength, because if it was un-stable or difficult at speed (like the bikes it replaced) it would never have achieved the journalistic and public accolades. It allowed owners to ride fast.
Modern bikes have benefited from advancements in materials and technology (like electronics & steering dampers) that allow superbikes to be ridden ‘relatively’ safely with aggressive geometry and engine mapping.
However this isn’t always the case – the 2018 Ducati Pangiale isn’t considered an easy bike to ride.
Read my section on Superbikes which explains that high-performancehandling road bikes simply didn’t exist before the GPz900R. Kawasaki clearly developed the geometry to ensure the GPz900R handled well at high-speed, And remember that unlike most modern bikes it was widely accepted as the absolute best of it’s generation.
So…what has this to do with 17″ rims? Diameter.
OEM ~666mm (lol), 160/60/17 = 624mm. Even a 180/60/17 is only 648mm with the 190/60/17 coming close at 655mm.
Note that although versions of bikes used different rims and tyres, the geometry of the bike remains the same. Front 120/80/16 (598.4mm) vs 120/70/17 (599.8mm) and rear 130/80/18 (665.2mm) vs 150/70/18 (667.2mm). Just because I can I have adjusted the rear axle by rotating the chain tension adjusters by 180° lifting the rear axle height slightly.
So that means with a 17″ rim in the existing swing-arm you are dropping the rear axle by 22mm. Which for a motorcycle is a lot.
I was recently reading about changes Suzuki made to the 2000 version of the Bandit (great bike & motor!) – which included a trail reduction of just 3.6mm and a rake angle reduction of 0.2°. That’s right, 25.6° reduced to 25.4°. When you consider the effort & cost to implement these changes into a manufacturing process it really highlights how tiny frame geometry changes must actually be quite significant – engineers simply wouldn’t change it otherwise.
The subtlety of geometry changes is again reinforced in this 2020 review on the Ducati V2 Panigale. The riders talks about a TWO MILLIMETRE rear ride height change (~7:40) from the previous 959 as having a major difference to the bikes handling. Yes he mentions possible other factors, but even pondering that such a small height change can have such a big handling difference is mind-boggling. We might not feel this, but I think most mere mortals can likely tell the difference of a 10mm change.
I don’t argue that the geometry relationships are incredibly complex, but recognise the general rule to raise or lower BOTH front & rear to maintain the same basic frame orientation. But due to the head stem, particularly with +A7 bikes wher the handlebar bracket covers the top tube, you can’t easily drop the front end of the GPz900R – by anything.
Dropping rear axle height reduces ground clearance and increases both rake angle and the trail. All of these geometry changes typically mean increased straight line stability at the expense of agility. So add this to her already old-school geometry and I often wonder if the handling changes of a 17″ rear rim aren’t quite what folks were expecting!
You can even install a longer shock (after market or alternative Kawasaki ones) raising the seat height, reducing rake angle & putting more weight onto the front wheel. This seems a good option if you have longer legs (not me) and a 2nd gen bike, where the weight balance is not quite as even as with 1st gen bikes (bigger forks, rim/tyre & brakes).
This exact handling issue is noted by MuzWaaMann with his 17″ rear conversion. He modified the front for the street fighter look and then modified the top clamp so he could drop in down an inch or so. “The other reason was to get back the handling that was lost after the rear wheel swap.”
I would always lean towards the huge 190/55/18 tyre used by WrenchMonkees because it has the same diameter as OEM keeping the bikes geometry and side profile almost unchanged. Nice.
But this rim/tyre combo is surely going to be significantly heavier. Here’s an interesting article discussing why tyre weight is more important than rim or brakes, of course none of these Michelin tyres come in an 18″ size! At the end of the day, the GPz900R engineers were pretty smart folks and created a motorcycle universally recognised for BOTH it’s handling agility and poise at high cornering speeds. So keeping the geometry as close as possible to OEM sounds pretty smart to me.
Yes it does give you more rear tyre options and they are wider. But does it improve grip? I doubt it.
GPZ1000RX (Ninja 1000R) engine
In 1986 the replacement for the GPz900R was released – the ZX1000A known as the GPZ1000RX (Oz) or the Ninja 1000R (US)
It is almost as interesting a story as the GPz900R, because after stunning the world with the GPz900R Kawasaki simply ignored all of the design criteria that made the Ninja such a great motorcycle and went back to old school 1970’s marketing for this release.
because perimeter frames were ‘the next big thing’ they created a new frame on the design from the excellent GPZ600R.
to increase power they simply used in a bigger capacity with a 74mm (72.5) bored and 58mm (55) stroked GPz900R engine (997cc)
What they got was ‘the world’s fastest production bike’ but it lacked the poise, the balance and the confidence-inspiring ride of the GPz900R. Reviewers weren’t terribly impressed, with nearly all preferring the old bike. It’s not that it’s a bad motorcycle, but to be less worthy than it’s predecessor and totally out-classed by the arriving Gixxers meant Kawasaki came crashing down from the performance pinnacle they had achieved just two years earlier with the GPz900R.
