rusefi.com

This is a test mule / yard truck. This truck starts and runs good. It should function well as a test mule. At any point the OEM ECU can be re-connected and if required can be verified and measured. I'd kind of like to sniff the existing signals, then tune the ECU separately to get the same fuel pulses over various loads and such, but meh, that's probably not going to happen. This first post will be updated to function as a kind of build history for this engine.

It will need
-- frank 1B 12V via truck A2 key switched with 10A fuse.
-- frank 3B GND via truck A4
-- frank 3D GND via truck A31
-- frank 3C GND via truck A32
-- frank 2O ASD via truck C3 low side drive Haynes 12-37
-- frank 2P Fuel relay via truck C19 low side
-- frank 3K 5V via truck A17 Haynes 12-39, 41
-- frank 2D CAM via truck A18 = sig Haynes 12-39
-- frank 2B CRNK via truck A8 = sig Haynes 12-41
-- frank 3T Injector 1 via B4,
-- frank 3Y Injector 2 via B15,
-- frank 3U Injector 3 via B5,
-- frank 3Z Injector 4 via B16,
-- frank 3V Injector 5 via B6,
-- frank 3W Injector 6 via B12,
-- frank 3S Injector 7 via B2,
-- frank 3X Injector 8 via B13 Low side drive Haynes 12-43
-- frank 1F Ignition via A7 low side drive single signal, high voltage. Haynes 12-42

-- frank 3O MAP via A27 = sig Haynes 12-39
-- frank 3M TPS via A23 = sig Haynes 12-40
-- frank 3Q CLT via A16 = sense resistor + Haynes 12-41
-- frank 3P IAT Via A15 = sense resistor + Haynes 12-39
-- frank 2N Alternator field control, Via B10 Haynes 12-19

IAC reference Haynes 12-41
-- A4988 ENABLE --> STM32 PE12
-- A4988 SLEEP --> A4988 RESET (JUMPER TOGETHER)
-- A4988 STEP --> STM32 PE14
-- A4988 DIRECTION --> STM32 PE10
-- A4988 GND --> GND (BOTH GND'S SHOULD BE NEAR HARNESS CONNECTOR)
-- A4988 VDD --> 5V
-- A4988 1B --> harness side of W20 --> Frank pin 1T --> ECU harness A20 --> step 4
-- A4988 1A --> harness side of W19 --> Frank pin 1S --> ECU harness A19 --> step 1
-- A4988 2A --> harness side of W17 --> Frank pin 1Q--> ECU harness A10 --> step 2
-- A4988 2B --> harness side of W18 --> Frank pin 1R --> ECU harness A11 --> step 3
-- A4988 GND --> GND (BOTH GND'S SHOULD BE NEAR HARNESS CONNECTOR)
-- A4988 VMOT --> 12V BAT (ADDED 47uF TO 100uF 25V CAP)
The -- A4988 is this board or equivalent. https://www.pololu.com/product/1182


It would be nice if it had
-- Tach via ECU C31 high side drive Haynes 12-23
-- Check engine light via ECU C17 Haynes 12-25
-- Low fuel sense via analog C26 and A4 = GND Haynes 12-38
-- Speed control servo via C4 and C5 Haynes 12-40
-- Starter relay sense via A6 Haynes 12-41
-- Downstream NB O2 via A25 = sig, A4 = GND Haynes 12-42
-- Upstream NB O2 via A24 = sig, A4 = GND Haynes 12-42

Frankenso will be connected by getting a junk yard ECU, cutting it up to get the connector, then soldering a pigtail to a Frankenso connector. It appears the ECU controls the primary voltages for the coil and will need to sustain high voltages. Hmmmm, J701?? Perhaps get an external coil with integrated ignitor then simply switch the dizzy coil wire?? This part of the puzzle needs to be figured out. Coil only has 2 wires, so I'm reasonably certain the ignitor is in the ECU. Probably wired with J701 as noted here http://rusefi.com/wiki/index.php?title=Manual:Hardware#J701_NEON_engine_schematics Perhaps I can use this board http://rusefi.com/forum/viewtopic.php?f=4&t=286

Any how, it's time to wait for the mailman and that pigtail connector.

http://rusefi.com/wiki/index.php?title=Vehicle:Dodge_Ram_1500_1996 shows fully-independent sequential injection? We really need a couple of extra low side drivers on Frankenso...

Are you considering batching injectors in pairs or an external http://rusefi.com/forum/viewtopic.php?f=4&t=577 or do you think you would be fine?

