rusefi.com

Hello guys,

My name is Sam and I drive a 2001 Dodge Ram with a 5.9L Magnum V8. The stock ECU is known for being a piece of junk, and I'm busy looking at options for a new ECU (going to be making 400+HP after a cam, and probably somewhere around 600hp after a turbo setup). I've already ruled out MegaSquirt, and I'm still considering FreeEMS. I figure I may as well test the waters over here as well. How can rusEFI help me, and how can I help rusEFI? I would love to contribute to this project, I'm pretty new to this sort of stuff though so I'm sure I'll have a lot of noobish questions, but I plan on beginning the computer swap around May.

-Sam

Good morning

v8 should not be a problem but noone have tried that yet, so far 6 cylinders is the most we've controlled in one engine. Noone have ever tried turbo charged anything, rusEfi is still work in progress - rusEfi is defenitely less mature then FreeEMS as of today. I hope that it's easier to dive into rusEfi because easier hardware situation and compatibility with existing tuning software but I could be wrong.

Hello Sam and welcome to the forum.

Feel free to ask questions. Have you seen the wiki? http://rusefi.com/wiki/index.php?title=Main_Page

What's your background? Mechanic, formal training, school of hard knocks, ect. I'm just trying to get a feel for what you may either want to do, or what you might be able to do. After I know you a bit better I could make suggestions on how you might be able to help the effort. Generic suggestions might include wiki help, or development help with specific features. Do you have software skills? Do you have soldering skills? Electrical skills?

It's been a while sense I was banned / left FredEMS. I'm not sure which project is more mature at this point. I think rusEFI has better validation of changes and new firmware builds, but FredEMS has more features that have been field tested. There is a build server that tests each newly compiled firmware against a live board to validate firmware functionality. I have found this effort to be friendly and pro-active which I'm a big fan of. From a hardware stand point, rusEFI has some reasonably good and available hardware. I understand the maturity issues to be a lack of software development, which will come with time. So I tend to think that rusEFI is a good platform which will continue to mature and continue to gain features.

Do you have a general layout of the engine, or some kind of electrical schematic? Is it VR, hall, other? I don't know of any real issues from the hardware stand point for a V8 turbo application. Are there any particular issues with MS that you didn't like? I can probably tell you if this has those same issues or not.

Any how, welcome along.

Issues with megasquirt? Closed source. I'm trying to do this all for under $200. That means computer, DIY harness (yes, I'd make a new harness), any necessary sensor changes, etc. Cost, and being closed source, is why Megasquirt turns me off. I was speaking with Andy about what a Jaguar board would cost me over at FreeEMS and the build cost if I were to do it myself is around $125-150, which honestly isn't bad. I'm young, I'll have time to mess around this summer (this truck is a daily driver and currently bone stock motor wise, but I obviously have big plans). As quoted from a megasquirt thread: "The 5.9 Magnum has a dual wheel setup - an eight tooth crank wheel and a one tooth cam wheel. Both sensors are Hall effect." It's a single cam 16V V8, fuel injected, OHV, using a MAP sensor... Etc. Not terribly complex. Right now it's equipped with a distributor but I want to run coilpacks when I convert the computer over. I'm sort of familiar with programming in Java, and I've messed around with an arduino a bit (I'm actually working on a digital dash using 7 seg blue LED displays).. I'm 17 and plan on majoring in EE. I can solder pretty well, my equipment is kind of crappy though(: I'd say I have a decent knowledge of electrical stuff but I still have a LOT (and I mean a LOT) to learn. Mechanic wise, well, I'm rebuilding my automatic transmission myself over the next two weeks (I also have plans to control it with an arduino). Putting good stuff in it. Basically, if it's a good platform as you say, I'm interested without a doubt. Is fully sequential ignition and injection possible with rusEFI at the moment? I've been told over at FreeEMS that sequential doesn't even matter and that running batch and wasted spark is just as good.

