rusefi.com

theflyingdutchp wrote: ↑

Fri Oct 19, 2018 7:17 am

Hey thanks! That will probably be easier to hack in, err I mean, install. It's been a while since I've covered MOSFETs, so please correct me if I'm wrong, but I'd attach my STM output to the G lead and then my second TC4427 input to the D lead? Goes hi when G is lo, lo when G is hi because the MOSFET sinks the current to ground?

In other news, I bought a TPS from an automatic Toyota from the same era. Its a four wire with a proper varistor position sensing. Looks to be bolt on, but not plug compatible unfortunately. Nothing too complicated to fix. Unfortunately I wasn't able to find a simpler ISC/IAC valve.

Yep that's correct, don't forget the resistor from D to + 5/12V so the D(/output) actually goes high when you put the G low. If you get that in place, you should have an easy fix which I find elegant in simplicity

Quick update!
Got the firmware loaded onto the discovery board. Tuner studio/RusEFI console connect (I should mention, not attached to the Frankenso board at this point).
Got the majority of the rest of the board and the connector soldered. Was able to use some rework equipment to fix up some of the fine pitch stuff, place the cap that shifted, etc.
Board w/ connector fits into the stock ECU case.

New issue: Crk/Cam don't have jumper wires at the board/connector like the others. I completely overlooked this. I've got a plan but I need to confirm the wiring in my car matches the manual I have first.

theflyingdutchp wrote: ↑

Wed Dec 19, 2018 7:29 am

New issue: Crk/Cam don't have jumper wires at the board/connector like the others. I completely overlooked this. I've got a plan but I need to confirm the wiring in my car matches the manual I have first.

I usually cut the long metal terminals on the back of the connector housing and solder wires right to the connector.

Progress and new issues: I've got 12V switched power going to W2 and ground at W41. LED D1005 lights up and I've got 5V across C1002. 12V from P902 to ground.

But at the Discovery headers I've only got ~0.7V at the 5V pins and 0V at the 3.3V pins. With or without the Discovery installed. I've got no LEDs on the Discovery when hooked up to the car (lights up when connected to the PC though).

Does anyone have suggestions for tests? I'm happy to take measurements, send pictures of details, etc etc.

Do I need to populate anything on page 10 of the schematic?
P604-P607 have no jumpers atm.

Thanks!

If you have 5V, did you install W23? This jumper is used because if the 5V is not properly soldered it can put 12V on your 5V, which causes bad things. So we verify the 5V is working before we install W23.

Nope, that's probably it.
Its the one in the top left of this picture?


May I suggest we update the wiki to reflect this need?
Do I need to populate P1001?

Yes, W23 is top-left of this picture.
P1001 is probably just holes if yo need to hook extra wires, not a real component?

Can you please make the changes and I will replace the picture?

Correct top left of that picture.

No need for P1001, that's just a test point. It will be 5V when you install W23.

I agree the wiki could use improvements. Do you have write ability? If so could you make changes?

I see @ made almost the same comments in at almost the same time as I did. At least we are consistent

Thanks guys, installed the jumper and its working now - board powers up and I can connect via tuner studio.. Dealing with all the jumpers now (slowly). I think I'll just repin the connector at the ECU plug for the VR issue. Gotta run to the yard soon and see if an automatic throttle body will swap on - the auto TPS won't bolt up to my manual throttle body.

I can take a new picture - although my solder joints aren't the prettiest, and write up a short procedure (IE: build the power supply, check the voltage at X and X. If good install W23).

Quick question - do I have to jumper 5V over to the connector for the MAF? The schematic doesn't show which pins are 5V (for the TPS and MAF).

https://rusefi.com/wiki/index.php?title=Manual:Hardware_Frankenso_board#Default_Pinout should be showing +5v pin 3K with W49

I'm blind... thanks Russian.

EDIT: I've got a 4-24 wheel, which one is considered crank and which is considered cam?

theflyingdutchp wrote: ↑

Sun Jan 27, 2019 6:36 am

I'm blind... thanks Russian.

EDIT: I've got a 4-24 wheel, which one is considered crank and which is considered cam?

what is 4-24 wheel? is that one single wheel? is it 24-4? is it something from https://rusefi.com/wiki/index.php?title=Manual:Software:Trigger ?

the one on crank is crank and the one of cam is cam, sorry.

russian wrote: ↑

Sun Jan 27, 2019 12:48 pm

what is 4-24 wheel? is that one single wheel? is it 24-4? is it something from https://rusefi.com/wiki/index.php?title=Manual:Software:Trigger ?

the one on crank is crank and the one of cam is cam, sorry.

https://rusefi.com/wiki/index.php?title=Manual:Software:Trigger#Honda_4.2B24

Sorry I should be more specific.
Two wheels, both inside the distributor, both driven off the camshaft. One has 4 teeth equally spaced and the other has 24 teeth equally spaced. Two separate VR sensors which share a ground. I figured that is the same as the Honda 4+24 Maybe not?

