Little Nugget from Steve Wong on ECUS, early Tachos and 3.2 Swaps

[tdw28210]

Meant to post this sooner, but just getting around to it now. In finishing up my Mexico Blue backdate build my 3.2 started stumbling, jerking and eventually completely cutting out.

The guys at Sonderwerks fiddled with the motor a bit and then one of us had the bright idea (Charlie probably) to swap the ECU out of my '85 Targa into the backdate build. About 10 minutes later the job is complete and the '86 3.2 is running perfectly again.

So, I now know I need an ECU repair at this point. When I think "ECU" in the air-cooled world, I think of Steve Wong. Even though I've NEVER actually bought anything directly from him. Mostly because of all the beaming posts here and me buying his work 2nd hand.

Anyway I fire off an email and Steve fires right back. We agree to hop on a call together and I describe what happend and what I need (repaired ECU) but, I might as well get the Steve Wong upgraded chip put it in too.

In the course of our chat, Steve asks me about my gauges. I explain I am running the original, but "backdated" '77 gauges. Steve then goes on to share with me that he has experience with other mid-year cars with 3.2 motors and original gauges having ECU issues.

(Hopefully Steve replies here), but basically it sounded like the '77 TACHOMETER gauge fires a signal back to the ECU that can actually damage it. Not every time, but enough for Steve to be wary of it. He recommended that I get an actual mid-80's 3.2 tachometer in addition to getting the ECU fixed.

Welp, it's only money and I am so far over budget already, what's another few hundred bucks? I actually get sort of lucky and ending grabbing nice used tacho for about $200 bucks online,

Anyway, just something to be aware of for the hotrod/engine swappers out there.

3.2 motors and mid-year tachometers can create issues for your 3.2 ECU and the combination of the these is best avoided if possible.

(Just trying to pay it forward, a little bit at a time).

[911pcars]

Yes. Tech docs say using an early, non-Carrera tach will eventually damage the 3.2 ECU. Incompatible signal levels.

914World.com - The largest online 914 community!

The Tach-Adapt unit installs between the DME tach harness and an early tach to make a compatible pair.

http://forums.pelicanparts.com/porsche-911-technical-forum/627935-tachometer-adapter.html

Sherwood

[Matt Monson]

Did someone say tacos?

[tdw28210]

Originally Posted by 911pcars

Yes. Tech docs say using an early, non-Carrera tach will eventually damage the 3.2 ECU. Incompatible signal levels.

914World.com - The largest online 914 community!

The Tach-Adapt unit installs between the DME tach harness and an early tach to make a compatible pair.

http://forums.pelicanparts.com/porsche-911-technical-forum/627935-tachometer-adapter.html

Sherwood

Thanks for this reply. Might have a nice mid-80's TACO for sale shortly f this adapter will do the trick for my backdated '77 with 3.2

(Yes Matt, someone DID say "taco")

[Tom '74 911]

I would love to know more about this issue w/using older tachs w/3.2L ECUs. I have a SPEC911 w/a 3.2 that started acting funny at the end of last year. The tack looks to have the date 01/81 on the back, but I can't be sure it's not been modified. The 10K limit on the face is not stock at the very least.

At first I thought it was a fuel starvation issue when the motor started cutting out intermittently on what seemed like right hand turns only. Then it started happening more often and in other circumstances - left hand turns and accelerating on straights. I've pretty much ruled out fuel as the issue. The last episode happened on grid when the motor quit after idling for a few minutes. I had a hard time starting it after that incident for the first time ever... That pretty much ruined my weekend as the symptoms worsened from there.

Since then, I've installed a new DME relay and replaced/rewired the fusebox. I did open up the ECU and did not see anything obviously amiss - but I don't really know what to look for. It's tough for me to do any test driving during winter, but I'd sure like to find the issue before the spring track season starts...

Is it possible to test the ECU to see if it's failing and if that's the issue?

Thanks,
Tom

[Bergo]

Interesting, looks like I should buy a new speedometer and a tachometer ! Throwing a 3.2 into my ‘74 backdate (coincidentally, also Mexico blue)

[Bob Ashlock]

Interesting discussion...

