AA80E 8-speed

[Jim Blackwood]

Sounds good Marty. I've made contact with an engineer in the transmissions division of Aiasn and hope to speak to an applications engineer on Monday. Hopefully we can button down the last details then.

I'm also waiting on laptop upgrade parts so I can load TunerStudio and test the basic board. That should take about a week. To get to that point in the assembly you add the main ampseal connector and some jumper headers mainly. I have not installed the two recommended (mandatory) jumpers as instructed but will do so later, using the supplied jumper wires until then. The main jumper header is arranged so that simple 2 pin jumpers connect directly across the header. I am sure our current assignment scheme requires the pins to be connected differently, meaning a tangle of jumper wires as likely as not. So we will need to look at that to see if it can be simplified. Maybe I'll be able to do that while I'm waiting on the parts but I'm also still working on the injector installation so I can't say for sure.

Jim

[bondango]

Well i have the basic board now built Jim, so need to get a good 12v power supply (old ATI pc maybe) and try and get the firmware onto it, i may need to look at my serial assignments as Rx and Tx may need reversed.

http://www.2urgse-supra.herobo.com/wp-c ... 151711.jpg

like you i just used the jumpers supplied for the mandatory 2 needed.

http://www.2urgse-supra.herobo.com/wp-c ... 151758.jpg

Marty

[Jim Blackwood]

Good. I finally got TunerStudio loaded on the laptop. (I had been using MegaTune) so next is the serial connection. I also need to swap out the guts in my 12v power supply box but that should be simple enough. Probably get back on that in a day or so. Marty, if you get the chance, look at the GPIO jumper assignments. I think we will probably have to move a few things around.

Jim

[Jim Blackwood]

Found the cord for the LC-1 wideband so went to radiosnark, got a jack and mounted it on the end plate next to the ampseal connector, there is just enough room and also a handy ground pad right below it. Removed the Tx/Rx jumpers and ran new ones to the jack but haven't soldered the pads at the other end yet. Got a 50/50 chance of having them right so I'll check it first to see if it works and reverse 'em if it doesn't. Put a good spare 12v power supply in the bench box and got that set to go so now I just need to download the instruction pages for the next check and get that done and it's back to building.

I've never used the CAN bus yet so I'm sort of looking forward to linking this up to my MS-II and checking out what I can see on TunerStudio when I get to that point. If I can look at the MS-II from the jack on the MegaShift GPIO maybe I can just put a cap over the serial port on the MS-II and leave it that way.

I've drilled down to an engineer at the USA Tech Center of AisinWorld and am giving him a couple days to respond to email before calling. At least I'm in the right neighborhood now. Also have a few more weeks before hearing something on my connector sample parts, and need to order the other connector and a couple pins. Those shouldn't be much of a problem.

What IS a problem are the wires for the Ampseal connector. The ones in the kit came with pins crimped to 18" wires. If the wires came right through the transmission tunnel they would be too short and that would still be a problem, for several reasons. First, every splice is a potential failure point and I dislike failure rather intensely. Second, I do not have and cannot afford to buy enough different spools of wire to color code the wires, meaning there will be no way to tell which wire goes to which pin short of running continuity checks and that's no way to wire a car. Third, the last thing I want in my wiring harness is a big bundle of splices. Is there anyone who sells longer wires (about 6-8 ft) with crimped pins for the Ampseal connector? I really liked the harnesses that Glenn sold for the MS and am very disappointed that I've not found anyone picking up the slack. Is there anybody doing this? Or am I stuck with crimping my own? I'm pretty sure some of you guys know, and really, this is some critical information that I need to have so I really appreciate your help on this one.

Jim

[bondango]

Hi Jim,
I was able to upload firmware and communicate ok with the ECU. I am using rs232 connector on the amspeal.
As for Canbus comms, we will be unable to implement that on our setup during testing as the engine ecu is a long way of final development at the moment. I will have to use an input for engine load, and also RPM (shaft speed most probably). So thats 2 inputs i will definately have to tie up on my setup.
so i guess its time to work out which cicuits and assignments that will be needed.

jim, for development regrding speedometer output being enabled i would say no. as this would be a feature of the LCD display via serial comms as discussed earlier, or via tuner studio and laptop until an actual speedo unit is needed to be driven???

SR always on - switch via ignition or megshift power on??





Marty

[Jim Blackwood]

Marty, as I understand it, SR is a shift trigger. Although it is used in every gear, you can't just turn it on and leave it. The troubleshooting procedures show that if it either stays on or stays off the OEM controller goes into fault mode and disables about half the gear changes. So it has to cycle for each gear change. If you go back a page I gave as detailed a description of it's operation as I have currently deduced, "-SR can be thought of as a trigger that initiates gear changes. On every gear change it diverts flow from the clutch apply control valve to the sequence control valve, and then back to the clutch apply control valve. I am not sure if this includes reverse or not, but suspect that it does."

For the speedo output I agree, we need solenoid control worse than we need a speedo out and can get that info over the CAN bus I would think.

I will try to test tomorrow.

Jim

[Jim Blackwood]

WOO-HOO!
Got a box from Sumitomo today. One MAJOR hurdle behind me, still several lesser ones ahead.
In the box were the main 30 pin connector bodies (4) and pins for them along with the little silicon sleeves that seal the wires in the body. So the biggest unknown in this whole swap is now resolved. The selector connector body is available online for maybe $15 or so and that leaves the 2 large pins for the backup light circuit which are available as pigtails from Lexus if I can't find a better deal. Wonderful, Wonderful, Wonderful.

