FLYWHEEL TRI-TACH -- I5 AUDI -- 135+1

romano · Post by **romano** » Mon Jun 03, 2013 2:26 pm

Hi to all,

After many tests, I must ask a support because I can't find by myself!!!

I hope than you can support because I'm a little bit disapointed on my engine reaction...

Let me describe the configuration:
MS3+MS3x, firmware 1.2.1, format 0262.07, TunerStudio 2.0.8
Stock sensors used and wired as follow
Single pin on flywheel to TachIn
Wheel speed on flywheel to PT4
Cam Sensor to Cam In on MS3X. The cam wheel was modified and have now 5 cuts (instead of 1 on stock wheel) due to COP installation.

The issue is the following: ECU doesn't see any TDC signal.
Have you an idea of the origin of this strange signal?

As you could see the arrangement of 135+1 on FW is not clear!!! Cam signal too

A 60-2 crank wheel adaptation is the only real good solution?

Thanks for your support!

Romano.

Post by **jsmcortina** » Mon Jun 03, 2013 2:58 pm

The composite logger gets a bit tricky as it only tries to record two channels but there are three in tri-tach.

However, even if that's ok, I doubt you'll get it to work because of your modified cam signal. EDIT: i.e. the code is requiring a single cam pulse for every two rotations of the engine.

Any suggestions of what cars/engines came with this tach arrangement in the UK? Does someone want to give me a running engine to test on?

James

elaw · Post by **elaw** » Mon Jun 03, 2013 4:27 pm

James,

Pardon my cluelessness / bad memory, but can you help unconfuse me as to which sensor is considered the "tertiary tach input" in this scenario?

I notice Romano says he has his ring-gear sensor connected to PT4, and in my notes I have mine connected the same. But on the ignition options menu, the selections for "tertiary tach input" are MS3X cam and JS10. I thought the ring-gear sensor was the tertiary input?

Romano: one thing I notice right off is in your ignition settings you have "cam input" and "tertiary tach input" both set to "MS3X cam". That definitely doesn't seem right!

Also, I assume you know the ring-gear sensor needs to be fed through a VR-conditioner chip - it cannot be connected directly to the Megasquirt processor?

I'm not sure if the composite-logger image you posted accurately reflects what the MS is seeing, but if it does it seems like the signal from that sensor (the ring gear sensor) is not coming through cleanly. I can tell you from experience that the signals from those two sensors are *very* finicky, and you often have to fool around a lot with your VR-conditioner circuitry to get clean signals.

And lastly I'll reinforce one thing James said: a 5-window wheel in the distributor is not a good idea. It's possible if your distributor is positioned just right it could work in "polled cam" mode, but in general the 1-window wheel is a better choice, especially given that the tri-tach code is somewhat experimental anyway.

romano · Post by **romano** » Tue Jun 04, 2013 4:17 am

more data concerning this config:
Engine: "3B" 5 cylinders on an Audi S2.

Sensors:
Cam wheel: hall sensor wired directly to Cam in of MS3x
Crank Wheel
Speed wheel ring: VR sensor wired to VR2 conditionner (supplied by JBPERF) then to PT4 of MS3x (appear as "MS3x Cam" in TS)
TDC Pin: VR wired to VR1 conditionner (supplied by JBPERF) then to TachIn of MS3.

Cam wheel modified as picture hereafter.

Stock wheel have only one window.

To be honest I don't understand how this modifed wheel could disturb the main wheel information (the crank one).

thanks for your support.

R.

elaw · Post by **elaw** » Tue Jun 04, 2013 5:02 am

Well I'll leave it to James to comment on the cam wheel - it seems to me in "polled" mode it could work (again, only if the distributor is positioned correctly) but in edge-triggered mode it definitely would not. Honestly to eliminate any possible issues, I'd recommend you replace that wheel with the OE one if at all possible.

But you are correct that the cam wheel would not cause the crank trigger signals to look wrong. What makes those look wrong is that the Audi VR sensors will not just hook up to a dual-VR board and work - especially the new v2.1 board (the small one that's about 3/4 inch square). The Audi VR sensors have a higher impedance than most VR sensors, thus they put out more voltage and are also more prone to noise pickup. And the signal from the pin sensor has a very low duty cycle, which causes problems with the "auto-zero" (I think that's what it's called) function in the signal-conditioning chip used on the v2.1 board. The LM1815s used on Jean's older VR boards can have this issue too, but there's a jumper you can install to disable that mode which helps. But the bottom line is you're probably going to have to fool around with putting resistors in series with the VR sensors to match their signal levels to what the VR board expects, in order to get a clean signal.

romano · Post by **romano** » Tue Jun 04, 2013 5:59 am

elaw wrote:... will not just hook up to a dual-VR board and work - especially the new v2.1 board (the small one that's about 3/4 inch square).

it's exactly the board that I have: the small one. Work fine on other engine (last one was a Fiat 5cyl too. All issues were solved with this board.

elaw wrote:But the bottom line is you're probably going to have to fool around with putting resistors in series with the VR sensors to match their signal levels to what the VR board expects, in order to get a clean signal.