-Sam

We have found you can get an enclosure and bit of harness from a Miata style ECU on e-bay or salvage yard for like $20 to $40. Which is low cost for sure.

Last I recall, batched injection is partly developed but not fully developed and not tested yet. It's physically possible, but that's where the software is a bit young.

Physically the specs for Frankenso can be found here http://rusefi.com/wiki/index.php?title=Manual:Hardware_Frankenso_board which includes 12 low side high impedance injector drivers. Of course 12 drives for 8 cyl's doesn't allow batched injection. It also includes 6 hi/low drivers which are typically used for ignition and wouldn't allow for individual ignition control. However the STM discovery board can physically do it, it just wasn't in the native plan for Frankenso. Frankenso was intended to allow many applications for low cost, but then add options with an add-on board installed above those Wxx jumpers near the connector. AKA, many of us are looking for low cost have noticed the cost drops when you use 4 or 6 cyl instead of 8. Less cy'l's is less cost. Also when you can crank 400+ HP from a 4 cyl, why do you need 8 cyl? Any how, we know the V8 is wildly popular here in the states. So we left options for V8, but expected it would also increase costs to get those options. We focused on 4 and 6 cyl applications to help keep the costs down.

Almost all IO pins go through a 0 ohm jumper, which has thru holes and allows remapping IO and allows for an optional add-on board. See wire mess in picture posted here http://rusefi.com/wiki/index.php?title=Main_Page We aimed to allow some common Miata's to simply work with 0R SMT jumpers installed. However the first engine russian had wasn't that particular motor, so he had to re-map the IO. Which is why you see that wire mess on the board in the picture. The long term plan is to make that wire mess into an add-on board which can be installed separately and would look much nicer. This add-on board can also optionally include other items. For example we know of a Honda ECU that uses the same harness connector but has a 3rd VR signal. The add-on board would allow the addition of that 3rd VR as well as it would re-map the slightly different pin out.

The STM32 discovery board has 9 unused pins, perhaps an add-on board can be made to get more cyl's added. I don't see any problems with adding them to PD0 - PD6. Also an option worth mentioning is that we have started development of the 176 pin replacement option for the discovery board found here http://rusefi.com/forum/viewtopic.php?f=4&t=749 So in the future we expect to have far more IO pins.

Have you compared http://libreems.org/ in your options? AKA Fred from FredEMS has caused multiple good developers to leave FredEMS, some decided to split off from Fred and created libreems. It's basically the same core software and core chip selection that Fred originally choose, but they are much friendlier and less leech-ish than Fred. Also has the brain board costs come down in FredEMS land? Last I recall the TA card price came down to something like $125. If the brain board costs are still that high, it will really nibble into your goal of under $200. Has anyone taken what Marcos and I did with Puma and made a lower cost brain board option yet? http://puma.freeems.org/ I seem to recall that just the S12X chip cost nearly as much as the entire F4 discovery board.

STMdiscovery is about $14 from multiple sources found here https://octopart.com/stm32f4discovery-stmicroelectronics-20382043, ECU enclosure and harness is around $40, which leaves you with around $146 for the rest of it. I think you are much more likely to come close to your goal with rusEFI than any other option.

What I would encourage, is that you start with wasted spark, stock injectors and simply get it running. I believe you can get rusEFI for your goal of under $200 with minimal soldering. Most of the hard soldering is done on the purchased discovery board. That setup alone should make a significant improvement and will get you into knowing how all the pieces connect and work. Once you have that, you can look for additional features, like batched injectors and additional ignitions for individual cyl control, as well as turbo control and other controls. This architecture has the abilities to do what you want, but some features haven't been developed yet.

I would also encourage you to draw a system schematic, one that includes the things like fan control, alternator control, ect. You may find you need more IO.

kb1gtt wrote:Last I recall, batched injection is partly developed but not fully developed

http://rusefi.com/forum/viewtopic.php?f=3&t=822 runs on batched injector. What is partly developed and not fully tested in driving by MAF with air integration

Great to hear. I didn't notice that was batched. I also seem to recall rusEFI is now sequential so batched and sequential if the hardware supports it. So I think the issue with Frankenso is that is only has 12 fuel and 6 igntion, which would be added via add-on module.