What is of interest is the engine – a bigger, more powerful (116hp@rw) version of the GPz900R that *almost* bolts straight in.
The RX also uses CVK36 carbs and the GPz900R airbox-to-carb rubbers will still fit ONLY if you have CVK34’s on the bike. I’d suggest that new rubbers will make this difficult process easier. If you have CVK32’s you can use the airbox from a GTR1000, the shaft-drive tourer version that uses a detuned (94.4hp@rw) RX engine using the smaller CVK32’s.
Forum members have said it only needs 8mm of material shaved off the left engine mount to fit.
The radiator for the RX is significantly taller so does not fit into the GPz900R frame, and any modification to the frame would probably mean the OEM fairing doesn’t fit either. Given the marginal static (not moving) cooling capacity of the GPz900R I suggest that improving the cooling with this bigger, more powerful engine will *definitely* be a required feature for urban riding!
A question I asked some forum members was why not use a big-bore kit. Their answer was cost.
However I suggest this will come down to the quality of the engine you purchase and the market you live in. Where there were plenty of bikes sold there will be a decent wrecking community and the chance of getting a good secondhand motor is higher. Due to shipping you also have to live quite close to the seller, so getting a good one does involve quite a bit of luck. In OZ there was only one secondhand engine listed on ebay.com.au in Nov 2020, the listing notes ‘worn cams, worn engine suits rebuild’.
So at AU$1,300 landed here in Adelaide, plus the cost of a rebuild, a new big bore kit is actually a significantly cheaper option.
Just for a laugh1
Why bigger capacity does always mean a faster motorcycle!
If the engine can easily handle the extra capacity (see below re: other engines) it does make you ponder why Kawasaki didn’t increase the capacity themselves. The GPZ1000RX is basically a stroked and bored version of the GPz900r, it seems that it was a marketing decision that determined this engine would be in a new bike. Faster but not better.
However big bore kits (usually 972 or 973cc) are still available and are installed into many bikes. BUT I have also read somewhere (can’t recall exactly) that there is potential issues with big bore kits due to the unavailability of OEM cylinder sleeves. But searching does suggest that even if they aren’t sourced from Kawasaki replacement sleeves are still available, although I have no idea why they are only listed as suitable for 84-86 bikes.
In 1995 the evolution of the GPz900r engine became bigger again for the GPZ1100 with 113hp@rw.
Although this is less than the GPZ1000RX (116hp@rw) reviewers say the bike pulls like a train from 4000RPM. But the bike was way to big & heavy to be considered sporty, especially in 1995 with bikes like the Ducati916 on the market! This engine was released right up to 1999 and being much more of a tourer decent second-hand units are a good bet.
The engine still uses CVK36 carbs but because the engine is no longer used as a stressed member the head lacks the mounting points to bolt straight into the GPz900r frame. I have contacted bergman at zrxoa.org who can provide you (for a fair & reasonable fee) a CD with the custom front bracket that you can use to fit these engines.
I have no information if the OEM fairings fit, but as per the RX the radiator would definitely not fit and your cooling system will have to be upgraded.
Ripley (believe it or not) the best performance upgrade I have done is – a seat.
The image above shows the difference in seat profiles that were available – this was actually off a rebuilt 750 Wayne Rainy replica with an even higher step between the rider and pillion – note bike is no longer for sale at https://www.racebikemart.com/
The stock seat is comfortable for both rider & pillion – with decent padding – so it’s a great touring setup. But it has almost no longitudinal support, so you slide forwards & backwards under acceleration & braking. This means that when riding with a bit of spirit you literally have to hang onto the bars, as it’s almost impossible to lock your body into position with your legs. Believe me I’ve tried shoving my knees everywhere…..
So from day one I thought about modifying the seat when this one came up locally on Gumtree. Pure fluke. The benefits are best described by the simple physics.
seat now has longitudinal support
body weight lowered
reduced longitudinal load on arms
reduced seating height means easier & better tuck behind windshield
All of these make for a significant change in the ride configuration from a tourer to a sportsbike. Yes there is a reduction in padding (on an already firm ride) and the lowered seat height won’t benefit taller riders – but IMHO it’s one of the first performance upgrades anyone should look at for the bike.
It just means spirited riding is much, much easier, plus it looks great too.
But nothing is ever simple.
Mine was definitely for the GPz, but it’s obviously not a Kawasaki OEM part. I have the ‘Shipped from Japan’ box but despite stickers and codes (Nuvo, 73-191 Smooth Type, barcode: *W0497081*) I can’t get a Google hit.
