Memory dump of ECU

[b3lha]

Originally Posted by svxistentialist

There are a few issues to sort with the car, some serious, most are trivial. Most serious for the moment would be the rear brakes. Rear left is dragging and gets heated in a short drive. The day I drove the car for registration the brakes "disappeared" through overheating, I had to pump the pedal to avoid the back of a bus! I'm going to replace the two rear disks, use Greenstuff pads and put on braided hoses all round. Naturally I will have to replace or repair the rear caliper; probably the piston seals are bad. As part of this renewal I will bleed down all the old fluid from the system and replace with Dot 4. I suspect what is currently in the system is there for years and has taken on water due to the deliquescent nature of brake fluid.

If you are going to the expense of new disks, pads and hoses, surely it makes sense to upgrade to DOT5.1 fluid rather than DOT4? I'm using it in the Claret car.

Originally Posted by svxistentialist

A labour of love

That's for sure.

While testing your TCU I found a clue to the electrical drain on my black car. I normally power the laptop from the cigarette lighter, but I discovered that on this car the cigarette lighter doesn't work. The cigarette lighter is on a 20A fuse, not the one I am having trouble with. Hmm. But then, stuffed under the centre console, I found an aftermarket cigarette lighter socket wired into the radio circuit That's the circuit that is intermittantly giving trouble. The radio circuit only has a 10A fuse and I found some spare 10A fuses in the glovebox. I think that one of the previous owners has been overloading the radio circuit and has probably melted something somewhere.

[b3lha]

Originally Posted by TomsSVX

if you guys need spare parts, I have oh, about a dozen TCU's stacked up ECU's are on the more rare side but I have one or two extras if ya need some toys to play with

Tom

Thanks for the offer Tom, I'll let you know. I was thinking that if the UK TCU mod that I did for Joe works, then maybe I can do it for other people on a core-exchange basis.

I think the TCU's are all the same hardware. So I should be able to reprogram any TCU with UK, JDM or USA software as necessary.

The USA ECU has an atmospheric pressure sensor, which the others don't. It sends the signal from that sensor to the TCU. So a USA TCU requires a USA ECU in order to work correctly. It won't work with any other kind.

But I think a JDM TCU should work with a USA ECU. If you guys could find a way to convert a US type transmission to a JDM type one then we could reprogram your TCU with the JDM software to control it.

[Nomake Wan]

Originally Posted by b3lha

Thanks for the offer Tom, I'll let you know. I was thinking that if the UK TCU mod that I did for Joe works, then maybe I can do it for other people on a core-exchange basis.

I think the TCU's are all the same hardware. So I should be able to reprogram any TCU with UK, JDM or USA software as necessary.

The USA ECU has an atmospheric pressure sensor, which the others don't. It sends the signal from that sensor to the TCU. So a USA TCU requires a USA ECU in order to work correctly. It won't work with any other kind.

But I think a JDM TCU should work with a USA ECU. If you guys could find a way to convert a US type transmission to a JDM type one then we could reprogram your TCU with the JDM software to control it.

Any chance you could find where in the code the TCU decides what RPMs to shift at in Power Mode? Maybe have a chip ready that changes it from 6500 to 7400?

[b3lha]

Originally Posted by Nomake Wan

Any chance you could find where in the code the TCU decides what RPMs to shift at in Power Mode? Maybe have a chip ready that changes it from 6500 to 7400?

That's one of the things I'm trying to figure out.

I was looking at it a while back. IIRC The TCU uses the "straight line equation" y=mx+c to depict lines on a graph. x and y are speed and rpm. The values of m and c are stored in lookup tables. The current lookup table is chosen based on shift mode(normal,power,economy,manual,cruise,overheat), stick position, and some other criteria. Then it decides what to do based on whether the point (x,y) is above or below the line. But there is some other stuff involved that I don't quite understand yet.

Hopefully I'll be able to get to the bottom of it in the next month or so. Depending on how much spare time I have to work on it. Now that I am able to change stuff I'm excited to start doing it.



I finished building the other two ECU memory boards last night. I'm really pleased with the way they turned out. I've never made anything like that before. It makes me want to learn more about electronics. I'll put them up for sale when I get back from Kenya. Probably around 50 GBP each.

[Nomake Wan]

Oh, that reminds me. Since the TCU can detect an overheat condition, does that mean there's an internal temperature sensor for the transmission? Or is it just a thermostat device with "normal/overheat" readings only?

[b3lha]

Originally Posted by Nomake Wan

Oh, that reminds me. Since the TCU can detect an overheat condition, does that mean there's an internal temperature sensor for the transmission? Or is it just a thermostat device with "normal/overheat" readings only?

Yeah there is a ATF temp sensor. You can read it with the tcuscan program. I have an aftermarket cooler between the rad and the condenser and I was quite surprised that a high speed drive, with lots of air flowing through the cooler, did not seem to lower the ATF temp I guess it's true what people say about it taking on heat from the rad and condenser.

[Nomake Wan]

Oh wow, silly me, that's what I get for not using the new program! Time to fix my adapter!!

EDIT: Oh, that reminds me. Could you do a run-down of your Solenoid C findings from your JDM car? Same as I did mine. Because a friend has been asking about the VTD, and while I know that the most it'll do is 50/50, I'm curious as to what happens on the other end. Or rather, how much rear bias will it do?

[oab_au]

Originally Posted by Nomake Wan

EDIT: Oh, that reminds me. Could you do a run-down of your Solenoid C findings from your JDM car? Same as I did mine. Because a friend has been asking about the VTD, and while I know that the most it'll do is 50/50, I'm curious as to what happens on the other end. Or rather, how much rear bias will it do?

