Weird problem on my Yamaha XJ650 Turbo

Hello,
I finished after many months of work my XJ Turbo but she has a weird problem.
The bike starts fine, she need only a just a little choke from cold and suddenly she idle without the choke.
If then I go and ride, the engine run smoothly and fine at any RPM but if I leave her at idle she start to go on 3 cylinders after let's say 1 minute, number 3 usually have problems! Then if I close the fuel petcock and wait a little then she runs on 4 normally. Basically I can't leave her on idle without previously closing the fuel petcock.
Last day at a traffic light I did not close the fuel tap and then she started going on 3 and even I think on 2 cylinders then she stopped! I was not able to restart the bike again, I removed a couple of spark plugs to check them then, as it was becoming night, I mounted them again. Then luckely the bike started normally and she ran really nice!
These are the works I did on the bike:

- Valve clearence checked, shims replaced and in specs.
- Compression test done and ok.
- Carbs cleaned and refurbished with carb kits (except the pilot screws and their o-rings)
- Fuel levels set with clear tube.
- Air Filter Changed (new).
- Spark plugs changed (with resistors).
- Sparks Caps changed (without resistors inside).
- Coils and wires changed (new).
- Check valve replaced.
- Fuel hoses replaced.
- Carb Sync done.
- Elettrical wires cleaned nearby the Ignition Unit.
- Fuse box replaced with a modern one.
- I checked many times the spark plugs and they are really fine, cleaned and no black carbon or oil on them.
My considerations on how to solve the problem, but I'm not an expert mechanic so I need your suggestions:

- Is the probem due to the fact that on cylinders number 2 and 3 the main jets are too large? At the moment 132.5 so as requested by specs. But from what I know main jets are not in use in idle...
- The problem could be caused by the ignition unit?
- Replacing the pilots screws and clean again the seats could solve the problem?

I don't have anymore ideas, hope someone can help me with this.
Thanks in advance.

 

First. What do you mean by close the petcock? The stock petcock is vacuum operated and has no OFF setting.

Second. What brand of carb rebuild kits did you buy?

 

This is incorrect for Turbo petcocks, which do indeed have an On-Off-Res (although I forget the order).
You say you replaced the check valve. Was this the FUEL check valve. I had issues with mine where I had replaced the seals, but after sitting for months it got sticky and would only let a little fuel by. This would starve the engine, especially at highway speeds. Sometimes sitting would allow enough fuel to sneak by to fill the bowl again and let it run.

However, you closing the petcock (assuming you're getting the positions correct) would suggest that the carbs may be flooding, and closing the petcock is stopping this. How did you wet-set your float heights? On a Turbo is particularly taxing because you need to set them with the fuel pump running (you can put the bike on the center stand, turn on the bike and ignition, select 1st gear and put the kickstand down. Pushing the start button you will only hear the fuel pump running). If you want to check it's not so bad on the bike, but if need to adjust, you'll want to build a rack and set up the fuel tank and 12V source for the pump (not a lot of fun).

If you use clear tube method to drain a carb bowl, then right the tube and run the fuel pump, you should see the level come back up very quickly and stop at the appropriate level. If it takes a long time and is inconsistent, then check your fuel check valve. If it overflows, suspect the float on that carb. If they all overflow, it could be your pressure regulator. The best way to check that (that I found) is a fuel pressure sensor, but that setup runs a little over $100 for the gauge and sensor.

 

I agree with @ManBot13 and suspect a float is not shutting off.
Shutting the fuel off may correct the overflow condition as you've found.
I'm not sure this is good for your fuel pump as it's a positive displacement pump pulling against a dead end.

 

there is a very obvious reason why a vacuum operated petcock won't work on a turbo.
As soon as you develop boost pressure it would shut the fuel off.

 

Thank you for all the comments first of all.

I did not check the carbs with the fuel pump running, I check them on the bench with the clear tube and a classic teporarily tank.
I checked the closure of the needle valve many times because it was the initial issue. I took out the carbs 3 times from the bike but last time I worked and changed again the valve seats and check them many many times, let's say 20 times!! Really 20 times because I had to be sure they were closing correctly.
In fact the spark plugs are really clean, no deposit at all. Before doing the work I always found them really carbon black.

Could you tell me more eventually on how to proceed with the fuel pressor sensor?

 

You are right about the fuel pump, I did not think that practice could damage it.
Another thing, It will be possible for you that the fuel pump has too much power and so it delivers too much fuel?
I'm asking because I don't think is an OEM one but there's no brand or written on it, all I know is that it is working fine and also the return line to the tank is very consistent.

 

If your pressure regulator is working properly then a larger capacity pump will not matter as the excess fuel is just returned to the tank.

 

you need to tee in a pressure gauge between the fuel pump and the regulator.
It doesn't matter which side of the fuel check valve you are on as pressure is the same on both sides.
@ManBot13 suggested a nice trick about putting the bike in gear, side stand down. When you depress the starter button like this the fuel pump runs without the starter motor.
You're looking for a pressure between 2-3 psi so you will need a 0-15 psi fuel rated gauge, they're cheap on eBay

 

A cheap fuel pressure gauge works if you only want to use it diagnostically. However, unlike a boost gauge, you can't run it up to the dash (you wouldn't want fuel coming up to your dash or somewhere you can see, then fail and spray fuel everywhere). An electronic gauge allows you to T off a sender in the line, and connect it to an electronic gauge. I liked this because it also helps you see that the fuel pump is keeping up with the boost pressure and regulator is working as well. If you forget to open the petcock, it'll also tell you because the fuel pressure will go from 2psi to 0 before the engine actually stops running (I had done this several times).