I could avoid lost of time if I directly go on 60-2 or 30-2 wheel?

R.

elaw · Post by **elaw** » Tue Jun 04, 2013 6:47 am

romano wrote:I could avoid lost of time if I directly go on 60-2 or 30-2 wheel?

R.

You definitely could... if you can find a good place to mount one, along with the pickup! If you can spend some money, 034 motorsport sells everything you'd need, in particular this damper pulley with a trigger wheel attached: http://www.034motorsport.com/034efi-eng ... -p-50.html

But before you do that, you might contact Jean and ask him. I think there are a couple of other people using that board on Audis and he might be able to tell you exactly what's needed to make it work, or refer you to people who know. Although the 60-2 setup works very well, installing a couple of resistors is a lot cheaper and easier than replacing a damper pulley, especially on an Audi 5-cylinder!

racingmini_mtl · Post by **racingmini_mtl** » Tue Jun 04, 2013 8:04 am

Unfortunately, I rarely get feedback when things are working correctly so it's usually easier to know what doesn't work than what work. In this case, I don't know what setup would work with the dual VR board. But as Eric says, the low frequency of the pin sensor signal does mean that the adaptive function is actually more of a problem than a help at low RPM.

You will need to put a shunt resistor between VR+ and VR- and try different values to see if you can find a value that will allow you to have a good signal when cranking. Once the engine is over about 700 RPM, the adaptive function should start working as intended.

Jean

Acki · Post by **Acki** » Tue Jun 04, 2013 2:52 pm

Flywheel with 60-2 trigger wheel are also available. ;)

Audi V8 has same setup with 136 tooth.
Months ago I made a thread with same questions.
Real solution wasn't available - or I understood wrong.

etsola · Post by **etsola** » Sun Feb 21, 2016 11:16 am

Hey Romano, were you able to make this work without installing a 60-2 trigger wheel ?

I have a CQ qith a 3b engine and am considering the Ms3 + Ms3x, but I can't seem to find a clear answer as if it's going to work well or not.

Cheers

jawier130 · Post by **jawier130** » Thu Jan 12, 2017 5:17 am

Has anyone managed to run the MS3 in an audi 5 cylinders?
I mean with the original sensors and the tritach-flywheel

jawier130 · Post by **jawier130** » Thu Jan 12, 2017 5:19 am

With the corresponding lm1815............

jawier130 · Post by **jawier130** » Thu Jan 12, 2017 5:21 am

I've been trying for a couple of days but I can not get it to work

elaw · Post by **elaw** » Thu Jan 12, 2017 5:38 am

I've done it... with great success (after a lot of initial headaches). But not with the native MS3 tri-tach code, I use a chip I got from Jean at JBPerf.com to convert the two signals from the flywheel to a single one that looks like a 45-1 tooth crank wheel.

I've made a lot of posts about it here... use the advanced search with my username to find them.

One big hint: the VR sensors that Audi uses are nonstandard and finicky... you have to get the LM1815 circuits set up "just so" for them to work. I did eventually manage to have a pretty reliable* setup using them, but have since switched to Hall sensors and ditched the LM1815s and it works even better* now.

* "Pretty reliable" = 1 or 2 sync losses a month on average in daily driving... "even better" = none in over a year.

jawier130 · Post by **jawier130** » Thu Jan 12, 2017 5:41 am

Ok,but I want to know if the tritach code really works

elaw · Post by **elaw** » Thu Jan 12, 2017 5:49 am

I tried it a long time ago and it generally worked, but I'd get fairly frequent (several a day) sync losses that I didn't get when using the hardware divider*.

If you're sure the signals coming out of your 1815s are correct, try switching the IRQ-1 and JS10 connections. There's a lot of confusion with the Audi flywheel sensors as to which is which.

* Jean's chip technically isn't a "hardware divider" either - it's a small microcontroller chip programmed for this purpose. But I have used a true hardware divider too and didn't get sync losses.

Megasquirt Support Forum (MSEXTRA)