Modern Vespa

I've got my Primavera back together again. I had the engine out to redo the seals, put in a new harness, and convert to 12 volt (which I did by installing a VAPE Sport).

Starts on first kick! Yay! But I could not get it to idle without the choke, and once it's warm, it wants to stall. Took a look at the crank case oil . . . well, crud.

When I was putting it back together, I took the collective advice and built a leak tester. I caught the head gasket leaking, fixed that, and then had it lose about 1/4 psi (from 5-6) over 5+ minutes. I couldn't find any leaks by spraying windex on joints, so I took my chances. Wrong bet, I guess.

How do you eliminate the clutch seal when you have the whole engine buttoned up after a rebuild?

Any thoughts on where I went wrong when I installed the new seal, or what I can do to improve my chances of success next time around?

Spacecat,

I know of two ways to check the clutch side seal on a large frame. Might not help on a small frame.

1. If you have an engine stand that can rotate the engine, rotate it so the clutch faces upward. Remove the clutch cover and clutch. Now squirt oil all over the exposed crankshaft bearing while the engine is pressured up to 5 PSI. If the clutch side crankshaft seal is bad you will see bubbles form in the oil laying on top of the bearing.

2. Replace the vent cap on the clutch cover with a bolt. Only hand tighten it up. With the clutch cover installed, pressure up the engine to 5 PSI. Spray soapy water right on the threads of bolt installed in the clutch cover vent hole. If you see bubbles on the threads you have a clutch side crankshaft seal leak.

What kind of seal did you use? Metal or rubber?

Hec

It's a smallframe primavera, right?

Might be your crank and not your seal that's the problem.

If you have deep scores, could be that.

Also could be crankcase bolts around the flywheel too.

Neither of these explains the gas in the oil, though. That's the clutch side seal, 99%.

Sadly, Hec, there's no good way to test the seal like you're describing, because the clutch basket obscures visibility to the seal itself. You could smoke test it or listen for bubbles, but visual inspection is pretty much impossible.

Yeah if there's scoring on the crank, the seal won't seat properly and you'll get a leak. Usually I test the seal against the crank before installing.

Ensures right size and seeing if there's any rubbing problems.

The bubbles in the crankcase, looking at gasket fail.

Thanks everyone.

Hec, I can't remember now, but I think it was metal. It was in the seal kit from Scooter Mercato, and I see the one they sell separately looks like the metal one.

108, that's right -- small frame. But the crankshaft was in great shape. It had been replaced just a couple years before I parked it. Thanks for the tip about checking the seal against the crankshaft before I install it.

Chandlerman, smoke testing would be like a pressure test but seeing if smoke comes out the clutch side? Somehow I think I'd need Willie Nelson and Snoop Dogg to help out on that.

I'm going to take a breath and catch up on house and yard work before I drop the engine again. I hope I will find a reason when I open it up, but it sounds like I might not. I don't want to jinx myself, but I think it will go faster this time as I've had a practice run now.

Accepting more advice and experience in the mean time.

I don't think smallframes used metal seals as far as I've seen when shopping for seals.

For the clutch side seal, you can try use oil to see if there's any leaks.

Used this method before to find a leak which actually ended up being a tiny one in the intake manifold screw.

It's a bit fiddly doing this method because the engine needs to be tilted.

Yes! That's what I was talking about!



Hec

The method itself works great. The problem is going to be *seeing* if it's leaking behind the clutch basket:



A smoke tester can be made with a cheap soldering iron stick in a jar with some baby oil and a rag. It'll find leaks that nothing else will.

Ah no, you're right, you can only test it with the gear off…

Now that you say it, I realize I meant the drive gear. But *I* knew what I meant.

removing that gear is not too difficult, if you want to use the oil & visual inspection method. Or the clutch breather/bolt swap trick; as HEC suggested is a good method too.

Thanks everyone.

I tried to devise a smoker using a beekeeper's smoker. I could not find anything with that, though perhaps I didn't get thick enough smoke.

So I moved on to the method putting gear oil on the bearing and pressurizing it. No bubbles. I figured I must have be doing it wrong, so I turned the crank slowly to see if I just didn't have the ports lined up to let pressure come through to the seal. No bubbles.

I know that there is an internal leak -- fuel in the crankcase oil confirms that. I am running on the assumption that I did not do these two tests right, and it is the clutch side seal. I will continue the teardown to see what's inside.

I have two questions for you:

1. Given that I just put this together with fresh seals, do I need to replace the other seals that currently are fine? (Flywheel, driveshaft.) I think the answer is no but I'd be mad if I was wrong.

2. Is there some other thing that could cause an internal leak, like not torquing down the bolts behind the flywheel tight enough, or anything else?