The link above from our mate’s across the ditch suggest that similar seats can still be sourced from Japan, but at AU$600+ even I would take a deep breath.
Whilst tracking down an electrical glitch I was going through the process of replacing any suspect parts and decided on some new Dynatek coils. As always noting is ever simple, because my research gave slightly conflicting information:
you should not install coils with higher than OEM resistance (increases load on other components)
higher resistance gives a slightly stronger spark hence slightly better performance at high RPM
Both make sense, the dilemma is that the Dynatek coils are specified at either grey 2.2 or green 3.0 ohms, yet the the GPZ900r OEM coil primary resistance ranges from 1.8-2.8 ohms! I was originally supplied with the green 3.0 coils – so its peak output is higher than the peak OEM output, but only by 7%. I ummed and ahhed but decided to exchange and go with the grey 2.2 ohm version.
I simply didn’t think I would ever notice the high rpm improvement so went the ‘safe’ option however I honestly don’t think it matters either way.
It came supplied as a complete kit containing leads and crimping lead terminals – so unless you have recently replaced these I strongly recommend replacing these at the same time. Obviously you need to measure these to suit your bike and then cut-and-crimp, and because I didn’t have the proper crimping tool (you can rough it but…) I simply measured & cut the leads and my supplier was more than happy to crimp the terminals on for free.
A really good example of why you should buy local!
They fit really neatly and haven’t had an issue with them at all, although because I was still having an alternator glitch with the bike I can’t really say if they perform better. Maybe its just the new leads!
2020 Prices & Availability
You can source them on eBay or from from many different local distributors in Australia. Prices on eBay seem a bit steep (AU$400) here is a local link for the coils (motociclo.com.au) at the much more reasonable price of AU$259.
Looking at the US parent site they only appear to offer the 3.0 ohm option, linked to the GPZ1100. So as this is the only older design they are offering (the new ones are called high output Mini Coils) I’m guessing that the 2.2 ohm versions are now old stock simply held by various suppliers.
Coil-on-Plug (Stick Coils)
Many forum members suggest replacing with the modern design where the coil is integrated into the spark lead. Apparently the stick coils from the ZX636 or the Yamaha MT07 are a pretty good fit, just requiring the rubbers to be trimmed/folded to reduce the overall height s it all fits under the tank. I asked some users whether they thought these were better than Dynotek and they did not know, however they do all feel it was better than OEM.
Mo K. Just done the conversion on an A7. It was fairly easy. I used Yamaha MT07 coils as recommended by Bill B. on this forum.
However new coils/plugs & leads may have been just as good, and some mechanical members suggest that the actual CDI system is the limitation here – not the coils themselves. Dyna themselves consider these ‘old’ tech, and can supply the Dyna 2000 system, a completely new CDI replacement from rotor, pickups, CDI box & coils.
Other advantages of stick coils are no need for the ‘old’ coils (so more space under tank for stuff) but it seems the main reported advantage was cost. These can apparently be sourced quite cheap (£20), however eBay suggests that in Australia this isn’t necessarily the case.
Secondhand 636 coils $160
A suggested wiring diagram is here as well, obviously you need to figure out your own method of actually wiring these directly to the CDI box & connecting the leads. I haven’t read of any issues with the OEM CDI box running directly to stick coils, if the resistances are similar there shouldn’t really be any.
An observation here is that because plugs are now are wired in series, a failure in any ONE of the wires or connectors will affect two cylinders, whereas a failed plug lead/connectors on OEM only affects that single cylinder.
If you didn’t get as a set you also need the actual plug connectors – search eBay for “Ford ignition coil connectors”.
Like many ‘modernisation’ mods there are pro’s & con’s the jury is out on this one. Everyone agrees the most significant limitation is the actual CDI system itself.
To ensure a fair viewpoint here’s an interesting link where the author suggests coil discussion is often just modern techno-babble BS and stick coils aren’t perhaps the best option for carb bikes!
MitSumi Hose Clamps
Tip: be very careful tightening the OEM hose clamps!
I suspect these are the original hose clamps used on the cooling system supplied by MitSumi. They don’t have any sizing stamps except for the letters 1081 which do match this product.
The issue I have with these clamps is that when I tightened a leaking one it ‘stripped’ (not really but you know what I mean) the thread hence losing all hose compression therefore sealing. You can actually see the metal distortion in the photo. So I then tried to tighten up another leaking one and the same thing happened. And the same with a third.
I do prefer to use a socket on the end of hand-held driver instead of a flat-blade screwdriver for tightening clamps, however I’m not using stupid torque loads. The clamps had to be tightened up because with an un-pressurised system the coolant would boil with the temp gauge only reading 1/2! I’m thinking that when the hoses were all nice and new (so flexible & uncompressed) the torque figure needed for sealing would be much lower than now, 30-odd years later.
