Balancing S38 ITBs

[wjtesquire]

Wow! Dean was kind enough to share the Mantis Magic Manometer with me and look at the results:

Before:

https://www.youtube.com/watch?v=x91Dia4 ... load_owner

After 5 minutes of adjusting:

https://www.youtube.com/watch?v=nINNnME ... load_owner

I can't tell you how much smoother and more refined my car is running now. I will report back with the MPG data as soon as I can verify it. I highly recommend every S38/M88 owner use this tool. Huge improvements in idle and acceleration.

[wkohler]

Wow. That sounds smoother too!

[tn535i]

My car will sound a bit like the first video if the idle is too low (700 ish) and more like the second at a slightly higher idle (900 ish). I wonder if the idle speed also came up before and after. The S38 is supposed to idle a little higher than most think it should based on other motors. But the balance (or lack of) is really apparent and much improved. I've done mine differently (2 at a time, many measurements/adjustments) but I'm due for a valve adjust and balance and I need to get set up like this.

Good stuff.

[wjtesquire]

I had my idle set higher so it would not struggle when the ac compressor cycled on. With the compressor off, I would see the idle speeds of about 900 rpm. When the a/c compressor cycled on the RPM would dip to about 700. After sync I set the idle to 700 and it doesn't change when the a/c compressor cycles on.

[tn535i]

That is really smooth then for 700 rpm. Nice. But I think BMW would say the idle speed on the S38 should be closer to 800-900 range. It also sounds like it is not dithering but the clip is only a few seconds. Did you look at the O2 output to see if it dithers from rich to lean every few seconds? If not you are supposed to adjust the AFM air bypass until it can dither.

[wjtesquire]

I just calculated my fuel consumption over the last two tankfuls after using the manometer and averaged 16.6 miles per gallon. That is up from 14.3!

[Bhart]

Would a motorcycle synchronizing tool like the http://www.motionpro.com/product/08-0411 do a similar job as the MantisManometer?

[HealeyBN7]

Would a motorcycle synchronizing tool like the http://www.motionpro.com/product/08-0411 do a similar job as the MantisManometer?

Yes, but you need six ports. Not many six cylinder bikes out there:)

Dean

[tn535i]

I've done it with just two port manifold. Almost like using a vacuum gauge but you balance 5 to one master cylinder that you pick, #1 for example. You might have to go through a couple times but you can certainly get it quite good. You'll want some sort of restrictor to take the bounce out and you must be sure to clamp off any tube you remove or the fluid may get sucked out. ATF is good to use because it's easy to see, low viscosity, and if you suck it into the engine it shouldn't hurt anything.

The six port manifold is probably much better.

[Randomg]

Made my first attempt with the mantis this evening. Boy were my cylinders off. I was able to even them out with the cylinder screws, but it looks like I'm going to have to tweak some of the other variables to get it right, as my vacuum average isn't quite right, and my idle speed is quite high.

My valves are adjusted, I set my butterfly valves with a dial indicator and all are resting on the stopper, my adjustment screws have been cleaned out and the seals replaced, and all the rubber intake boots have been replaced. There aren't many places left to have an intake leak. I currently have my idle speed screw tightened all the way and my idle speed is still high with one of the individual cylinder screws tightened all the way down, so I'm at the limit of the screws adjustment range.

I'm thinking 3 variables are left: intake leak (either in something that affects all cylinder, or in the boot or few seals that each cylinder intake has), butterfly valves are adjusted wrong allowing too much air to go past them when they are closed, idle valve is malfunctioning (stuck fully open), or there is some ECU variable at work (the war chip does have a tab for idle speed). I have some investigation to do this week/weekend.

[HealeyBN7]

Double check the butterfly gap specs. If you read back through the thread I think there was some confusion on how it was quoted in the instruction sheet. It is not clear enough.

The gap should be .1mm-.15mm. That's about .004in - .006in. Mine was happier on the "closed" side of the adjustment.

Here is a photo of the service manual.



Sorry for any rework this caused you. I'll get the pdf updated.

