I've had discussions with JohnK about producing these and offering them on our site and I have decided to pass based on liability. My business is fully insured regarding product liability but one must manage risk. If someone home builds these and they fail (to the state that does not set off a SDL) and someone's car burns to the ground or worst, big problems to come. That said, I'd be very careful installing a non OEM unit. YMMV
I was thinking the same thing. I think these things crossed model development and were adapted along the way. I would bet they had cross functionality as far as the amplified voltage signal is concerned but it seems that the warning lights and possibly the bank shutdown capability was onboard them in the 348 model. I don't have a 348 to try things on but it might be worth the effort.
Based on the 348 2.7 diagram, the TCU provides only one signal, the amplified voltage, to the ECU (so that the ECU knows when to shut down the bank). The SDL is controlled directly by the TCU, lit at certain temp. point before the "shut down" threshold. In the case of the 355 2.7 (two connections to the ECU, one possibly not really functional based on the test results posted earlier), there is no direct connection TCU-SDL so I believe that, in this case, the ECU controls and decides (based on the amplified voltage from the TCU) when to switch the SDL light on and, if the temp. continues rising, shut down the bank.
I wonder if there is more than one iteration of the SDL circuit for the 348. Some people have posted that they have completely removed the control units without any SDLs or CELs showing, while others have said that they get lights when the control units are disconnected. I'm almost positive my car falls into the latter category. I will check next time I fire it up. The WSM wiring diagrams are for early cars. It's possible that the circuit was altered at some point to warn of a complete failure (or disconnection) of the control unit by illuminating the corresponding SDL. BTW Bob, thanks for the response to my previous question. I'll be in touch soon via PM.
Answering my own question from a few days ago, the following temps and actions are listed in the WSM: Motronic 2.7 SDL Flashes at 900 C (1652 F) Steady at 940 C (1724 F) Cutoff at 980 C (1796 F) (All temps +/- 20 C, +/- 68F) Voltage across ECU terminals 52 & 19 goes to 13 VDC at 980 C (cutoff), otherwise voltage across terminals 52 & 19 is 0. The WSM seems to imply that there is no cutoff built into the 2.5 cars but that could be an error of omission. There are plenty of mistakes in the WSM (shocking, I know, since that's not at all like the Italians).
This is good info. Voltage between 52 & 19 means 53 to ground as 19 is grounded. It means that the TCU first talks only to the SDL (the 348 wiring diagram I posted earlier is for the 2.7, shows direct connection of the TCU to the bulb) with no signal (voltage) to the ECU. Then, when the temp reaches 980C, the TCU just switches on 13V to Pin 52 of the ECU, i.e. there is no varied voltage signal going to the ECU. My 2.7 348 is a UK model, no cats, and I do not get any SDL or CHECK ENGINE when I disconnect the TCU. So, it appears that my marked-up 348 diagram is actually for the non-cat models as, based on how the TCU is connected, it does not seem to indicate that there would be any warning when it is disconnected (the voltage on ECU Pin 52 would be always zero, all's well). It is possible that the Cat 2.7 348's have the TCU connected differently so as to cause warning when disconnected. However, I do not have any different wiring diagram for 2.7 than the one I posted.
So, the TCU-ECU communication on a non-cat 348 is very simple for the bank shut down: switch +13V to ECU's Pin 52 and this shuts down the injectors. In the case of 355's 2.7 ECU, Pin 52 is left unconnected so the voltage on it is always zero (allowing the injectors to run). The signal from the TCU goes to Pin 44 instead, which is unconnected in the case of the 348. The question is what is the function of Pin 44. It is possible that this pin requires a varying voltage, not just off-on like the Pin 52. If this is the case, then the signal from the newer TCU is a varying voltage as compared to the off-on of the older TCU in which case the swap would be rather difficult, unless the 348's ECU can understand the newer TCU's signal when brought to the ECU's Pin 44.
Is it a full engine shut down in 348 as opposed to banks? Are both SDECU connected to pin 52? I guess same question and answer in 355s. How does the ecu know it's right or left bank with one wire and one pin? What voltage would illuminate the sdl?