I'll have it open soon and will look carefully at all the mating surfaces. This is not a tuned engine, and I can't think of any precipitating event that would cause any internal damage I did not previously notice. Right now I'm just planning to replace the clutch seal and get it back together carefully, but your tips about what to look for are appreciated.

Hey there, spacecat.

To answer your questions:

1. If the other seals were just replaced, I don't see any reason to put new ones in. The clutch seal is the only one that would allow gas to leak into the gearbox.

2. Not torquing the bolts behind the flywheel tight could possibly allow leakage. The case gasket could also allow a leak if it gets mashed/folded/torn or slips as you're putting the halves together. Pay attention to the part that seals the crank off from the transmission. I always use Motoseal. It's never let me down.

Once you get the case buttoned up, I'd strongly recommend giving the engine a pressure test. If you can't hook up a smoke tester, just pressurize it with air. You can locate leaks by process of elimination.

+1 on SoCal's advice.

When I do a smoke test, I remove one of the plugs, then slowly pump smoke into the motor until it's filled with smoke. Then I replace the plug and pump more smoke into it.

Also, if you have a cheap laser pointer, it may also help you find where smoke might be coming out when it reflects off the smoke particles. Something that puts out a line, like a laser level, works even better.

You are not going to want to hear this but....maybe take it apart again and ensure the seals are good/installed properly? The second time goes a lot faster (ask me how I know).

You can do it while you wait for your sharks to charge.

I'm way ahead of you -- it's already out and I'm going to split the cases this weekend. (That's how I managed to do the test with the oil in the bearings.)

It is going faster, with good appreciation for what I learned last time. When I get it back together, I'm going to try the smoke test again -- thanks for the laser pointer tip, sdjohn. [ETA oops that was chandlerman]

But this time I will do it before I put the clutch cover back on. That's probably obvious to the veterans, but I hadn't thought it through and tearing it down and having it not turn out as expected directs attention in new ways. I did read several of the threads about air leaks and how to build various devices to test for them, but did not read anyone say explicitly to focus on leaks in the crankcase and cylinder first, and then after, the gearbox. So I'm saying it here for posterity and for people at their first rodeo.

I think you guys are enjoying your "Smoke Testing" a little too much!!

Hec

Takes a while to charge the ill tempers.

I started down a youtube rabbit hole of austin powers after that one, I forgot how funny that stuff was

I am losing my mind.

I split the cases and replaced the clutch side seal. It seemed fine, perhaps the lip was a mm or 2 above the edge of the case. The only thing I saw off was some of the Threebond sealant in shreds everywhere. [This is a story and I think it's a red herring. Local buddy recommended Yamabond and a motorcycle shop to buy it. They hand me Threebond, I ask is this Yamabond equivalent. Yes, they say. After I get the engine back together I learn it is the white Threebond that is a thin silicone sealer, not the Threebond people usually use. Now disgusting bits of what looks like silicone caulk on bits of things. I have seen Greasy rail against using silicone willy nilly and add this to the evidence.]

I put in a new seal and put it back together, going slow and deliberately over each step. [I used Permatex Permashield, the hylomar blue equivalent -- toss up with Motoseal mentioned above.] I pressure tested and had a little leak at the intake. Do that over and get it back together. I found a leak actually in one of the plugs I used. Replaced that and got a perfect pressure test -- needle stayed put for 90 minutes. [I now think that my previous pressure test probably was ok -- the leak may have been in the plug, because I found nothing with soapy water and the leak rate was the same as when I found that leaking plug this time.]

So I had high hopes when I got the engine back in and started it. Could not get it to run with the choke off. I checked the timing, fiddled with the carb. Checked the crankcase oil and it is green.

So, it's still leaking into the crankcase -- perhaps just when it's warm. I think the thing to check now is the torque on the nuts around the flywheel shaft. I am terrified of over-torquing them. I torqued them in a criss-cross pattern last time. Any advice on how to do this?

Any other ideas about what it could be, if not this?

This is a head scratcher.

Trying to imagine how you can get "a perfect pressure test -- needle stayed put for 90 minutes" and still get gas in the gearbox.

Seems your leak only appears when the carb's on and the engine's running.

Are you 100% sure you're what you're seeing in the gearbox is gas?

If a clutch or fly seal was leaking you'd have a racing engine at idle. My guess is a bad seal at the carb.

As for the case nuts around the flywheel, they need to be snug. Don't be afraid to torque them. If I remember correctly two are studs, the other two are bolts. They don't require much. I think spec is 4-6 ft/lbs.

https://archive.scooterlab.uk/quick-tip-2-stroke-drive-oil-seal-leak-down-test-feature/

Try the oil test

I second this.