I'm so confused right now... You were saying sequential is possible but batched isn't? Obviously Russian just refuted that but I was so confused reading that. Lol. I prefer sequential anyway. In my case with FreeEMS, there are already boards available where I simply get the 112 pin S12X and solder it into place. $15 for a PCB from coolEFI. I can get sample chips from manufacturers. I have estimated total cost with FreeEMS to be under $200. If you want to talk politics, PM me. I'd like to hear the stories about that stuff. LibreEMS seems terribly inactive, so I'd rather not jump into that boat. When I made an account there, Fred noticed. I don't really understand all the controversy, you can clarify in PM if you'd like. I ask a lot of questions and I'm bad at navigating forums, if there's any way to get in touch with you guys on Facebook or whatever, that'd really help me. Right now my truck is on jack stands in my driveway and I have the driveshaft halfway unbolted, all the transmission linkages and electronics disconnected, O2 sensors unplugged, and exhaust unbolted. About to pull the transmission. I'm in it deep. Lol. Here's a clip of what it sounds like open headers:



Also a couple images (hopefully these show up, I'm on Tapatalk.)

About drama from FredEMS, sure we can talk. I can send my skype or phone number and a time to call. I'm less than interested in the drama stuff, I'm more interested in getting things done.

Cool to see coolEFI, Looks like DeuceEFI has made a low cost BSDM tool. You'll either need someone to program your chip the first time, or you'll need a BSDM programming tool. That programming tools comes on the $14 discovery board, so not needed with rusEFI. I see a couple boards noted on that page for $34 to $65. I'm not sure what you are getting for $15. Is that a board with drivers, or is that just a brain board? Does he share the layout files? From a very basic review of the posted PCB, it looks like he's not using an automotive friendly 5V regulator. The rest looks reasonably good in terms of layout, analog and digital are separated in a semi star pattern, ect. However I'm concerned that the power supply isn't friendly to automotive applications with load dumps and various voltage spikes. I'd like to measure track widths on power devices, and I'd like to look at the GND plane floods. Those are some common issues with people when they layout a PCB.

Hmmm, as luck has it, I have a 96 Dodge RAM. It's probably the same motor. Can you post a picture of the OEM ECU? I think we have the same physical connectors.

All 2nd generation Dodge pickups share the same PCM, although there may be a different setup with the diesels, but the 3.9 V6, 5.2 V8, and 5.9 V8 all share the same JTEC PCM. You can read about that here:
http://dodgeram.org/tech/repair/Pcm_diags/Page1.htm

The 3.9 V6 is like a 5.2 with two cylinders chopped off. The 5.2 and the 5.9 share mostly everything, the differences being internal, like with the bore and stroke, camshaft, top end, etc. Same family of blocks though.
I would prefer NOT using the stock harness and enclosure. What I'd really like to do is get like, a 61 pin Deutsch connector, like the military spec ones, and have it right behind the engine in the firewall. Then I could have the ECU in the cab, so my engine bay looks gorgeous. Lol.

I seem to recall CPC style connectors were on the order of $.25 per pin. That's some pennies for male connector, pennies for female connector, then some for the male shell, then some for the female shell. It seems that many connectors are cheap for one side. I really like CPC's, but I'm not crazy about the costs.

Yup, same general PCM. Your truck is in much better shape than mine.

120k miles, thanks! I built the headlights with some cheapo $85 aftermarket housings, retrofitted some $40 Mini H1 clone projectors... They work insanely well. I want the engine bay to be incredibly neat and tidy looking.

Well, in great shape, except that the transmission is messed up. Lol. I'm building it up to handle at least 600hp 24/7 though.

What I found in the pan today... I have everything pretty much ready to pull it tomorrow. I suspect the rear servo may have broken, because that looks like rear band material to me... Assuming it is, the cost of this rebuild just went up ~$140. Fun fun!