Dean

[Randomg]

I had done the adjustment in the past, so I didn't use your part of the kit for it. I'll go through the butterfly adjustment again and hopefully that fixes it. If memory serves me, I already adjusted it to as closed as possible (I've had trouble getting my idle speed low enough since the rebuild), but I will try to do it more carefully with your tool. I think I'll just put them all at full closed (butterfly resting against the intake runner) to see if I can get the idle lower than 800 rpm to verify that there's not something else that also needs to happen to lower the idle to the right rpm. I'll try to do that tonight.

[tn535i]

Have you checked your TPS to make sure it is sending the idle signal or even jumpered it on the harness side to force the signal. Does the War chip still have the provision to jumper the TPS input and put the computer in open loop mode without idle control. This is where you should be balancing everything anyway.

I'm not a fan of micrometer adjustment of the mechanical linkage. I would unplug the TPS to see where the default ICV position allows it to idle. If too high (>1000) I will close down the t/b's and sort of balance the t/b's with a gauge before fine tuning. I'll bet this is where you are right now. If you have adjusted the t/b's you'll loose it as soon as you tweak anything mechanical and have to repeat. So if that's the case just start over. I would set all the bypass screws about 1-1.5 turns open then balance the mechanical movement of the t/b's using a vac gauge and trying to hit about 900-1000 rpm on a warm engine with the TPS inputs bridged to run in open loop. Once the mechanical adjustment is about right you can proceed back through the individual adjustments. Then remove the jumper on the TPS harness and adjust the mixture on the AFM bypass to get slightly rich. Then re-adjust the TPS switch if needed and plug that back in. It should now be in idle control and dithering and you can use the big idle bypass screw to get the speed you want, but it should be about right.

If the t/b's are too far open the ICV will not close enough to bring the idle down and this may be where you are now.
If the t/b's are too far closed the ICV can't open enough to raise idle on a cold engine and keep it running right.
You need to be at the proper mid point in the ICV range or actually just above mid point with ICV slightly closed on a warm engine IMO.
I don't think you ever get there with measurement of the t/b's.

[Randomg]

My tps works, and I was doing the tuning with it jumpered. I think setting the bypass screws to the default position and using the butterfly stoppers isn't a bad idea to get it close, but another thing to keep in mind is that I'm not just adjusting idle here, ideally you want the throttles all in the same position so that at all other throttle positions are as closely matched as possible. I think the two goals with the mechanical linkage are that the butterflies don't rest against the intake wall (slowly wearing down the aluminum, or potentially binding and getting stuck), and setting them all to the same angle. If their resting angle is not to spec but you can get your idle rpm, afr, vacuum and wot angle correct, I think that's the way to go.

After writing the above I think my current plan is: set bypass screws and idle speed screw to default, adjust using the butterfly stoppers to get the vacuum readings as close as possible, then go in and measure the butterflies and try to set them to the average reading (so if 1-2 is .8, 3-4 is 1.1, and 5-6 is 1.2, I'll try to set them all to 1.03), then check WOT to make sure they're close at the other end of their range, and then use the bypass screws to fine tune. I think this should work assuming that I don't run into the limit of the bypass screw range (maybe I should start with more than 1 turn when doing the butterfly stopper adjustment), but it's possible that one of the intakes is worn so much that in same position as one of the other sets, it just lets too much air flow past the butterfly to let it stay at the same angle as the other two sets.

[tn535i]

I think you have a good plan. Absolutely agree you don't want any of the throttle plates completely closed so checking that out is a good idea. I think if they were fully closed you would need to open the bypass screws way far though. I think the adjustments you'll be making won't make a whit's difference at WOT, maybe at cracked throttle. For that try using the manometer both at idle and with the throttles all just off idle slightly, maybe 1500 rpm or so with a load on it from A/C, and if all is well you should get good balance both ways. If not you may have some slack to dial out of the linkages. Of course all that testing with bridge on TPS so it will not try to dither.