I was in my garage and could just drive down to the local hardware store for some replacements (Tridon 21-38mm). Probably should have used stainless steel but the sizing was different – either too small or too large.
But If you were on the road and were simply trying to stop a small leak, then this clamp failure would make the leak much worse and would be really, really annoying.
Fuel tank internal rusting
Rust is bane of old petrol tanks and a potential nightmare to carburetors. I recently read a forum post where a Kawasaki GPz900r FB Group member (Dave.P) suggests adding a capful of methylated spirits, about every 4 tank fills, to help minimise this. Googling suggests that this is indeed a sensible idea. Certainly falls into the ‘can’t hurt’ category.
In 2020 a full rebuild of my front forks (A8) from a local Kawaski dealer cost AU$330 – which I thought fair & reasonable. However check that the rebuild kits they include new dust seals – mine didn’t.
Aside from needing a nice big, clean bench the challenge with forks is being able to pull them apart – to stop the cylinder from spinning you will likely need something to mimic the Special Tool 57001-183 or 57001-1057. Note that these aren’t ridiculously expensive (~US$50) and can be used on many different Kawasaki’s.
However forum members offer the following solutions.
file a square taper onto a wooden broom handle;
use the hex head of a bolt, from the forums I think 21/22mm for 1st gen bikes and 27mm for 2nd gen. One member grinds a 1/2″ square onto the thread so you can use a standard extension bar;
sean.j even reckons you can use the plug spanner from the OEM toolkit!;
Fluid doesn’t compress, so having water or fuel in a cylinder can easily blow the head gasket or bend a conrod even when simply turning over the engine on the starter. So if in any doubt remove sparkplugs FIRST.
Spark Plug #1.
Although moving the cam chain to one side of the engine was extremely beneficial and universally acclaimed as brilliant, it has one major design flaw. The rocker cover has to be raised at one end to provide clearance for the cam sprockets, which means that for the GPz if it’s on the side stand water will drip down towards plug #1 and has no-where to drain.
The hex body of Graham.B’s #1 spark plug is completely rusted away. Not ideal.
If water leaks past the plug then it’s into the cylinder.
Some of us are old enough to remember manual fuel petcocks – you had to turn OFF when finished and ON again to start. Modern machinery (!) like the GPz uses vacuum (connected from the carbs) to only allow the fuel to flow when the engine is. Clever & convenient but as per the norm the more complex something is, the more things can go wrong.
ON/RES protected by vacuum shut-off. PRI (ME) fully open
Unless there is vacuum draw (from carbs) valve stays closed when ON or RES
William C. notes that the bore of the fuel tap vacuum is very small and can easily clog with debris. He suggests removing and drilling out to about 2mm. Note that the entire petcock can be dis-assembled and rebuilt with kits.
If the petcock AND the float bowl shut-offs fails the Keihins have no in-built overflow. GPz900R tanks are notorious for rusting, if the petcock filter fails this potentially allows debris to enter the system damaging the seals. A stuck petcock is usually a first sign of debris! Many owners upgrade with a Pingel petcock., expensive but generally considered excellent.
A second in-line fuel filter between the petcock and the carbs is also a consideration, although some owners feel this can contribute to fuel starvation. A full-size car filter is installed on my A8 without this problem an it may be due to the different vacuum system used on early bikes.
If fuel enters the cylinder, the ‘good’ result is just contaminated oil as the fuel seeps past the piston rings. Karlis G. has owned many older bikes and had this happen to a couple of them the GPz900r and a GPZ750. He suggests learning the exact level of your oil (not easy on a GPz BTW) so if you would see an increase in the level before you fire it up.
The worst case scenario is a hyrolocked cylinder and bent conrod. It’s rare but possible, and because there is really no warning (Alan W. says ‘maybe’ the smell of fuel) so some owners go the extra mile and modify their carb bowls to have an overflow by drilling the base and installing an riser. If the fuel level gets past the top of the riser it automatically drains away……..somewhere!
Philip.S modified carb bowls
As with many things there’s not a simple answer. A manual OFF petcock does significantly solve the ‘fuel’ hydro-lock potential, but isn’t as convenient as a vacuum lock. What would be ideal would be a 4-position valve with ON, RES, PRI and OFF!
Keys & Lock Barrels
Available from cmsl.com!
The GPz key is itself a work of art, but also very collectable for many reasons. If I was selling the bike and had TWO, I myself would probably consider keeping one as a memento. The bike has that kind of effect on owners, so perhaps this explains why many subsequent owners, like me, only have ONE key.
And of course the key doesn’t work in the luggage or helmet locks. Not sure why, I may be that they had different key profiles, this was certainly the normal for many cars of that era.
