355 Replacement slow down cat ECU prototype

This is great news, GothSpeed produced a similar thing, I run a pair on my 355, but they seem to be no longer available.
Look forward to the results, good luck
Mark

 

Awesome!!

I have been involved in this arena for a few years now and your posts on this have been very astute. I would expect a well done product.

I’d be happy to beta test for you. Let me know. My car is a SDECU lab with outputs tapped with leads in the cockpit already from all three locations.

 

Dr Bob would be a great beta tester.

 

Would these be compatible with 348?

 

Amazing. Looks like you've put a lot of effort and expense into this project.

Would the heat from the chemical reaction of so much epoxy affect the components?
What are the main functions of the epoxy? Waterproofing? Insulation? Reinforcing the plugs or just an easy solution for a backing?

If epoxy insulates, it that a bad thing for some components due to heating issues?

Adding epoxy just seems to make the units hard to repair.

 

Yuuup he's the slow down light dude for sure

 

No, the 348's SDECU has very different outputs.

 

The kind of heat generated by the epoxy during setting is nothing to be concerned about, way below the heat that the components are exposed to during soldering - about 200 C. For example, the whole BGA chips (chips attached to the PC board by solder balls) are heated to some 230 C during soldering.

As for the repairs, I agree with you. I much prefer "black boxes" that you can open and access a clean PC board.

 

Unlike the 355 SDECU which sends a variable voltage, i.e. 0.5 V while the Thermocouple voltage is below 5 mV and 100 times the Thermocouple voltage thereafter, to the ECU, the 348 SDECU does not send any variable voltage to the ECU but processes it by itself. After the amplifier of the Thermocouple signal (which may be similar to that in the 355 SDECU), it has circuitry that triggers at least one relay that I could hear. Maybe there are two relays inside but I do not know exactly as I was unable to expose the internals (would have been easy had they not been epoxied). Anyhow, during bench testing (and reading some info in the manual), I have observed/concluded the following:

1. At ignition "on", a relay inside is activated for ~ 1 sec. and sends a pulse of +12V to the ECU (see the diagram below showing to which ECU pin) and to the SDL (for the bulb test). This brief pulse to the ECU tells the ECU that the SDECU is functional. If no pulse, the ECU will generate a fault code. However, in the case of my UK spec non-cat engine, the fault code does not effect the running of the engine in any way, it just sits in the ECU. Also, if the SDECU is completely disconnected, there is no effect on the engine running.

2. No signals are sent out until the Thermocouple voltage reaches the first warning temperature, same voltage as in the case of the 355, I guess (see the temperature graphs below; I do not know whether they are for the 348, 355 or both).

3. When the Thermocouple temp. reaches the warning point, the SDECU starts flashing the SDL.

4. When the Thermocouple temp. reaches the critical point, the SDECU switches the SDL permanently on and switches a permanent +12V to the relevant ECU pin which shuts the bank down.

Image Unavailable, Please Login

Image Unavailable, Please Login

Image Unavailable, Please Login

Image Unavailable, Please Login

 

One item to add the excellent write up from Miro is his info applies to the 2.7 ECU. The 2.5 ECU on the '89 cars functions somewhat different. The main thing is it will flash then go solid with the light but not shut the bank down.

 

The info I provided is for the SDECU with a 6-Pin (3+3 connector, I couldn't find it anywhere). You can use a beautiful AMP Superseal 6-Way connector:

Image Unavailable, Please Login

The "+12v from relay D" is just another ignition switched supply. For some reason, this SDECU has two ignition switched supplies and two grounds. I suspect the later 141037 is the same as 142581 except that the 6-pin connector has been rationalised to a 4-pin connector (+12V, Gnd, SDL and ECU) by merging the two +12V supplies and the two grounds inside the unit. So, instead of a wide 6-way AMP connector, a 4-Way can be used (it would require just connecting the two +12V and the two grounds to one pin at the loom side).

I have actually adapted two 355 SDECUs for use on my 348 by addition of a volt sensing circuit/relay and a flasher. The volt sensing relay is programmable so I have set it to 3.6 volts (instead of 3.7, a bit earlier warning) and it will only flash the SDL but not shut down the bank. I did not replicate the 1 sec. pulse on start-up since my engine is not affected by the fault code. I could have added a one time 1 sec. on-off circuit but I did not want to risk bank shut down if something goes wrong with this circuit and it goes on but not off. the good thing about my installation is that I can see the the voltage (= exhaust temperature) on the displays of the volt sensing relays. Some pics:

Image Unavailable, Please Login

Flasher (the black cube) placed next to the volt sensing circuitry:
Image Unavailable, Please Login

Thermocouple heated with a torch:
Image Unavailable, Please Login

Ignition "on":
Image Unavailable, Please Login

I can see the displays through the bonnet:
Image Unavailable, Please Login

 

NOT on MY car!

 

This is great news for the community. Thanks for your effort.

 

Received wbt prototype sdecu today!! It looks great and am installing it now and will report back

 

I have a SDL that I can't track down as of yet.

I have bench tested all of my cat ecu's thermocouples and they bench test fine.

I'd love to test these as well. Hell what's the costs to ship I'll pay just to test them after DR BoB.

 

In my opinion, the best way to determine what the issue is is to tap the outputs of the sdecu and drive around with a dvom and observe real time what is going on.

I’ve been involved with these things for awhile now and if you do the above you will have objective info instantly to act upon.

In my case it was a faulty TC (thermocouple) that bench tested fine but installed in the car would create voltage spikes triggering sdl. I have a video of the dvom showing the spikes I could post.

Ultimately you will need known working or new components to overcome the issue as there is no repair process per se unless of course there is a wiring problem. But with three locations and six components and watching a dash light and waiting for the start up to full heat cycle and in the BP valve location, the opening of the valve, it is just a crap shoot without doing as I said above.

 

Amazing. Any plans on making these?

 

I'd take one or a set once these are in production.

 

Bob,
Post your video. Sounds very interesting and educational.

 

Skipp,

You made me go hunt for it.

You can see the random spikes in the SDECU output at idle. This was the right cat location. OEM SDECU and TC.

Hopefully anyone with a sdecu issue they can’t solve will see this. Both the sdecu and the TC bench tested fine.

This is where a lot of unknown sdl come from. Bad TC in this case.

If you look at the graph wbt posted above you will see that on cold start the output is .5v and will usually max out around 2.8-3v under normal conditions. When outputs over 3.5 are seen by motronics it triggers sdl and some combination of higher of repeated triggers bank shutdown.

 

Not really. "wbt" will most probably make the 348 SDECUs (as one piece) soon as he has already made the 355 SDECUs to which he will just need to add a simple small circuit for different output signals (and different connectors) and it will be good for the 348.