Wet setting with the fuel pump running ensures that the float needles are set for this extra 2psi fuel pressure, as well as showing that the regulator is working.

 

So, resuming, do you think that with a fuel pressure I will solve the problem?
Do you have a link for the regulator or pics?
About the wet setting with the fuel pump I really don't know, it seems quite complicated and I'm confident that the needle valves are closing correctly otherwise the spark plugs were not so clean.
What about changing all the pilot screws and relative o-rings? May be the problem?
What about turning the main jets of carbs 2 and 3 from 132,5 to 127 as the outers?

 

Does it stay running if you raise the idle to 2,000 rpm?
How many turns out are on the mixture screws?
I think your regulator is probably ok if the bike runs good, delivers turbo boost on the highway.

 

I did not check this, I will do and let you know. The strange thing is that even during cold start she don't need choke, not full at all, just a little. Then turning choke of she keep idle very well...

Pilot screw are moreless 1 and half to 2 turns out.

Yes the turbo is working fine.

 

I'm still trying to find the idle sweet spot with mine. Last weekend I tweaked them an additional 1/4 turn which leaves them at 3-1/4 turns out.
This is too much and I'll be moving them back to about 3 turns.
In your case I would give them a little more fuel for idle conditions.
This still does not explain why shutting the fuel pump off helps your idle condition.

 

Simmy, 1/4 turn adjustments to idle are really too large. Rickomatic had a post long ago showing the sweet spot is the width of a dime (when colortuning). Also, each one may need to be adjusted slightly differently. I do remember my turbo being at least 3 turns out (vs the 2 1/4 that is usually prescribed).

Note that idle really should be what is specified in the manual (1050 i think?). 2000 rpm is too fast and will not be running on the idle circuit. You may also be off in sync if you don't do it at idle. These engines, when properly tuned, will idle like a sewing machine.

When I had my turbo, it ALWAYS needed choke and would take a mile or two to warm up enough to turn it off. If you don't need any choke, then somehow too much fuel is getting to the engine, either because your idle screw is set to high (this usually accompanies a bad sync and not idling on all four cylinders), or and overflow of fuel (in one bowl, or possibly a leak in the "choke" circuit.

Returning to your original post, you definitely want to replace pilot screw o-rings, and this should be standard in the kit. If the pilot circuit is clogged, bleeding air, etc, you can't idle off of it. By cracking the idle screw just a bit, you may get everything to "balance out" and get an idle, but you'll have problems off idle, etc.

 

I was only suggesting the 2000 to see if @RussellNash77 's stalling went away.
If so I suspect his mixture screws maybe too lean.
I was not suggesting this as a fix, I can see how that could be interpreted.

With my bike I had it idling pretty smooth but the RPM's were hesitating a little when returning to idle, when hot.
My experience in past has always been remedied by tweaking the mixtures a little richer.
I agree with you 1/4 was too much.
BTW I've never had any success with the Colortune.
Whenever I start moving a mixture screw looking for the appropriate coloured feedback I end up way out of whack and usually just return them to 2-1/2 and tweak from there, with the Colortune back in the box.

 

Yesterday I tried to raise the idle to 2000 rpm and there was no problem. Anyway I put it back to 1050rpm, the bike has no problem at all and idles correctly.

I managed to set the pilot screws without removing the fairings and I faced the problem from a different point of view. I compared the temps of the cylinders at the exaust pipes, trying to take the temp at the same distance from the engine, after few minutes at idle I noticed that cylinder 1 was way too hot, cylinder 2 too cold, and 3 and 4 at the same temperature. So I turned out a little pilot screw 1 to rich the mixture and turned in a little pilot screw 2 to lean the mixture. Doing another temp check they are now all 4 at the same level of temperature.
Then I tried the bike from cold and she reacted normally to choke and ran really nicely! I left her at idle during stoplight and in other situations, she ran really well and no smoke at all.

I hope this problem was only due to the pilot screws settings, like Simmy I had no luck with colortune in the past...
If randomly the problem will raise again I think I will try to change the ignition unit, hope is not the case.

 

If a temp check works this makes tuning a 4-cyl bike much easier.
What are the temperatures you saw?
How many degrees do they differ between cylinders?

 

It depends where you take the temperature, I decided to take it just below the initial curve of the exaust pipes (to have a reference point) using a
thermometer gun so to have continuos readings. They were about 98 degrees C. on 3 and 4 but 135 C. on number 1 and something like 80/82 C. on number 2.
Let's say, a part the numbers, that cylinder 1 was about 35/4o degrees above and cylinder 2 about 20 C. below the other two.

Tomorrow I will try again, after leaving her at idle for a couple of minutes, to check the temperature.
It is of course very important to check the temperature at the same distance from the cylinder as it changes really fast.
I don't know if this procedure is already "normally used" and "acceptable by purists" but, from what I saw and at the moment it is working and think it can help to set the pilot screws, as you said, on 4 cylinders bikes.

 

Russell, are you still out there?
It's not the ignitor unit.
I've had the same problem, but am unable to find mixture adjustment screws on my BS30s.
Will you post a photo of their locations?