I think we may all measure the closed throttle position slightly differently and that is why I don't bother anymore. Factory trained with factory tools might work but I don't have that option. I tried my best the first time around I did this and got poor results because with the TPS unplugged my engine ran way too fast, maybe 1500 rpm. So I had to close down the t/b's ever so slightly with mechanical linkages in order to get the control range back and also off idle transition right, I think that's about 1000 roughly with TPS unplugged.

The M30 has the exact same issues and I follow the same procedure. Warm engine, unplug TPS, adjust throttle to about 1000 rpm in open loop, adjust and plug back in TPS. On the M30 you are done at that point... easy... maybe check the AFM bypass air.

[Randomg]

Tried again today, followed tn535i's plan, set all the screws to default and used the butterfly stoppers to get them as close as possible first, then tuned with the bypass screws. When I was done and had the right vacuum on the gauge (little less than 30) and all of cylinders at the same place, I was back to where I was yesterday, which is that my idle was still too high.

With the tps connector jumpered, the idle speed was right around 1000 like you had suggested. This was with the cylinder one bypass screw tightened all the way and the 1-2 stopper set to right as it started to move from resting against the intake manifold (I can probably get it more closed with a dial indicator instead of eyeing it, which I'll do tomorrow). I also had the idle speed screw tightened down all the way. I expected to see the idle drop when plugging the connector back into the tps, maybe my idle system isn't working right, I guess I'll check the tps is grounding correctly tomorrow as well.

Not really sure what's going on, I'll go in with a dial indicator tomorrow to make sure the 1-2 set are as close to closed as possible and try again.

I think you're right about WOT, but even if you weren't, I don't care anymore, I'd rather have a good idle than a perfect WOT at this point.

[tn535i]

When the TPS closes at idle is it changing anything? It sounds like you should test it next to be sure you're getting an idle input. First check for voltage at the outer pins, 5vDC IIRC, and then the center pin should be ground (0v) on the connector. Backprobe the connector and plug it back in the TPS and 5v on the idle input (pin 1 I think) must go to 0v when the throttles are closed for idle control. It may just need adjustment or maybe the switch is bad. I've seen several of these TPS have bad solder connections inside after years of thermal cycles.

Idle control on the S38 has always been a bit of a mystery to me and I would not be surprised if some act differently depending on the chip (most are chipped). If the idle loop can 'control' why do you need to adjust the bypass screw on the t/b's anyway. Balancing the t/b's make sense but not adjusting idle. My big screw seems to do little when the TPS is active. In fact I've always thought maybe the instructions I've seen were wrong and should tell you to first unplug the TPS then use the big bypass screw to set the warm idle up to about 900-1000 so that when you plug the TPS back in it controls properly. Sometimes this is in essence what I do. Then when I plug the TPS back in the idle drops a little to about 800. But maybe that's my specific chip ?

Not sure how you would adjust perfect WOT, almost no vacuum and really 80-90% open is probably flowing the same air as 100% open. Above 3500 rpm or so the S38 is just open loop and dumping fuel in based on the AFM and speed map, and making nice sounds. I think if you do get idle right the rest will be fine.

[Randomg]

Went through the whole procedure (I guess I should have done that in the first place) and had better luck. I was able to get my idle down to around 1000 with throttles sync'd at the right vacuum. I think my idle control valve isn't working correctly at this point, but I think I should be ok to fix that later and use the large bypass screw to adjust if necessary without ruining the vacuum syncing of the individual screws. Cylinder 1 and the large bypass screws are tightened all the way, but I guess that's fine if all the measurements are right.

What I meant by perfect WOT is that all butterflies are at the correct angle when resting against the WOT stopper.

For what it's worth tn535i, tuning the idle by the butterfly stoppers didn't end up working well for me. During my drive the on/off throttle transition was worse and my wideband showed noticeably different AFRs at part throttle. Maybe my engine required more variation between the throttles than yours did?

I'll be shipping it back to Healey tomorrow.

[tn535i]

This S38 is a tricky beast IMO. Not the smoothest idle and basically ready to make power with just a crack of the throttle. Isn't that what the ITB's are all about.