Hmmm, your magnet looks like it didn't catch very much. I've seen them with far more build up. Those metal chips are a bit alarming. Good luck.

Keep in mind that I cleaned off that magnet and the whole pan just a month ago so all of the material you see in the pan is entirely just from one month of driving.

Eh, that's not good.

Got her out today. Used a piece of Powermac G5 to hold the torque converter in place.

Hi Sam,

Glad to see someone else from AR on here let me know how this project goes for you sounds like an awesome build.

Just an update for you guys:

As of now, I have my transmission completely disassembled, with the pump, front clutch, and rear clutch (input shaft stuff, basically) totally rebuilt and ready to go. What happened was that the rear servo cracked, probably from high line pressure and the fact that they're known to break, and the rear band chewed into the rear drum and annihilated all the friction material. The rear drum was ruined as well. Rear clutch steels were warped from excess heat, pretty much all the clutches (front, rear, overdrive direct, overdrive brake) looked worn but okay. The accumulator piston was broken into two pieces... Explains the erraticness in shifts, presumably from hydraulic pressure to the front and rear servos being super erratic. Long story short, stock crap sucks. These aren't poorly designed transmissions at all, Chrysler has been using this design since the 50s. It's just poor implementation of a few things in production; cost cutting measures. But, I'm replacing all that crap. So it'll be good.

Are you modifying for harder shifts and elevated RPM shift points? Does this have electronic control, or just electronic feedback? I know older tranny's you needed to drill out some holes, change some springs, ect to get a better preforming transmission. However I hear you saying things like solenoid is broken, so I wonder if this has electronic controls. Mine is also an automagic, and if it has electronics controls, I might find some time to figure out how to control it, electronically. However as it stands now, I haven't looked into it, but if you happen to encourage me by dropping some data like schematics or something like that in my lap, I might develop a board to help wake that tranny up.

Developing a controller for these transmissions would be HUGE, and change the Mopar hotrodding world quite significantly. The RE series automatics are essentially electronically governed 727s with overdrives (same as an A518 but with a governor). The RE series transmissions (except for those used with diesel engines) all bolt up to pretty much ANY small block Mopar engine, as well as the newer 5.7/6.2 hemis. Why don't people use them? There's no controller on the market. Not at all. I've been thinking about this for quite some time. If you have a 2nd gen with a magnum series engine, you have a transmission either no different from mine or differing only in how beefy it was built.

Yes, I installed a shift kit over a month ago, and yes, it involved tearing into the valve body, swapping springs, drilling holes, swapping valves, adjusting screws, etc. You do that to increase hydraulic line pressure for harder, faster shifts.

There are a few electronic things:

Governor pressure solenoid
Governor pressure sensor
Overdrive clutch solenoid
TCC solenoid
Fluid temp sensor
Output shaft speed sensor

That's it. The three solenoids share a 12V common source, and are ground controlled. The governor pressure solenoid has a resistance rating that would mean it'd draw 4amps if directly fed power, but I'm assuming the computer maintains a limited current ground. The computer controls a 512Hz PWM frequency to control governor pressure. It uses the governor pressure sensor as feedback to actually know what the hell it's doing. @1amp, or 100% duty cycle, governor pressure is 0psi, which would be like when the vehicle is stopped. At 0amps or 0% duty cycle, the governor pressure is at maximum psi. Governor pressure with a stock transmission usually corresponds to a psi for every mph. So @60mph, governor pressure would be 60psi. You get the idea. The output shaft speed sensor tells speed. The other two solenoids are simple on/off controls. It's not particularly complex. A controller wouldn't necessarily "wake it up", but it would be a major breakthrough in Moparland, and suddenly unlock all kinds of combinations never achievable before. Believe me, I've been thinking about it for a long time now.