Allen S. nicely describes the way to ‘re-program’ the locks.
I just enacted the process for “re-programming” the locks for the seat and helmet locks, to match the only key I have. Re-programing is really just filing off the bits that go up and down when you don’t want them to. Took 30 mins and works a treat.
I also wondered if you were modifying them all at the same time you might be able to interchange the wafers between them to maybe save a bit of filing. Also remember that there are lots of little bits here, including springs.
The inner most tumbler (wafer) is the one that acts to keep the barrel in place so that it does not come out with the key. If you just hammer the barrel out like I did, you destroy this tumbler as attached picture. What you need to do is carefully depress the tumbler from the inside with a very fine screwdriver and the barrel will be released (without a hammer!). You can then file/modify that tumbler to match your key and retain the barrel in position. (I was able to recover the damaged tumbler). Next tip, you can file off the various high spots without removing the tumblers from the barrel, but this can leave burrs which stop them moving freely, so best to carefully remove them and make sure they are smooth after filing.
The most frustrating glitch I have had with the bike was electrical.
After a long trip to and from Melbourne (pre Covid-19!) I noticed a slight drop in performance that just got steadily worse, to the stage where after an hour or so of highway riding the bike would just lose power and stop. Sit on the roadside for 30min and away I would go again. So was it fuel or electrical?
To cut a long story short it was a problem with the alternator. Although the battery charged fine and there wasn’t any dramatic voltage drops with the headlight, If the environmental & riding conditions were in the wrong alignment (ie it didn’t always do it) the alternator probably overheats and then internal resistance (hypothesising here?) falls over a cliff and the whole lot loses efficiency.
This was tricky to diagnose (fule or electrical?) especially as it was happening at the same time as my fuel vapour lock glitch -which had remarkably similar symptoms. Probably the most consistent symptom was – consistent engine revs! So droning along the highway was what it didn’t like, under normal varied riding it never played up.
Ultimately this was repair by replacement, replacing it with a second-hand one solved the problem hence I was able to pinpoint the culprit.
Note: the one I sourced had simply been cut (no plug) so required a bit of electrical splicing to plug back in. But you can’t easily just plug in a replacement alternator, the wiring is buried deep within the frame and impossible to just pull out without more disassembly. More than I was prepared to do anyway. So whilst sourcing an alternator with the plug is beneficial it’s highly likely you will be cutting and splicing it in anyway.
Regulator (2020 prices)
One of the curious replacements for the alternator is the internal regulator. On the left is the OEM part on eBay for the bargain of +AU$400 – a quick search shows this is not an unusually high price for this component. Obviously the internals are solid gold! Of course you can use the same regulator from other models (GPZ1000RX/ZX1000-A) but it seems everything online is the high side of $350.
On the right is the aftermarket version for under AU$100. Here is the YouTube link to installing the aftermarket version, although curiously the RMStator website no longer offers this part. Given the cost saving tracking down one of these seems a sensible option, but I have no information on how it performs.
I love this video! LOL. Seriously though you can see the damage to the clutch basket.
This is a very common flaw with several Kawasaki’s of this era – probably all that use the ‘piggy-back’ alternator. It’s also often referred to as a Starter Chain problem as the symptoms always start with starting!
The alternator chain tensioner spring loses strength over time (heat?) and eventually doesn’t put enough pressure on the chain. First noticeable as disconcerting clunking when cranking the motor, and in extreme cases it gets so bad that allows the tensioner to push back and contact the clutch plates.
A failed starter clutch requires the complete disassembly of the engine! Therefore it is highly recommended to correct any issue that may put any additional strain on this part!
There are plenty of custom solutions available on the web. Several years ago (2015?) I investigated the simple sleeve pressed over the guide shaft pin, which I thought elegant & simple, however this solution has been replaced with more robust designs.
All of these work by limiting the movement of the tensioner preventing contact with the clutch. But amazingly I managed to source a brand-new tensioner from Kawasaki, and you can still get them from gpzzone.com. So far after about 40K it’s still actually OK – but I’m sure I’ll probably have to do one of these fixes eventually.
I’d suggest having a good battery to ensure nice clean starts helps here and is nicer on your starter clutch!
I’m going to be devil’s advocate here. This part has a *lot* of back-&-forth rotation, so whilst I do not think there is any issue with the following solutions, they all replace a virtually fool-proof riveted pin with either a bolt+nut or pin+e-clip. Depending on how you view it, the jamescompton suggestion of putting RTV on the e-clip is either a good or bad thing!
Ultimately if your tensioner is clunking you need to fix it.
The instrument cluster connects using a couple of allen key bolts into captured nuts welded onto mounding plates welded onto the fairing sub-frame.