It sounds like something isn't right with your idle control though. Maybe the TPS or maybe a sticking or bad ICV? For example, unplug your TPS and see if any on/off idle transitions disappear or anything changes, mine will do that. But when the TPS and ICV are active some slight transitions can happen as you go off idle and the ICV goes back to a home or neutral position. I find it worse on deceleration.

The procedure as I read would have you probably set the t/b's for a slightly open or higher idle with TPS unplugged. Opening the big bypass screw will let more air through and should make the computer close the ICV. If the ICV is quite closed and then goes back to about 1/2 as it comes off idle control at the same time the throttle opens it might make for a bigger bump in speed that you want. Is that what you have now?

If the ICV is more open and closes as it comes off idle you might even get a dip in speed. In this case you will also get a bit more engine break if you close the throttle while in gear as the computer tries to close the ICV all the way down to reduce speed but the drivetrain is trying to turn the engine faster. You'll feel that when you lift off throttle in gear. If the ICV is too far open on normal idle it won't have the range it needs for a cold idle either. So this probably worse than the previous.

Somewhere in between is the best compromise. I'm not sure there is a complete win on the S38 as the computer only knows to try to control or not control idle and the range of the ICV is what it is. I don't think there is any logic for anything based on road speed or gear ratio like a newer car. It may take some experimentation to find the sweet spot you like.

I've been tempted to put a scope on the ICV and watch it's open/close times and transitions. That might be a better way to get the base throttle set about right on a warm engine and then balance from there.

Also, I think as long as you don't adjust the t/b stops or individual bypass screws you are not going to loose your balance made. And what I meant on WOT is the minute adjustments at idle mean almost nothing at WOT, certainly < a 1 degree at idle and the difference of 99% vs 100% is nothing at WOT.

[Pattonky]

I just got through syncing my throttle bodies with a carb tuner. Only 1 and 6 were off at all. 1 was a little high and 6 was a little low. This made a big difference in the idle. The car seemed to kind of cycle before the adjustment and the idle fluctuated from about 870 to as high as 1050. After adjustment she rides more steady between 890-950. I don't think the adjustment of the throttle bodies altered the RPM so much as it allowed the motor to maintain a more constant rpm. RPM taken with Fluke 88. I suppose I am just kind of posting with the hope that this might be useful info to somebody.

[bobby in Derby City]

I just got through syncing my throttle bodies with a carb tuner. Only 1 and 6 were off at all. 1 was a little high and 6 was a little low. This made a big difference in the idle. The car seemed to kind of cycle before the adjustment and the idle fluctuated from about 870 to as high as 1050. After adjustment she rides more steady between 890-950. I don't think the adjustment of the throttle bodies altered the RPM so much as it allowed the motor to maintain a more constant rpm. RPM taken with Fluke 88. I suppose I am just kind of posting with the hope that this might be useful info to somebody.

Fluke 88 and a carb tuner!! I should come down I-64East to see ya so we can do mine ...

[Pattonky]

Are you having trouble with your idle or are you just pulling my chain over the tools to do the work? I will confess that I am a believer in having good tools to work with when I do something.

[milarsky]

Dean,

I was lucky enough to meet Alim this summer, who lives in ABQ. I am in Santa Fe for the summer. He currently is tuning his throttles using your Mantis. We threw it on my car, which runs very well. The throttles were very uneven and after 5 minutes of adjusting them, it was amazing. The thing runs like a scorned dog! What a difference. I cannot thank you enough for sending this thing around. Now we are working on Alim's car, on which we have some baselining to do over this next week.
Many thanks,
jeff

[alim_h]

Thank you Dean. I just finished getting my throttles adjusted too! Worked out very well and I did notice a difference at idle and just off idle.

The Mantisometer was simple to set up and use - genius idea! Thanks for loaning it out.