It's native sequential, batched would require me to add injectors. This has CNG as an option, so who knows perhaps some day I'll do something crazy and add CNG as well as gas fuel. For now I'll focus on 8 injectors. I might need the aux board to get more controls, as this needs transmission controls and it has several dash lights that could be nice to have like MIL and low fuel. I'm running out of IO quickly, while the Frankenso board has several unused pins in the connector. I might have to make the first add-in board which would be placed above the pile of Wxx jumpers near the connector. Any how first things first, make it start.

I forgot about this guy for ignition http://rusefi.com/forum/viewtopic.php?f=4&t=286 I have the parts, so I soldered it up and will be using that for my ignitor. I'd like to find some kind of heatsink for it, then glob the entire thing in hot glue to make it a nice in-line option.

kb1gtt wrote:It's native sequential, batched would require me to add injectors.

I do not follow. I thought that all you need to get batched would be connecting pairs of injector wires together, so this case you are controlling 4 wires, with each wire controlling two injectors wires in parallel?

I think you are seeing the common 12V wire. This has 8 injector wires, one for each cyl. The common wire splits the left and right side, as well it's powered via that ASD relay and is kind of complicated about how it provides the 12V. However at the end of the day, it's just 12V supply, then each injector has one wire that runs to the ECU connector, for the typical low side drive.

Rather than batching to get extra IO, I'd be more likely to find a way to add extra IO.

It is indeed fully sequential injection with a distributor. There is an 8 tooth crank wheel located in the back of the engine, and a cam sensor that communicates the position of cylinder 1 in relation to those 8 pulses from the crank wheel.

The JTEC ECU in these uses a motorola MC68HC16Z3 chip as the main control with two simpler Motorola chips controlling ignition and injection. It's actually a pretty well designed system, it's just limited by software. There's some interesting info on it here if anyone's curious, although it's kinda irrelevant to our interests: http://thespeedfreaks.net/showthread.php?7503-Open-Source-JTEC-tunning

Also, Jared, I made a little video here for you:

Thanks for the vid. Yours is much cleaner than mine. Mine is rust colored.

It's a bit early, I'm focusing on running engine first. Do you know the typical frequencies of the output shaft sensor? I'm sure it goes to 0RPM, but how fast does it go? I'm also sure it's linear, so if we know any speed and it's RPM, then we'll know the scale. Also do you know what the min sensing is? For example, at some low RPM the sensor won't trigger an output.

Do you happen to know if the alternator is controlled similar to what we did here http://rusefi.com/forum/viewtopic.php?f=4&t=747 It seems this one has an isolate sensor. The one we did before had a common GND. Do you happen to know if there are special reasons for doing that? If not I'll just try connecting one side to 12V the regulating the side that connects to the ECU.

The OSS is a halls effect sensor that sits over a 23 toothed gear, which is also the gear that the parking rod jabs into to lock up the transmission. So, for every rotation of the output shaft or driveshaft, whatever you want to call it, there's 23 pulses. That'd get you something like 52k pulses per mile depending on your axle ratio and tire size. It's really simple, there's nothing else to it than that. I don't see why it'd have some sort of minimum speed. I have no clue about the alternator, but I'll try to research it and ask around.

I think this is VR not HALL. It has 2 wires correct? See this snippet Being VR is not a huge deal, as the frankenso board has VR inputs. The slight problem would be that the software doesn't exist to decode it. There would be a min RPM as VR sensors don't go down to 0RPM. They need a certain amount of change before there is enough energy for the sensor to detect the motion. For your CRANK shaft sensor with around 36 teeth, I generally expect the min RPM to be around 50RPM or below. With a 23 tooth wheel, I would expect a similar RPM. I wonder what typical speed 50 RPM of the output shaft correlates to. I have no idea what gear ratio the truck has, and it's on stock tires.

I expect the CAM and CRANK to be something similar to this

Then it might be VR, either way, that's the basic principle of its operation. I've taken it out of the transmission before, it seems to have a magnetic tip that protrudes. In my case, since my transmission failed, I had a lot of metal stuck to it. Lol. 50rpm is well under 1mph. 60mph is around 2300rpm or somewhere in that range, you get the idea. You probably have 3.55 gears, all you have to do is crawl under the truck and check the axle, there should be a sticker.