Bob, it appears that each SDECU is connected only to its bank's Motronic ECU in the 348, so only one bank would be shut down. I'd think a 12VDC signal would illuminate the SDL, but according to the archive post you quoted that's not the case. It's something I hope to establish once you confirm what connectors are needed to make a breakout connector. Not sure how much info I can get from my existing control units but I'm willing to try. I've gotten random SDLs, but never a steadily flashing one. At least it should be possible to determine the SDL output voltage though.
Thanks Mike. I wasn't aware that there were two ecu in 348. That explains that. 355s have one. Have to delve into the wiring diagrams some more but the sdecu must use different pins or how else would the ecu know which bank is involved?
I understand the operation of the 348's old TCU to be as follows: 1. When the voltage from the ThC reaches the level equivalent to 900C temp, the TCU starts flashing the SDL; simple on-off circuit (chip) is activated switching the SDL bulb via a switching transistor, I guess. 2. When the ThC voltage reaches the equivalent to 940C, the switching transistor is caused to be permanently "on". 3. When the ThC voltage reaches the equivalent to 980C, another circuit switches +12V (or 13V as the book says) to Pin 52 of the ECU. Rather simple, I think. Of course, the very small voltage from the ThC is amplified first before it can control the mentioned switching circuitry.
Mike, I have a 1990 348 ts with known bad TCUs and I have run my car for years with those units unplugged, no SDL, no CEL. I borrowed a set of known good TCUs and installed them. Everything worked as expected. I would just like to have the proper TCUs in place and working. This is a very interesting thread!
From yelcabs pinout for 5.2 motronics: 2 input pins R/L SDECUs SDL output pin Image Unavailable, Please Login
Chuck, it is my recollection that the cars with SDLs (or was it CELs?) after unplugging the SDL control units were very late cars like Spiders. My car is a late '91 and I would not expect it to share that characteristic but I'm pretty sure it does, perhaps because it has had a host of factory updates applied. I will be running the car in the next few days so I'll confirm or deny at that point. I had hoped to be able to just remove the control units altogether and plug the TC ports when I finally get around to installing my test pipes so I unplugged a control unit once to make sure it wouldn't cause problems. My recollection is that it did cause a warning light, although I don't recall which one. I'm guessing it must have been a CEL since otherwise I'd just remove the SDL lamps when the test pipes go on. But my memory is bad and I doubt I took notes, so need to try it again.
I have a '89 with the 2.5 and the left bank light will at times light it's SDL in the flashing to solid condition and not shut the bank down. I have had it occur at startup as well as at operating temp. It is a malfunction of either the TC or the box because I have measured the cat temp at full operating condition with solid light on and it is normal. It has never been an issue to me because I know the cat is not getting to an abnormally hot condition and the engine doesn't shut down. I would venture to think the WSM is correct for the 2.5. Great thread with a lot of good info. I guess I have to get out and swap the boxes to isolate my particular issue.
I'm not yet clear on the distinctions between the two different 348 SDECU OEM parts that are out there. Now I'm hearing of a 4 conductor on the early ones. Can someone post a pic of the early SDECU ? It's possible that it didn't have the shut down capability and only lit the sdl.
For what it's worth: No SDLs with boxes disconnected (both wires disconnected). With them connected, the SDLs come on momentarily as part of the start up routine (key in, turn to on, before starting). Guessing this is to let you know the SDLs themselves are working. This doesn't happen with them disconnected so it would seem the SDLs are only related to the control station boxes. With the boxes connected, but not the sensors (one set of the wires only), the SDLs come on and stay on! Not sure about the CELs except to say been driving with the boxes disconnected while testing and they aren't on. So far anyway. This is on a 91 348ts. What about opening the boxes, refurbishing the electronics inside and then sealing them back up?