What would be really good to know is what the actual output structure design of the 3.2 ECU tach signal interface looks like. With that info, a determination could be made about what kind of load (sinking or sourcing) it can tolerate without causing damage.

Some thoughts: There are a variety of 'early' VDO tachs out there, the most notable differences being the transition from the high-voltage triggered ones from pre-1972, which have a pretty high input impedance, and those after. From August 1971, the tachs were changed to trigger direct from the distributor points/CDI pin-C trigger signal which toggles between 12V (points open) and 0V (points closed). The 3-pin CDI has a low value 120 Ohm or 33 Ohm pull up resistor to +12V, capable of sourcing 300+ mA of current to keep the points happy and clean. Perhaps some versions of those tachs, designed to trigger from that 12V signal, had considerably lower impedance, and were capable of sinking or sourcing higher currents, but I have never observed that on any that came across my test bench.

The next big change was in 1978 for the SC cars which used the 6-pin CDI. Since that box was triggered by a reluctor pickup distributor, a separate tach drive (TD) signal was generated in the CDI. The tachometer provided a weak "pull-up" to 8V internally on the TD signal and the CDI 'pulled' that signal down at the rpm rate. Fairly low currents are involved, so I am surprised the SC tach could actually 'force' enough current back to the 3.2 ECU to damage it. BUT that being said, I've been wrong before and I don't have enough details at hand of the internal design of each version of VDO tach to be certain. Would love to have my hands on that to share because we all want to know about this stuff.

Also interesting to know is within some of the mid-70's versions onwards, the VDO design no longer used discrete components. A 'custom' IC appeared bearing a Texas Instruments marking and part number. I once investigated this with TI and was told they had no more info on that old part, and even if they did they could not share it as it was custom made for VDO. Since this part was not available, even to the repair shops (who I understand 'harvested' spares out of old tachs) Accordingly, I developed a small driver board to take the place of the factory VDO board. This little board has been useful for customizing tachometers, converting almost any version to different scaling (10K) etc. It provides a very weak 12V pull-up on the trigger signal through 10,000 Ohms, so max current it can source is only around 1 mA. It is doubtful that such a low current could cause any damage to any ECU source or sink tach signal. This is also the case for my TACH-ADAPT product that is often used to "buffer" the signal between the ECU and Tach as discussed earlier in this thread.

There, that's my brain dump for the morning ....

[FrankM_]

Bob, I would love to see a basic schematic of the internals of the VDO instruments but they are hard to find. If your question is on the Motronic side, let me know which part you want to see exactly (which pins), I do have a reference schematic of the 3.2 ECU.

[Bob Ashlock]

Frank, I'd love to have that Motronic schematic. Most important the tachometer output signal section! Bob@ashlocktech.com

[mysocal911]

The pin 21 (tach output) of the 911 3.2 DME ECM has an output resistor of 510 ohms (1/2 watt).
Pin 21 can be either shorted to ground or to a 12 volt supply without any effect or damage to the ECM,
either intermittently or continuously.

[FrankM_]

Ok, I’ll send you the schematic when I get home. Just for my understanding, it is pin 11 and 21 that go to the tach ?

[mysocal911]

[mysocal911]

Originally Posted by FrankM_

Ok, I’ll send you the schematic when I get home. Just for my understanding, it is pin 11 and 21 that go to the tach ?

Pin 11 is the up-shift input to the '84-'89 original tachometer, and should NOT be connected to anything other than that,
or the DME ECM will be damaged.

[Bob Ashlock]

Mucho thanks MySoCal and FrankM.

It appears that pin 21 provides a 'sink' through 510 Ohms. Not knowing for sure what's inside the tach, but even at 12V the most current this would sink would be around 20 mA. That represents as much as 1 Watt of power. They placed 1W resistor there (R500) suggesting that is the case, at least if something went haywire in the tach.