That sounds reasonably easy and straight forward. Were are the electronics located? I've seen some tranny's with the electronics in the ATF, others with them remotely mounted. I don't follow what features it would offer over the stock electronic controls. Perhaps datalogging the tranny RPM and pressures would be handy, but it sounds like it wouldn't change the shift points, or shift speed. So I wonder what features would it potentially add. Do you know the internals schematic? AKA a solenoid can be controlled with a fly-back diode (free wheel diode, snubbing diode, what ever you want to call it), or it can be controlled with a voltage limiting device like the MOSFET we are using for the fuel injectors. The different topology's change the voltage peaks that are seen by the solenoid and how much heat the solenoid has to dissipate. So I wonder about the internals. Also what type of sensors are we dealing with, is the RPM via Hall or VR? Is the pressure a variable resistor, MEM's or other?

How do I verify what kind of tranny I have? The snow is almost low enough that I can take a look under the truck.

In all honesty you should be able to find what tranny you have by looking up your VIN online. Otherwise, there should be some numbers on your transmission, pretty much any of them will probably tell you something. Look for "518" on a sticker or something, that, or the huge casting number on the bell housing.

The electronics are mostly in the valve body. The only pressure you're able to monitor is governor pressure. Modulating governor pressure DOES change shift points, since shifts depend on how governor pressure interacts with throttle pressure. I.e., when you step on the gas, suddenly throttle pressure shoots way up since the throttle valve is directly actuated by the accelerator, there's a cable that hooks up to the transmission that goes to the throttle body. When throttle pressure is high, and overcomes governor pressure, it's gonna downshift. That's basically how it works. There ARE 1/8" NPT pressure test ports on the side of the transmission that you could fairly easily hook electrical pressure sensors too if you really wanted. I believe the pressure sensor is a variable resistor... I have a spare solenoid and sensor now since I replaced them both. The temp sensor is a VR too. RPM is a pulsed halls sensor.

On the Variable Resistance (VR) sensors like RTD's or Thermistor's, (not to be confused Variable Reluctors) do we know the thermal coefficient or profile of the sensor? Worst comes to worst I guess we could generate the curve, but getting it in paper somewhere is helpful.

Can you measure the DC ohms of the solenoid with a multi-meter?

I'll see what I can learn from my VIN, thanks for the tip.

I'm gonna eventually harass some Chrysler engineers to get answers.



^

Lots of great info in that PDF. I see qty 3 less than 1A drive solenoids, and generic temperature sensors. I didn't see the hall RPM sensor though. I'm also ignorant about some of the terminology, like TCC and mechanical TV. Where I come from TV isn't mechanical

I guess TCC = Torque Converter Clutch and I see a note in this video that you often can't just jumper it to GND and that OEM's have been turning them on to early which has lead to pre-mature failures

I think that TV meant Transmission Valve in that article.

Looks like those are common connectors on the tranny, do you know if OEM connectors can be obtained via TEconnect, digikey, or some supplier?

TV is the throttle valve. I doubt it'd ever mean "transmission valve", since automatics have an entire network of hydraulic valves that each do different things.

By the way, the truck's driving.

Speak of the devil though, I'm still having a few hydraulic issues. She's running and driving just fine, actually really freaking great, but a little funky in the WOT shifts (especially when flooring it when already at speed, maybe 50mph). Seems like if you floor it when cruising at 50mph, it'll wanna downshift into 1st gear (NOT GOOD).... Motor will rev to like 5500rpm, hit redline... Weird crap. It's supposed to shift into 2nd gear at 45mph WOT. And it does. But for some stupid reason, it's deciding to be funky if you're up to speed and you floor it. Other than that, it's doing quite well. I'm actually pretty sure having a custom controller for this transmission would be really freaking nice right now... If I could just tweak it a bit to raise governor pressure, it'd probably solve all my issues.

Fast forward to today, she's running AMAZING! No issues at all! So I'm guessing either the computer figured things out, or it just needed a little time to break in. It's shifting so quick now that my tires chirp and my accessory belt squeaks. I think I'm ready to begin working on the controller for this, and then move on to the engine.