I should add that the governor pressure is not temperamental enough where you'd have to worry about speeds that low. Generally even at 1mph you'd be feeding it pretty close to a 100% duty cycle to achieve around 0psi govenor pressure. It's not picky, it tends to shift into 2nd gear around 10-15mph at light throttle on mine, however my transmission is modified. Was around that range before though. 50rpm is probably around 1mph now that I think about it. I'm just rambling at this point, don't listen to me haha

Analog channels on your board are soldered as described on at http://rusefi.com/wiki/index.php?title=Vehicle:Mazda_Miata_1994 - at some point I want this analog setup to be reflected in Frankenso schematics

Ann inputs and outputs could be re-mapped any way you want, but suggested default pinout is in http://rusefi.com/images/Frankenso/frankenso_default_pinout.png at http://rusefi.com/wiki/index.php?title=Manual:Hardware_Frankenso_board#Default_Pinout

[video][/video]

The basic time line behind that video was about like this.

Over the course of the week before the video, an ECU was obtained and the harness connector was removed, as well the schematic was obtained in a Haynes manual which was studied.
-- Sat morning, was when real works started, by first wiring 12V and GND, then verifying 5V and power works. Then the Hall sensors were wired, and a log was capture and sent to russian Total time was like 1 hour which was mostly spent soldering wires and finding missing jumpers.
-- Sunday morning, install new firmware with crank decoder and wire the injectors, ignition, fuel relay and ASD relay. Total time was again about an hour, we tested each injector and ignition, but as a system it was not functioning properly. Some diag tools were needed and I ran out of time for the morning. During the day russian developed some handy test tools in the dev console, then I got to get another hour or so on it in the evening. We found a missing jumper which was causing the fuel pump to not turn on. Once we got the fuel pump working, it almost started right out of the gate. After some tuning adjustments to the spark advance and fuel, it was running smoothly only 5 to 10 minutes after it fist coughed. As well it was a "cold" engine. The above video was taken and it's called a night.

-- TODO items include connecting the MAP, IAT, CLT, ect sensors, checking the timing with a timing gun, and continuing to do a bunch of diag now that I have a live test mule. I plan to pester russian about trans controls such that this can be put under some load. However getting it to rev and function a bit better is first priority. I also need to make a break out board for that connector. I'm running the risk of breaking a pin off of that connector which would suck.

So all in all it was a good day, we have 8 injectors operational.

My mouth is gaping open. Just read this thread. I had been going back here, refreshing it, hoping you were gonna reply, but I had no clue you were already well on your way to having it running! Honestly I'd like to thank you, this is a huge amount of encouragement for myself, and I'm extremely excited. Definitely can't wait to see what happens next.

kb1gtt wrote: -- Alternator field control, Via C12 and B10 in the ECU harness Haynes 12-19

Do you know anything about how this is controlled? I suspect that alternator is one of the potential noise sources so I wonder if we can expedite the alternator. I guess MAP and WBO would move us forward most.

Meanwhile, this is what I did Saturday. Best pass I could get out of mine.

1/4: 17.279s
1/8: 11.225s
Trap speed: 80.87mph
60': 2.691s

I'm sure with some better tuning it should see 16 second times, maybe.

I would avoid controlling the alternator with the ECU. The easiest, noise free way to do this is to use an old school external regulator. The volt regulator is kind of like a pulse, activating a magnetic field through grounding of the unit, keeps noise out of the ECU if you go this way and it lasts forever. Bulletproof, foolproof, I'd say ignore using rusEFI to do it.

3 wires - ground, power to batt and green wire goes to regulator



http://store.alternatorparts.com/partnoc8313.aspx This would work fine, a friend of mine is using it with his 5.9 magnum, which originally did alternator control via the JTEC.

Sam wrote:The easiest, noise free way to do this is to use an old school external regulator.

To some extent we need some noise in order to find the weak/weakest point on the board.

That's a good point, lol. Alternator is probably gonna be the noisiest thing though, I dunno.

double post

Hmmm, it even has the magic green wire. I think we can do the same setup as on the NEON. I'll have to look at it closer. I seem to recall there were 2 wires going to the alternator, so I might need to do something slightly differently.

It's great to see so much progress on this! Iv been wanting to convert my old 89 TBI 5.7l to rusefi full sequential injection for a while but it only looked the Frankenso board supported 4/6 cylinders.

I'm thinking i could add some bungs for the injectors and sensors to an old intake manifold, plop the old tbi throttle body on top as a backup and just try to get it working / running off some of the tbi sensors that are there, there's no place for a crank sensor though. After that id spend the money on a proper intake and injectors etc.

Any old tach signal would allow for fuel. If you have spark now, you should be able to get a signal for ignition.