Its a rubber mount connection to two short legs extending from the bottom of the cluster which means that the cluster does have a reasonable amount of leverage as it vibrates, especially at speed. The plates nor the welding are particularly heavy duty, so after years of use (in my case 75K) the welded plates simply break.
On my bike there were four nuts you could use to mount the cluster (no idea why), and when the first set broke I just moved the entire cluster sideways to use the others. Of course that would be way too simple, and since it didn’t reverse years of metal fatigue it wasn’t long before the whole lot simply fell off.
FYI it’s surprisingly how much the cluster moves around when unsecured, even at really low speed!
As a bush fix I used a small diameter exhaust pipe clamp to go around the central tube of the subframe – the threads of the clamp then went through holes drilled into strip of steel plate and tightly secured with nyloc nuts. Matching holes were drilled and tapped for the instrument cluster mount. Rotation isn’t an issue because the cluster naturally settles against surrounding parts and this hack actually holds the instrument cluster really firmly.
A bit of bending in a vice has the gauges angled really nicely.
I am pondering buying another frame, really strengthening the mounts (ie welding bigger plates on!) and then going through the annoying process of pulling everything apart to replace.
The 3-roller starter clutch of the early bikes (84-90) was common for many Japanese motorcycles of that era, not just Kawasaki’s. The issue is that over time and (mis) use the clutch housings may crack, some users have even reported problems with the setting of just one of the springs. The negative effect of just one of the springs seem plausible from this video. And whilst the repair itself isn’t terribly difficult actually getting to the starter clutch is – requiring the engine crankcase to be split. Here is the link to Sid Youngs comprehensive web page on the process.
Kawasaki must have realised this was a problem and in 1992, when bikes were only released in Japan, they upgraded. But when they re-released the 1993 bike in Germany & Netherlands it still had the old clutch design! Charming! FYI the new style clutch still looks available on Impex however like the other options I’m sure it’s not a direct replacement.
Up to 1993 EU clutch
Upgraded clutch for non-EU bikes 1992+
If you want to keep your early bike as original you can simply rebuild, here is Ride Restorers video on the process. However the general thoughts are to replace with the newer clutch design from the 2nd gen bikes, the1000RX (A2/3), the ZX10B or the most suggested the 1995-98 ZZR1100. Note that this is a completely different design (sprag clutch) with different gear ratios and they were different within the ZZR range, so when doing the upgrade you also need the matching idler gear – buy as a set! FYI these can fail as well however it does seem to be the preferred design for modern bikes.
Thanks to Andy.B for image showing a GPz clutch &idler on the left, newer ZZR parts on the right.
Owners like Sean H. have reported this coming out, consensus is that the clutch had previously been repaired with parts perhaps from a 1000RX.
Oil hole Alignment
A curious observation & question is the alignment of the oil holes on the starter clutch & shaft. It is not documented in the service manuals therefore I personally don’t think it is critical, but nor have I found a definitive answer. If you are dismantling it would be interesting to know how it was aligned OEM. FYI here is a recent online discussion to ponder.
Joe P. If your fitting the parts yourself please ensure you align these oil holes. My factory manual does not show any reference to it but is extremely important that the clutch gets pressure oil flow through it.
Tony M. Thank you for showing that. I am just about to rebuild the cases after fitting the clutch, and I didnt know about the oil hole. At Least I can put it right before I go any further. Just thought I’d add some pictures from mine today. Mine wasnt inline but it is now thanks to Joe P.
Tony M. On closer inspection I can see that there is a slight recess in the starter clutch that will allow oil to flow around the shaft if the hole is not aligned. So not critical after all.
Joe P. I know mine failed just after removing my starter clutch on a rebuild. On advice from a bike mechanic I checked the alignment which were not aligned. Fitted new clutch aligning the holes and never had issues after that. If the groove does in fact supply the clutch with sufficient oil l would never assemble one without aligning them… I’m sure it will allow better oil flow which means better cooling. If the jury is still out on this I know how I would assemble without a doubt.
Ride Restorers. Hello Tony. I heard this theory before starting on the engine but don’t recall it mentioned in the workshop manual.In my opinion the hole in the shaft doesn’t need to be aligned but always make your own call. There is a turned cavity at the oil holes that will fill with oil and distributes the oil around and to the starter clutch. Doesn’t hurt to align the holes but I’m not convinced it’s actually needed.
In late 2020 Ride Restorers began to post some great videos of a GPz900R rebuild online. I personally would have updated it but each to their own!
Sid Young has several detailed tutorials on the GPz900R and has released a book on restoring Japanese bikes. Some feedback swings both ways, but if you use the info on his site and feel it helps consider adding it to your collection.
See Overheating for more detailed discussion on temperature senders & gauges.
Glitch: temperature gauge drops to zero only when bike is gear and in motion.