For Honda 4+24, "4" is "primary" and "24" is secondary

you need to make sure that your trigger shape matches perfectly or you would need your custom shape

Honda "4+24" was a side effect of trying different combinations of the three Honda signals. I am not sure if that trigger is enough to synchronize and identify your exact shaft position, I wonder if maybe yours on Toyota is a bit different? Does your toyota have mechanical ignition distribution? in that case you are OKish to start without exact synchronization but your fuel phase would be a bit random. So maybe later you would need to modify things if you want sequential injection and ignition.

Ok, I'll take a closer look at the schematics I have.

If they don't line up is there a way to log the pattern with Frankenso? Otherwise I can probably use a scope at school but that will take some time to get access too. EDIT: nevermind https://rusefi.com/wiki/index.php?title=Manual:Software:UnknownTrigger

The distributor is a single coil mechanical setup, and it's batch injection, so really I think it only needs to know which pair of cylinders is at TDC. But, in the future I think I'd like to move to COP ignition. In that case I figured I would just remove a tooth from the 24 tooth wheel. What do you think?

EDIT2:
I think its different enough that it will require a custom pattern. I'll hookup the 4 tooth signal to "Cam" and 24 tooth to "Crank" and produce a log + schematic from the manual.

russian wrote: ↑

Sat Jan 26, 2019 10:00 pm

Can you please make the changes and I will replace the picture?

kb1gtt wrote: ↑

Sat Jan 26, 2019 10:01 pm

I agree the wiki could use improvements. Do you have write ability? If so could you make changes?

I created an account on the wiki, but I do not have edit privileges. If y'all want me to edit directly its the same username. Otherwise I'll make a quick write up in my spare time and post it here.

First of all you should be able to start your car with just the "4" without "24"

Once we have it running we can implement your custom Toyota 4+24 or maybe we will go the route of modifying the trigger.

russian wrote: ↑

Sun Jan 27, 2019 8:11 pm

First of all you should be able to start your car with just the "4" without "24"

Would I select one of the Honda trigger shapes with a 4 tooth wheel and only hook that wheel up, or do you mean create a new pattern?

It'll probably be another week or so before I can make any more progress. I blew a few input traces this weekend in my rush to get everything hooked up so I want to slow down, plus life is getting a bit busy.

russian wrote: ↑

Sun Jan 27, 2019 5:47 pm

Honda "4+24" was a side effect of trying different combinations of the three Honda signals. I am not sure if that trigger is enough to synchronize and identify your exact shaft position

Honda wheels are evenly spaced. 4+24 can get you running batch fuel, but no sequential. No idea if the Toyota wheels are the same, but if the car ran sequential fuel from the factory I don't think they can be without another sensor.

theflyingdutchp wrote: ↑

Mon Jan 28, 2019 3:49 pm

Would I select one of the Honda trigger shapes with a 4 tooth wheel and only hook that wheel up, or do you mean create a new pattern?

You would select existing universal skipped wheel configuration, 4-0.

ZHoob2004 wrote: ↑

Mon Jan 28, 2019 4:52 pm

Honda wheels are evenly spaced. 4+24 can get you running batch fuel, but no sequential. No idea if the Toyota wheels are the same, but if the car ran sequential fuel from the factory I don't think they can be without another sensor.

The car is batch fuel from the factory. Can you tell me more about the Honda wheels/point me towards some reading?

russian wrote: ↑

Mon Jan 28, 2019 6:17 pm

theflyingdutchp wrote: ↑

Mon Jan 28, 2019 3:49 pm

Would I select one of the Honda trigger shapes with a 4 tooth wheel and only hook that wheel up, or do you mean create a new pattern?

You would select existing universal skipped wheel configuration, 4-0.

Ok gotcha, sorry to ask so many simple questions Russian. If I could suggest another wiki edit, perhaps add that trigger pattern to this page? https://rusefi.com/wiki/index.php?title=Manual:Software:Trigger

I've done a lot of research on this over the past few years, so get ready.

Honda has 3 wheels in the distributor (all cam speed). 24 tooth for primary trigger, 1 tooth for cam position, and a 4 tooth for quick startup (about 10 degrees BTDC for cranking)

With the Honda wheels, one can do batch fire by treating the 4 tooth wheel as a 2 tooth crank wheel and then do batch fuel. The 24 tooth can be added in for better resolution using the 4 tooth to identify tdc on each piston.

Here's a picture I found of a 4a-ge distributor being modified for megasquirt. In this case, they have removed a few teeth to emulate a 12-1 crank wheel (24 tooth spinning at half speed with a tooth removed from 2 opposite sides).