Hey Gents

In response to the video posted on FB:

Download the FSM here >>> http://dodgeforum.com/forum/2nd-gen-ram-faqs-diys-and-how-to-s/301336-2nd-generation-dodge-ram-service-manuals.html

96-97 is what im going to reference. On page 21-299, or in the general area, will explain how the EPC functions, and its pretty simple compared to a GM or Ford trans. Note 95 model years used an external TCU, so if you can find one to reverse engineer that may help. I will dig into my ATSG for the pressure data to help concoct a curve, and according to the FSM and ATSG manuals, there are 4 programmed pressure curves. Basically, higher DC% will cause a higher shift point, DC% drops to force the fluid pressure to act on the corresponding shift valve, causing a shift, then the DC% rises again until the next shift point. The 4th and TCC is controlled by the integrated TCU, unplugging the electronics should cause the trans to lock into direct drive, or 3rd gear. Out of time for now, will be back later...

Great and thanks. I'll start thumbing through it to see what I can learn. For now I'm most interested in the OEM MAP signal. I'd like to learn is it linear, and what pressure correlates to what voltage.

the MAP is generally linear AFAIK, but the scaling is standard for every calibration I ever looked at.. The MAP on these trucks is a standard 0-5V 1 bar sensor, I think earlier OBD1 Chrysler engines used a standard GM MAP. Keep in mind the initial baro read and pseudo baro read to set the baro scale, this is for elevation correction. If someone wanted to bolt on a hair dryer, a 2 or 3 bar MAP would need to be used.

http://injector-rehab.com/shop/mapsensor.html

http://home.comcast.net/~fierocave/ecmqrg.htm

http://delphi.com/manufacturers/auto/sensors/engine-and-transmission/diesel-and-gasoline/map-mat

34blazer wrote:the MAP is generally linear AFAIK, but the scaling is standard for every calibration I ever looked at.. The MAP on these trucks is a standard 0-5V 1 bar sensor, I think earlier OBD1 Chrysler engines used a standard GM MAP. Keep in mind the initial baro read and pseudo baro read to set the baro scale, this is for elevation correction. If someone wanted to bolt on a hair dryer, a 2 or 3 bar MAP would need to be used.

http://injector-rehab.com/shop/mapsensor.html

http://home.comcast.net/~fierocave/ecmqrg.htm

http://delphi.com/manufacturers/auto/sensors/engine-and-transmission/diesel-and-gasoline/map-mat

I'll be running a GT45 setup with mine, either 76 or 80mm, shooting for around 700-800hp, but I'd be happy with even 500.
Interesting information about the transmission, by the way. Info on that stuff seems to be limited... My ATSG manual is unfortunately an older A500 series manual so I haven't been able to figure out a lot of the details of the REs. Sounds like you know more than me though.

If I could go ahead and get a Frankenso, I'd like to do exactly what Jared is doing and go ahead and start. Use the factory stuff for testing and prototyping, then once it's reliable and we have stuff figured out for the most part, I'd go ahead and hardwire everything the way I want it. If anybody has a spare Frankenso, name your price, I may be interested since I really would rather skip having to solder it... Lol.

Someone has been working on building a stand alone controller, but the thread is a dead-end, much like all the other stuff I have found in the past. The good part is, the MS code is open source AFAIK, and the parameters seem to offer a good base. My knowledge is severely limited for assembling boards and components, I need to buy an assembled piece. Same goes for the code. Now if everything was defined, I could work with that, and figure out most issues logically.

The 46/47RE transmissions are really simple, electronically, compared to most modern transmissions. When you look at the hydraulic flow of the valvebody, the only thing the EPC solenoid controls is the pressure acting on the shuttle valves, and all they do is re-route pressure to different circuits. If the shuttle valve has governor pressure acting on only one side, call it LEFT SIDE, then it will move opposite of the applied pressure, then as pressure increases on the opposite side, enough to overcome the pressure acting on the LEFT SIDE, then it will start to "shuttle" back, opening other circuits which enables other gears. RH models, and other models alike, use a mechanical governor, and the TV(throttle valve) controls the line pressure.

The main rule of thumb is 10psi governor pressure per 10mph, but that's a rough wag and should be used as a baseline.

The 4 pressure curves are:

Indicated temperature below -1*C

Indicated temperature above 10*C- Logically thinking this is the normal baseline operating range

WOT-

T-Case in Low Range

http://www.msgpio.com/forums/viewtopic.php?f=4&t=25


This thread is still somewhat active, but hasn't reached real world testing.

http://board.moparts.org/ubbthreads/ubbthreads.php/topics/961847/1.html