I know there are types of simplistic Tachometer Adapters available that are basically large inductors,(sometimes people even use a relay coil) used to generate a high-voltage reactive 'spike'. I can imagine if one of those was connected to this tach output port, the high-voltage spike that would occur when transistor T500 releases would damage something! That is a much different scenario than connecting a device that presents a resistive load or sources a voltage through a resistance. I doubt the latter, if a light load or source could cause that circuit any damage.

Oh, unrelated to the tach rate signal (pin 21), but now I'm curious about the up-shift signal .... is that just a logic-level signal ?

[FrankM_]

Bob,

file is sent. What I would like to know is if I can use a 930 tacho on my non-turbo MY85 ? I would like to 're-use' the boost pressure indicator to indicate the air fuel ratio from a wideband O2 sensor. Most of the widebands give a 0-5V analog output which I could easily transform into whatever the gauge needs.

If so , which partno of the 930 tacho can I safely use in my non-turbo Carrera ? Beter do the research up-front as these gauges go for 600 - 900 USD second hand !

[FrankM_]

Bob, on the upshift indicator (pin 11 I believe) ? It is a straight output of the Bosch custom IC 127, I think it is their custom peak & hold driver IC. Connected through a '0R' resistor.
The IC 127 is a kind of a mystery too, no datasheet, no documentation, no spare parts :-(

As a side project I am currently building a small injector driver (for testing & cleaning purposes) using the LM1949 to mimic the exact output my ECU gives these low-impedance injectors.

[mysocal911]

Originally Posted by Bob Ashlock

but now I'm curious about the up-shift signal .... is that just a logic-level signal ?

Yes, just a simple voltage change to indicate a certain engine load. A logic chip in the tach using that and RPM to turn-on a light in the tach, indicating a shift point.

[Jonny H]

Originally Posted by FrankM_

Bob,

file is sent. What I would like to know is if I can use a 930 tacho on my non-turbo MY85 ? I would like to 're-use' the boost pressure indicator to indicate the air fuel ratio from a wideband O2 sensor. Most of the widebands give a 0-5V analog output which I could easily transform into whatever the gauge needs.

If so , which partno of the 930 tacho can I safely use in my non-turbo Carrera ? Beter do the research up-front as these gauges go for 600 - 900 USD second hand !

Depends on the year of the 930 tacho. The tacho in the early Turbo (75/76) that used the larger 8 pin CDI boxes require a high voltage drive (similar to a 69/70). They won’t work with your ‘85. Bob’s adapter would probably fix that though.

I would imagine you’d be ok with the later 930 tacho although I haven’t tried this combination.

[mysocal911]

Originally Posted by Jonny H

The tacho in the early Turbo (75/76) that used the larger 8 pin CDI boxes require a high voltage drive (similar to a 69/70).

A simple internal mod to that tach solves the problem!

[ischmitz]

Let's do away with the myth here. An early tachometer is not "sending a signal back" or does anything bad that damages a 3.2 DME. The issue is something else:

A 3.2 DME has two outputs that sometimes get confused with each other. Pin 21 is the engine speed output and it is a buffered signal from T500. The BC546 transistor can easily drive any tachometer out there. More so, it requires the tachometer having an internal pull-up resistor to VBat. R500 limits the current through the transistor to less than 30mA even with a hard short to Vbat at 14.4V. This doesn't damage the DME at all as that transistor can handle 100mA continuous current.

The second output of the DME on pin 11 is a signal required for the shift light in the 3.2 tachometer. It comes directly from the Bosch 0127 IC and contains engine load information. And therein lies the problem. Once you connect pin 11 to anything but the shift light input of a 3.2 tachometer you are in trouble. The tachometer will initially work since the fuel pulses have the same base frequency as the ignition pulses. However, the 0127 IC gets really upset in the process (read: it destroys itself). This is due to the internal pullup resistor of the tachometer biasing pin 3 of that IC to Vbat. And within 10 - 15 minutes the DME will fail with no fuel signals. The higher the RPM the more stress it put on the IC and this affects how soon it fails.

I have repaired multiple 3.2 DME that were affected by this phenomenon. And trust me, a simple re-soldering will not fix that kind of damage.

Ingo