In our case, this is unnecessary as you can simply use the 4 tooth as your reference position.
http://www.aeu86.org/forum/Thread-Megasquirt-I-v3-0-and-toyota-distributor-igniter

In the case of the Honda trigger (if I recall correctly), the 4 tooth does not line up exactly with the 24, and I think it's pretty unlikely that the toyota wheel is the exact same, but it's worth a try. Your best bet is to hook everything up, pull the spark plugs and crank with the sniffer open to capture the cranking pattern and get a custom trigger defined.

PS: Honda engines of that vintage spin counterclockwise, and so does the distributor. Distributor is on tail side of motor, ans spins clockwise when viewed from the top.

ZHoob2004 wrote: ↑

Tue Jan 29, 2019 7:39 pm

Honda has 3 wheels in the distributor (all cam speed). 24 tooth for primary trigger, 1 tooth for cam position, and a 4 tooth for quick startup (about 10 degrees BTDC for cranking)


In the case of the Honda trigger (if I recall correctly), the 4 tooth does not line up exactly with the 24, and I think it's pretty unlikely that the toyota wheel is the exact same, but it's worth a try. Your best bet is to hook everything up, pull the spark plugs and crank with the sniffer open to capture the cranking pattern and get a custom trigger defined.

PS: Honda engines of that vintage spin counterclockwise, and so does the distributor.

That is exactly the information that I was looking for, thank you!

I'll post some FSM snippets later but that sounds pretty much identical to the Toyota setup. I've suspected for a while now that this system was engineered by a separate company (Denso?) and sold to all the Japanese manufactures. Maybe other parts of the whole EFI setup too (TPS, AFM, etc).

Anyway, I'll post those pictures later - if you don't mind, could you take a look and tell me whether it looks the same?

Please register on wiki and tell me your wiki username, i will grant you write aceess to improve the trigger page. Need your help here!

Happy to help! It's Theflyingdutchp.

Here is the snapshot from the manual. I'd suspect that if the Honda motor spins the other way that the 24 tooth wheel will be offset in the other direction from the 4 tooth wheel. Although I suppose it depends on what side of the head you put the dizzy. Anyway, I'm curious what you think ZHoob2004.

theflyingdutchp wrote: ↑

Wed Jan 30, 2019 6:00 am

Although I suppose it depends on what side of the head you put the dizzy.

Now that you mention this, I thought it through and realized the Honda distributor is on the tail side of the motor, and thus spins clockwise when viewed from the top. I don't have a good source for the angles, but luckily we already have a trigger pattern for the Honda wheels that we can use to derive the angle.



Looking at this next to your drawing, it appears that both the Toyota and Honda patterns have the 24 tooth lagging slightly behind the 4 tooth wheel, so they may very well be the same angle, though I'm not sure what is meant by the "5, 7, or 10 BTDC" note.

I'm not sure of the OEM for the Honda distributor. It wouldn't surprise me if it was Denso, I think the MAP sensor is...

Sure would be convenient if the pattern was the same.

theflyingdutchp wrote: ↑

Wed Jan 30, 2019 6:00 am

Happy to help! It's Theflyingdutchp.

you now have wiki write access

ZHoob2004 wrote: ↑

Wed Jan 30, 2019 6:53 pm

Now that you mention this, I thought it through and realized the Honda distributor is on the tail side of the motor, and thus spins clockwise when viewed from the top.... <snip>
....I'm not sure what is meant by the "5, 7, or 10 BTDC" note....
...Sure would be convenient if the pattern was the same.

Its looking good - once I get everything plumbed from the connector into the ECU I'll try it out. I think the 5,7,10 degree BTDC refers to the base timing which can be set by rotating the distributor.

russian wrote: ↑

Wed Jan 30, 2019 7:34 pm

you now have wiki write access

Made some edits - hopefully they're suitable. Do you know what the functional limits of the universal wheel are? TunerStudio allows up to 500 teeth with 500 missing.

It was a busy two weeks + weekend. I'm hoping to have a little time this weekend to wire/solder. We'll see.

Great question about limitations! it's probably around a hundred or two. Would not have time to confirm now sorry

russian wrote: ↑

Sat Feb 09, 2019 4:33 pm

Great question about limitations! it's probably around a hundred or two. Would not have time to confirm now sorry

No problem! I imagine that would be plenty for any standard trigger wheels. I'm just curious, is that a processing speed limitation? Is rusEFI structured as a finite state machine?


In Camry news, I've got a MAF reading now, it picks up IAT and CLT too although the curves are WAY wrong . I still haven't gotten an automatic throttle body for the TPS. I also drilled a hole in the case so I can connect to it with the ECU in there.

A question for y'all, maybe @ specifically, my two VR sensors share a ground wire. I'm considering connecting the left side (connector side) of W1002 and W1003 together with a short wire so they can share the ground. Does this sound reasonable? The pins are next to each other so I don't think it'll be too big of an antenna, but my understanding of AC is pretty weak.

EDIT: Drew a picture to (hopefully) explain better. Jumper is purple. On the PCB they're next to each other, about a close as they can be.