The VTD is a lot different in the way the torque is delivered to the wheels. The Duty cycle signal is a secondary contributor to the system.
The actually gear ratios in the epicycle gear set in the center diff, set the torque ratio at 36% to the front, 64% to the rear. This is the base division, the Clutch that the VTD uses is mainly a Limited slip clutch, to prevent wheel spin.
The duty cycle signal is used to tighten up the ratio progressively from the 36% / 64% to the limit of 50% / 50%.
The ratio also varies as the car is cornering, as the different tracts of the front /rear wheels, changes the rotational speed of the two sets, to move the torque to 40%F 60%R. This is done without the TCU being involved.

This is how the WSM describes it.
http://www.subaru-svx.net/photos/files/oab_au/3071.gif

Harvey.

[svxistentialist]

Harvey, very enlightening, thank you.

Regards the LSD they describe; they say it is in the rear transfer case, and between the output from the front differential and the rear drive.

Does this imply that this particular LSD regulates the amount of torque distributed between the front and rear axles, or is it more complicated than this?

Joe

[oab_au]

Originally Posted by svxistentialist

Harvey, very enlightening, thank you.

Regards the LSD they describe; they say it is in the rear transfer case, and between the output from the front differential and the rear drive.

Does this imply that this particular LSD regulates the amount of torque distributed between the front and rear axles, or is it more complicated than this?

Joe

Hi Joe, This.is .a drawing that I did to show the way the VTD is set up.

Not the best, but you can see that the clutch housing, in yellow, is attached to the first transfer gear, through the pinion carrier. So it is attached to the front drive. The rear drive from the rear sun gear, in red, carries the clutch hub.
So the clutch spans the two drive outputs.
Increasing the pressure on the clutch, will reduce the torque difference between the two.

Harvey.

[b3lha]

Originally Posted by Nomake Wan

EDIT: Oh, that reminds me. Could you do a run-down of your Solenoid C findings from your JDM car? Same as I did mine. Because a friend has been asking about the VTD, and while I know that the most it'll do is 50/50, I'm curious as to what happens on the other end. Or rather, how much rear bias will it do?

I monitored it on my drive to work this morning. I had the stick in D and the power mode on the whole journey. When driving at low speed it seems to stay around 25%. When accelerating it rises to about 45% and at cruising speed it falls back to about 5%. It never went above 45%. I need to find somewhere with a gravelled car park where I can spin the rear wheels and watch what happens. I guess I should also try putting the diff lock fuse in.

I guess I should also have been looking at what gear it was in and the Sol B duty cycle, but it's hard to drive and study at the laptop at the same time. I really need to add a feature to the program to log the values to a CSV file for later analysis. That would be cool.

I think I can speed the program up a little by not scanning every item on every pass. The ATF temp and Atmospheric pressure for example don't need to be scanned every couple of seconds because they don't change that rapidly.

[Trevor]

Phil, you are certainly blazing a trail and Nomake is a good mate to have along riding shot gun.

Please, please set up a method of logging results. We do not want you to become a result.

As Ken would wish, take special care.

Sincerely, Trevor.

[Nomake Wan]

If the VTD transfer layout is as Harvey's diagram shows, then it seems to coincide with my understanding of the BNR32's ATTESA-ETS transfer mechanism. Save for the fact that the front transfer shaft is internal on VTD and external on ATTESA-ETS.

In which case, I would wonder why putting no pressure on the clutch pack doesn't result in a 0/100 split. In ATTESA-ETS, when there is no transfer clutch application, the transmission is fully rear-drive. All they have to do is pull the ATTESA fuse and they've got RWD.

On our cars, you put a fuse in and get full lockup of the transfer clutch, for 50/50. Hm...

I mean, I realize that Nissan and Subaru aren't the same, but they're quite similar, and were fairly close entities back in the early 90's. So, I wonder. Does this mean that no pressure to the transfer clutch still results in partial application? The clutch can never fully disengage?

Otherwise, I cannot figure out how torque is getting up front with the clutch disengaged.

[oab_au]

Originally Posted by Nomake Wan

If the VTD transfer layout is as Harvey's diagram shows, then it seems to coincide with my understanding of the BNR32's ATTESA-ETS transfer mechanism. Save for the fact that the front transfer shaft is internal on VTD and external on ATTESA-ETS.

In which case, I would wonder why putting no pressure on the clutch pack doesn't result in a 0/100 split. In ATTESA-ETS, when there is no transfer clutch application, the transmission is fully rear-drive. All they have to do is pull the ATTESA fuse and they've got RWD.

On our cars, you put a fuse in and get full lockup of the transfer clutch, for 50/50. Hm...

I mean, I realize that Nissan and Subaru aren't the same, but they're quite similar, and were fairly close entities back in the early 90's. So, I wonder. Does this mean that no pressure to the transfer clutch still results in partial application? The clutch can never fully disengage?

Otherwise, I cannot figure out how torque is getting up front with the clutch disengaged.

If the VTD clutch does not work, we just have a open center diff that will spin one set of wheels or the other.
The drive to both F/R wheels is mechanical.
The clutch plays no part in driving either end. It is only used to prevent wheel spin, and to alter the torque balance between the F/R.

Not the same as the Nissan.
Harvey.

[Nomake Wan]

That makes no sense. "Open center diff?" In all the diagrams I've seen, it's clearly labeled as a limited-slip differential.

I'm lost, here. The VTD clutch does what, exactly? If it is fully disengaged, then what? It makes absolutely no sense, from an engineering standpoint, to have a setup so that failure of the VTD clutch to engage results in front or rear wheels rotating arbitrarily. A design in which the failure of the clutch (or disengagement of it on purpose) results in one or the other rotating every time, on the other hand, makes infinitely more sense. Such as Nissan's ATTESA-ETS.