I have some pics and videos here: https://www.dropbox.com/sh/sh8u2e3g8gmu ... ZISJa?dl=0

[HealeyBN7]

Great to hear that it worked out. After getting it back I replaced the tubing as it started to yellow and took it and my M5 to the SoCal Vintage BMW meet for a little show and tell. Hooked it up to a couple S38s but the owners weren't brave enough to turn any screws. I don't blame them. Parking lot tuneups never end well. You really need to work through the mechanicals (valves and throttle plates) first. This is the last step.

It was sent to Matt - Defacto in CT after the show. UPS says it arrived on the Oct 13th. Matt, if you see this message, or the PM or my email, can you confirm that you do indeed have it and let me know when you are planning to return it.

Of course if you have any questions post them here or PM/email me.

Thanks,
Dean

[MicahO]

Hi Dean - if it's made it out to the East Coast and doesn't have another date booked, perhaps I can pick it up from Matt and go through the process with my e28. Maybe even the Touring? Should be physically the same process, I think.....

Maybe get a balancing-fest together? November gets relatively open, in-between sports seasons for the kids means the weekends open up a little.....

[MicahO]

So I'm done with the use of the Mantismanometer on both the e28M5 (s38B35) and the e34 M5 Touring (S38B38), and it went very well! I went through cleaning the needle valves and new orings, valve adjustment and butterfly settings in the run-up to the Mantis' arrival. I had also roughly set the vacuum on each intake with a gauge - new Oring, full-tight, then out one turn, then set to a rough-bouncing-about-9-inches-of-vacuum with a gauge and a pinched hose. Even that was an improvement for both cars. Then with the Mantis local and a couple of decent weather days I got the real balancing act done.

The touring started fairly balanced right from the rough-set of one-turn-out. On the e28 I did end up moving the screws more than I thought I would to attain leveled readings. The e28's screws were highly variable before I removed them, two almost totally closed and one about 1.5 turns out (that was the factory set with yellow caps). Position isn't what matters, the level of vacuum is, but it did seem strange that there was so much variance in how tight the needle was set.

Also, I'm really not sure what I was smoking, but I did connect the hoses 1-2-3-4-5-6 front to back on both cars, and I'm almost certain that when I adjusted the touring, the cylinder 1 port moved the right-most line of fluid, and on the e28 it adjusted the left-most.

Either way, I got both cars nicely set. The e28 in particular is now more manageable on-off throttle than it had been, and has back all of the immediate off-idle 'pop' that it had been missing after a tank of bad gas on the way home from TedFest in June. It's FAR smoother idling around at parking-lot speeds, and even sounds better. The B38 in the Touring is a far different engine management, which may be why the difference is not *as* noticeable, but there is still a notable improvement all-around.

Really great tool, and MANY THANKS TO DEAN for making it available to the community!

If there is anyone relatively local (NY Tri-State) that wants to do a round of tuning please let me know ASAP. Dean has someone lined up on the Left Coast as the next destination. If I don't hear from anyone semi-local in the next week or so I'll send it westward-ho!

[Valve_clatter]

Now that I have finally finished with the exhaust, and purchased all the requisite other bits means I actually should probably do this...

[BenGerman]

What is the current schedule for the Mantismanometer?
Any chance to get a hold of it before the Vintage?

[HealeyBN7]

Ben,

chances are good. It is ready to ship. Send me your contact information (real name, shipping address, phone number, etc) off line and I'll get it in the mail this week.

I have had to refurb the unit a few times, so if I am pivoting off the free rental program to now include a use fee to cover incidentals.

2017 pricing is as as follows:

$25 to get it shipped out.(USPS) plus,
$100 deposit (includes use fee) for a total of,
$125 due at time of rental (pal pay to healeybn7@yahoo.com)

Two week rental time unless agreed-to up front.

You will have to pay for shipping it back to me and when returned, you will receive $75 back on paypal.

I can send you an updated pdf that includes corrections and tips from others. Be sure to read the pdf before ordering the Mantis as you may want to have a set of o-rings on hand before starting. This is the last step in the tuning process, so valves should be dialed in before hand as well.

Dean