When stationary the gauge looked fine , gradually increasing as the engine warmed. Revving the bike would give a tiny little wobble on the gauge, but not significant. But when in gear and moving the gauge would drop – right back to zero. When moving you could engage the clutch, let the engine revs drop back to idle, and the gauge would return to normal reading.
Note: the gauge fluctuations were far to instantaneous & extreme to be related to thermostat operation, and the possibility of air pockets were discounted due to several system bleeds
What made this behaviour really confusing is the the temperature measuring system has nothing to do with the clutch or transmission. The gauge is simply a voltmeter with the sender acting as a variable resistor to ground. As the coolant temperature increases the resistance decreases.
I gave up trying to figure this out in my head and decided that I might as well start by checking the gauge by removing the sensor lead and grounding the circuit. A simple process of removing the wire to the sensor and then directly connecting this wire to an earth (basically anywhere on the frame). As the sensor connector is shielded and the wire quite short, you do need another length of wire – if one end has a male tab connector it is absolutely perfect.
Anyway I removed the seat, pulled back the tank and tried to remove the yellow wire to the sensor. This was the first clue of something not quite right. The female spade connector didn’t want to slide off easily, yet the actual tab (on the sensor) seemed loose. I had read that you need to be gentle here so you don’t break the tab off, so I carefully teased it off with a set of pliers. Access is limited because I simply lift the tank, removing the tank completely would give much better access.
Grounding the unplugged wire (yellow) confirmed that the gauge was operating fine (swings fully to the right), but I could also see considerable rust inside the spade terminal, I know rust is a PITA with grounding connections, but due to the restricted access I just did a quick bush fix by scraping away as much rust from inside the spade terminal as I could.
What was also odd was that although the sensor male tab itself looked clean, it isn’t a tight fit on the sensor. It can be wiggled and even had a few degrees (~5°) of rotation. I don’t have another sender to compare so this might be normal, but it does seem a little odd. When down at the local Kawasaki parts store they didn’t have any for any bikes – and weren’t of any help. As the tab is riveted there is no mechanism to tighten it – and all options are again limited by accessibility and the knowledge that it can be broken. But I had already decided I was going to replace this part anyway so bending the tab slightly did seem to restrict the movement of the tab – but that might have just been in my head.
But bugger me – everything now works perfectly. It just cannot be a coincidence.
IMHO confirming this ground connection is one of the first checks (along with ensuring no leaks – even little ones)) to do.
Vapour Lock, Fuel Caps, Petcocks & Filters
Lots of discussion on vapour lock revolves around the OEM petcock and many owners recommend changing to a *quality* manual petcock like Pingel. When the OEM tap is set to ON or RES (REServe) the petcock has a diaphragm valve that is closed. A vacuum line (which can also restrict under heat/vacuum) connects to the carby so only pulls opens this valve and allows fuel to flow when the engine is running. The way to isolate the petcock when diagnosing vapour lock is simply to remove the fuel line to the carby and put the tap into PRI (for PRIme) which bypasses the internal valve.
If fuel flows freely out of the petcock then connect back up to the carbs and go for a ride with the tap set to PRI.
If switching the tap from PRI to ON turns the vapour lock problem ON & OFF then it most likely IS related to the petcock (rebuild kit) or vacuum line (replace with bigger/stronger). FYI William C. recommends drilling out the small internal bore of the vacuum line barb (on the petcock) to around 2mm to help prevent blockage. Logical.
IMPORTANT: And although this might allows the bike to run it is not recommended to leave the petcock set to PRI as this increases the risk of hydro-lock (see above) or fire (if in an accident).
The primary is simply gauze around a cage on the petcock. It’s always struck me at just how flawed this common motorcycle install is, especially with a fuel cap that opens directly upwards, however that’s a different story. However the obvious limitations of this design do support the argument for a secondary filter between the petcock and the carbs.
Lots of owners immediately point the finger at any secondary fuel filter installed. This is such a common post there must be some merit to this, although I feel there is still some debate as to whether this is due to the smaller size of motorcycle filters, the filter orientation (hard to position so there isn’t an air bubble) and whether the problem is associated with the 1st gen tank design. On my A8 I have (bike came with it) a car sized in-line filter (Ryco Z92) specifically selected and installed in a way which should have very little impact.
Ryco Z92 Car Filter
How it fits on my A8 without any fuel flow problems
Honestly? I’m not convinced.
After modifying the fuel cap I have not had an issue with vapour lock so this STRONGLY suggests the problem was NOT related to the second filter. Ultimately it’s a pretty in-expensive component and relatively easy to install & then remove for direct comparitive testing. Mine is going to stay on.
First gen bikes (A1-A6) looks like they have a breather hose & valve running off the tank, interestingly 2nd gen (my A8) only has the single tank overflow. The caps are the same for both so I’m guessing Kawasaki found a different way to manage pressure equalisation. Of course it could be something completely different, but if it is anything to do with pressure it could affect fuel supply issues so if possible remove this from the equation when testing.