I had to send my GPIO board back to Al to have a look at it, couldn't establish comm but maybe he'll have better luck. That went out yesterday, and Matt gave me a price on longer wires for the ampseal connector and I intend to order those later today along with extra jumper wires (they come 10 to a package for seven bucks). May need those to make all the connections as several will be different from the default and the little monoblock jumpers won't work for those.

That leaves the solenoid currents but I still have a few things to try there.

I'll try to get a photo of the connectors later.

Jim

[Jim Blackwood]

The build is back on track. First, preliminary testing *seems* to indicate that the trouble I had with the board MAY have been caused by the 2.5mm phono jack used for the serial connection. If the follow up testing shows that this is indeed the case I will send Al a payment for the second board and then look at where I can use it in the control scheme. I have yet to get the first board to work correctly but I need to properly install a 5v regulator (which I cannibalized for the 2nd board) and then we'll see. We still want to do the AA80E controller as a single board stand alone if we could and definitely want to avoid a double GPIO solution, however it may come down to a choice between a double GPIO and a MS-II/GPIO and we will shoot for the latter choice here. I'm thinking that in my build I may be able to use the second board as part of my sequential/COP/EGT solution but will still need additional elements. It may prove useful, and because Al was gracious enough to send me a second board to help find the problem and both boards are now used the least I can do is try to find a purpose for it.

Now back to that audio jack. If you source your jack from Radio Shack (Will I never learn?) be aware that it is not made to handle a very thick mounting panel. In fact the end plate of the standard MS boxes are just about twice too thick. As a result the nut prevents the LC-1 plug from seating fully and shorts out one of the contacts to ground. It may be intermittent, and will show continuity. I solved the problem by counterboring the end plate to half thickness. At any rate this is something that should be checked with any phono jack.

I have loaded the 4.112 code, and installed the matching TunerStudio ini file. I am not sure what exactly to do with the 2.905 MS-II™/MicroSquirt®/Sequencer™ code file but have saved it and will need to sort that out eventually.

Marty and I have been in touch and have begun assembly. I have just reached the first decision point, the last step of section "D" regarding the speedometer output. Section "E" also requires decisions so now is the time to review the documentation and the spreadsheet in light of the actual physical board to determine the final circuit assignments. I will be out of the shop for about a week but should still be able to access the web and review documentation so that is on the agenda.

My inclination at this point is to configure the speedometer output as a solenoid driver. Marty's digital dash uses CAN signals to provide a speedo function, some electronic speedos may be able to accept the output shaft's sensor without signal conditioning, leaving just the scenario of a drive motor running a mechanical speedo unsupported and possibly some electronic speedos. In my case I currently have a mechanical speedo but a dashboard upgrade is in the future and I may have a second GPIO which can drive a motor, see above. TunerStudio also has a speedo display. I think we need solenoid drivers worse than we need a dedicated speedo output. If it doesn't move it doesn't matter what the speedo says. There also may be some way to get a speedo drive out of the MS-II. Anyone disagree?

GPI 1-5:
GPI-4 is the brake light input, that needs to stay as is.
GPI-3 is temperature. (How badly do we need this?)
GPI-5 is downshift. I think we need this for the paddle shifters.
GPI-1 is 2/4WD, I think maybe we can use this as a solenoid output. Anyone using this transmission for 4wd applications may need to come up with something different. (You really do not need an accurate speedo in low range, is that the function of this input?)
GPI-2 is swB(Output2) Solenoid output?

OK that's it from me at the moment. Lots of study ahead and I'll post my thoughts as I review the materials. Lance, Marty and anybody else, please jump right in with suggestions as I'm in pretty far over my head at this point and have no desire to pull the lone wolf routine here. I'm confident we can sort this out, but nobody said it'd be easy.

Jim

[Jim Blackwood]

OK, I've checked the spreadsheet posted a page or two back against Lance's draft wiring instructions, the jumper header and the build sheet and it looks like we're consistent in all areas except for a few details that we still have to define, mostly the ones we've been discussing. Here is what I noticed:

1) On the jumper header PEO and VR4 are not assigned. Maybe I can read back through these posts and find the reason. I'm guessing it has something to do with PEO only being usable as an input.

2) The input shaft speed sensor on VR3/PT5 is not an essential input I do not think. We can get engine speed over CAN. However, engine speed and input shaft speed are not the same thing. Is input shaft speed used anywhere?

I started back at page 1 and will re-read everything, on page 2 and already it is illuminating. May need another chart.

Jim

[Bernard Fife]

Is input shaft speed used anywhere?

Jim,

A number of newer trans use a sensor on the input shaft speed, it lets the controller keep track of converter slip as well as internal transmission slip (so the controller can reduce engine power and prolong the life of the trans if there is excessive slippage in the friction elements). MegaShift can do that too, to a limited extent, through spark and fuel adjustment over CAN to the engine controller.

Also, if a input shaft speed sensor is connected and enabled, the ISS is used to:
- report the input shaft speed,
- report the converter slip, and/or the internal slippage.

There is much more on this here: viewtopic.php?f=4&t=240&start=50#p3155

If none of that input speed information or slippage prevention ability is important to you, then you don't need to connect an ISS to MegaShift since it is not used in any transmission control functions.

Lance.