As for tightening the case nuts, this is what I use. I start at 6, then 8, then 10 ft/lbs.

That's why I am losing my mind.

So, no I am not 100% sure that it's gas in the gearbox -- I'm trying not to delude myself (nah, couldn't be!) while also trying to not delude myself in the other direction (what if it is?). Honey colored motor oil is now greenish after about 20-30 minutes of operation over several days. Smells different -- chemical, not gas exactly. But it does not make sense at all that it's fuel unless something changes when the engine warms up. (I also did the pressure test with motor oil over the clutch side bearings before I put the clutch cover back on.)

So, setting that aside. Even if I run it with the choke on, I can't get it to idle slow. It is racing. It does sound like an air leak there. I am sure the inlet pipe isn't leaking at the engine because that was tight at the pressure test. The first thing I did was make sure the carb was seated all the way and the clamp snug.

The other condition that has changed since the pressure test is the exhaust. I used the exhaust elbow with a plug in it for the test; I had to remove that to put the Pollini on, so I cannot be sure that gasket isn't leaking.

Is there anything on/in the carb itself that could cause an air leak? It's the Dellorto 19/19 attached to the inlet pipe. I've cleaned it a couple times over these last few months. Yes I have the little felt packing there, though I think I will check the whole situation over again.

Right now, I thinking:
-check over the carb once more
-run the engine and see if it's not resolved
-if no, pop off the flywheel and check the torque while it's warm
-at some point drop the engine still in the scooter and redo the exhaust gasket, though I'd like to do these in separate steps for science

When I retorque the nuts, should I loosen them a little then re-torque, or just tighten/check them from where they are? (I read an article from this forum about torque accuracy but it seemed to be about higher values.)

The connection between the manifold and carb is a notorious source of leaks on smallies. Try sealing the slits of the manifold with RTV or Motoseal - not too much - then clamp the carb on. Let it dry before firing it up.

Did you redo the pressure test after finding the gear oil turn green?

Yes or not yet, depending on when you mean. Before I tore it apart again, I pressurized it to do the motor oil over the clutch side bearing test. No bubbles. I went ahead and opened it up because I figured it must be leaking somehow anyway and wanted to check the gasket. I wasn't able to identify anything that could have cause it to leak fuel into the crank case.

This time, no, I haven't yet. I will try these other things and then when I go to check the exhaust gasket do it then.

I'm stumped over what could cause discoloration in the crankcase oil so quickly besides fuel. Could new clutch plates do this?

No I meant doing the test after the latest oil check?

I usually pressure test again after the engine has been through some heat cycles, just to make sure the expanding hasn't separated/moved anything.

Might be overkill, but it's just good to know especially when carb jetting a new setup.

Doubt it's clutch related.

Im wondering if it's scratches in the cases or something along those lines where there's gaps in the gasket.

Not overkill if the alternative is losing one's mind or splitting the cases a 3rd time with no reason to expect a different result.

Are you absolutely sure it's fuel? Does it smell like fuel compared to fresh oil?
Are you using a green 2 stroke oil? How about trying one that's red instead?

I do find it takes no time at all for the oil in the gearbox/clutch to colour up.

Given that you're finding no more leaks, it could be an answer?

Well Ginch, that is the most elegant solution to the given facts and I hope it's right.

The oil smells chemically -- I wouldn't say exactly like gas, but not like fresh motor oil. So, no, I'm far from sure. I'm using Motul, which is green. What two stroke oils are red?

I'm out of town this week but will get back to sorting it this weekend.

There's Red Line. I have some left over at home but can't recall the brand, I'll check.

Isn't Ipone also red?

Looks like it. And, you can get Ipone Samourai Racing strawberry scented. I guess if you're racing Samourais, you want to smell your best.

I buy strawberry ipone just to troll other rally attendees.

Thanks everyone for the considered suggestions. I'm troubleshooting now on the hypothesis that there's an air leak at the carb. I will update you when I have results. But I have a couple questions.

First, I have this Polini exhaust:
https://www.scootermercato.com/core/media/media.nl?id=5789&c=643441&h=pA0M6FWGXf6pvYDNW6m4MegcthKmraZ6OakQNQ7GMGeFAPFg&resizeid=-6

It is dripping fuel-laden exhaust from the union by the spring. Is it normal for it to leak exhaust there, and if no, could this be part of my problem?

On the Dellorto 19/19 carb, does the fuel/air mixture screw go in for lean and out for rich? Or the reverse?

I saw this the other day and saved it... I was really surprised that this could be an issue. A number of people replied that it had made a difference to theirs as well.

Of course yours could be something else entirely but it's an easy fix if it works!