My A8 tank – clearly only one hose despite all schematic diagrams showing two!
Fuel Flow: Chris R. races his GPz and always posts ‘remove the filter & seal the tank’. He never elaborates but I’m assuming he means that rust/dirt/crud causes the petcock filter in the tank to block, and restricting flow will always negatively affect any fuel system. This can be tricky to isolate as sometimes it only occurs when there is high fuel draw, so if your engine runs fine at idle & low rpm but restricts under load this is worth investigating. How tricky? A few years ago I had the throttle body cleaned on my car but when returned it just wasn’t right. Ran smooth, it red-lined, no error codes but I knew it was down on top end power (noticed most going up a hill), even the kids noticed the change in exhaust note. However as the car appeared to be running perfectly the mechanic understandably (but frustratingly) couldn’t help. I had the engine replaced (different reason) and still the car felt lethargic but as it wasn’t the bike I lived with it. About a year later the fuel line popped out of the filter mid-corner (only time I appreciated traction control), and as soon as this was put back in: tah dah – car back to running perfect again. So it seems that this very minor mis-location only affected the engine ‘slightly’ at high revs when there was a ‘slight’ fuel flow restriction, something the car ECU simply doesn’t record. Tricky. Anyway, this was not related to my issue!
From day one my A8 had shocking vapour lock problems that would stall the bike as the fuel level dropped. It was worse when the level was low but it was related to drop in volume, so it often happened even with plenty of fuel. Incredibly annoying.
This could only ever be solved by opening the fuel cap, which is surprisingly hard to do when riding. On one occasion I got so annoyed I simply left the cap open, but of course I stopped on a soft verge and the bike did the ‘wait-for-it and then fall-over’ trick. So man-handling the bike back upright after fuel spilling everywhere I acknowledged that leaving the cap un-done is a really, really bad idea. Duh – nice one Homer!
Incredulously that is the ONLY time I have ever had the bike tip-over like that. And to rub more salt into the wound, that tip-over scratched more of my fairings than did a previous low-side crash that totaled the CDI cover and my rear wheel. Go figure!
In case you didn’t know (I certainly didn’t) the GPz cap has a complex internal venting design – so I’m not sure aftermarket versions are an equivalent swap. Only minor version changes to the 
Original 84 US [51049-1079] – still looks available at some online stores
Original 85 EU [51049-1053]
I haven’t been able to find an image online but I recall when pulled aprt the internal are surprisingly complicated. All the little valves looked OK, but no matter what I did the vapour lock problem was still there unless the cap was open.
In the end I just removed the internal valves and ta-dah, the vapour lock problem was fixed. As there is still a very restricted internal pathway that significantly limits fuel flow (maybe even still prevents it) so I’m satisfied that removing these valves isn’t a significant safety concern.
GPz900R (ZX900) Database
Number of owner contacted & confirmed bikes.
About the Site
My family loves older vehicles, the newest one we own is 2003! But I am acutely aware of the ownership complexities especially:
they often need more 'hands-on' mechanical work &;
there often isn't any local expertise from the service centres;
there is often no new parts available from the manufacturer;
parts often have to be sourced 2nd-hand or from overseas.
So we often end up doing a lot of the research & work ourselves and this information gets stored either locally with the bike or online forums - although finding the useful parts in these forums isn't always simple.
The original goal of the site was simply somewhere for me to record service work & contacts on my GPz900r so that my kids (the one that likes bikes anyway!) could easily access it - it doesn't concern me if it was publicly available.
I then realised that with this online structure in place I could also offer it to other owners, and the site could potentially expand to record other owners experiences and expertise , meaning we can learn from others but also pass on this knowledge to subsequent owners of these wonderful motorcycles.
At least Covid-19 has given me plenty of spare time to pursue my passion for the motorcycle!
The information provided on this site (or links) is personal experiences from non-professional home-mechanics, so neither it's accuracy nor it's validity can be confirmed. If you need professional advise please visit your local Kawasaki dealership or a qualified industry professional.
Like riding any motorcycle, at the end of the day the only opinion that really counts is your own!
The site is 99% free to use & your participation is most welcome.
In Apr 2021 I started a GoFundMe campaign to purchase unusual but relevant resources such as microfiche. As it is often not economical nor logical for individuals to purchase these items (eg. why purchase A1 info when I have an A8?), this campaign shares the purchasing & processing cost amongst many owners and over time potentially benefits everyone.
For their support only campaign contributors will have access to purchased resources, however this will only be a very small part of the site and all other information will remain completely free.
So if you enjoy the site and share my passion for the GPz900R, please consider donating to the campaign.