Oddly, I've never experienced a steadily flashing light on mine. Once it comes on it may stay on briefly and then go on and off randomly, or it may stay on for a long while, go out for a short time, and then back on. Seems to be completely random and also seems to occur only after the control unit has had time to get hot. Bad thing is, with the converters still in place it's pretty much the light/boy who cries wolf-- you never know if the warning is legit and you can only ignore it at your own peril. For what it's worth, a new thermocouple from Dr. Bob significantly reduced the incidence of SDL illumination on my car. In fact, for quite a while after installing the new TC the SDL did not come on at all, leading me to believe that the issue was completely solved. Bob has warned before that thermocouples become faulty more often than most people realize and based on my own experience I believe that. I have a fairly low mileage car in great shape and the factory TCs looked new but they were clearly out of spec by enough to cause a whole bunch of SDL issues all by themselves. I'm hoping to get rid of the whole warning system once I get around to installing the test pipes but I'd like to have the option of returning to a fully functioning stock configuration at any time. IMO replacing the original TCs is necessary for the system to work as intended. Fortunately, we have better-than-OE replacements available for those, but the control unit remains an issue. Fortunately, we have at least one group working on a replacement and as time allows I hope to do what I can to help.
Do you happen to know the part number of your Motronic ECUs? I used to think it might be related to later ECUs, although I'm not seeing how that's possible based on the diagrams as posted by Miro. But I'm still curious about your Motronics for more than one reason. I actually hope you're correct and am looking forward to having an excuse to fire up my car so I can refresh my memory. Maybe I goofed and disconnected the TC instead since you get an SDL with those unhooked. As to opening up the originals, I'd **love** to be able to do that. My electronics knowledge is not extensive enough to be able to design a replacement circuit from scratch but I can dang sure find the best available spec replacement components and swap them out (as I did on my speedo repair). Problem is the potting in those boxes is deep and strong. IIRC others have tried to defeat it and failed. I'd try heating one to see if it's possible to liquefy the potting without destroying the board/components, but since I don't have any spares I'm gonna bide my time. But if anyone has any outstanding ideas on removing potting I'd love to hear them. Maybe freezing with liquid nitrogen and chipping it out a little at a time? But what do you do when you get down to the board itself? The only potting Ive successfully defeated was pretty gooey, so Im at a loss as to how to deal with the plastic concrete our SDL control units are filled with.
In many years of electronic engineering and trying to reverse engineer competitor products I can confirm epoxy potting is near impossible to remove. But! It's a fairly easy circuit to mimic functionally. We have a few posts that detail the function of the unit, and thermocouple amps are a fairly standard and commonly available chipset these days. Just get the ambient temp offset voltage correct and it should be a matter of following the chipsets datasheet recommended circuit. For the 348's that require a ~12 volt signal to enable bank shutdown a comparator tied to the thermocouple amp output set to switch at the right point could do the trick. Has anyone considered older 348 CCU's are open collector and essentially create a ground path for the warning bulb? This would explain why some have success simply removing the box on older 348's while my 2.7 355 has the warning light on permanently with the CCU removed. It certainly appears 2.7 and 5.2 motronic ECU's are a little smarter with their inputs running proper A-D conversion on the signal line (smarter but still easily defeated with a bit of know how). I've easily defeated my 355 slow down warning lights with a constant voltage signal, but I don't have cats. If I still had cats I would spend a weekend designing a new circuit but unfortunately I have too much else going on to help the team out here just at the moment. There must be a few other electronic engineers or skilled hobbyists on this forum that could pick his up. Happy to offer advise to any willing developer.
Based on my quick look at the wiring diagrams, the 2.5 348-s have the following arrangements: Non-cat version - no ThCU-s, SDL connected to the ECU; I believe the ECU lights the SDL up based on the engine temp as there are no Thermocouples on the exhaust. Cat version - two ThCU-s with 4 pins (+2 separate pins for the Thermocouple); only 3 out of 4 used: +12V, ground and connection to the SDL; no link to the ECU so it appears no bank shutdown function, only Light SDL. 2.7 348-s came with ThCU-s with 6 pins (+2 separate pins for the Thermocouple) on both, Cat and Non-Cat, and had the additional bank shutdown function.
