[rusEfi] 1996 Dodge RAM 1500 V8 (#15)

[kb1gtt]

Frankenso R0.4 is getting close to it's next spin. Once it's done then it's like 5 to 6 weeks waiting for slow boats, and such. I believe the last hold up is a USB issue. It's been observed that when the dev console is connected via the STM discovery USB and in a real setup not on the bench, after something like 30 minutes it will fault the USB and you have to cycle the USB power to get it to come out of it. I suspect the issue is a GND issue with the PC, but it's all speculation on my end as I didn't have a test mule. Once I get alternator control I expect I'll see the issues on my test mule, and will be able to make a better analysis.

So my current set of priorities with this is to get IAT, MAP, ect going to the Frankenso board, then get a better tune on it. After that I'll get alternator control going. Then I'll see if I can reproduce the USB issue.

About alternator control, I it appears it's low side drive, and I'll assume it's driven by the same 400Hz pulse's that are noted in the below links. In that Neon we used a high side drive.

http://rusefi.com/forum/viewtopic.php?f=4&t=747&p=13364&hilit=400Hz#p13364
http://rusefi.com/forum/viewtopic.php?f=4&t=747&p=12829&hilit=400Hz#p12829

Does anyone have more details about what kind of current I can expect on this low side field control driver? I plan to use a low side drive, but those are technically only driving about 1 amp. Also they are over voltage clamping, I'll have to install a diode to change that to the more slug-ish snubber diode topology. I guess I'll have to clamp the OEM wire and take a measure. I wonder if it's PWM or analog control.

After alternator control is functional, I'll have one more low side driver. I believe that low side driver can be used to control the transmission solenoid. But first things first, get alternator control and get a better tune.

[Sam]

The problem is, if you only have one low side driver left... You'll need three since the governor, TCC, and OD are all low side solenoids.

[34blazer]

This should free up some I/O....

http://store.alternatorparts.com/hd-external-voltage-regulator-conversion-kit.aspx

[Sam]

This should free up some I/O....

http://store.alternatorparts.com/hd-external-voltage-regulator-conversion-kit.aspx

Lol, that's what I was saying to do earlier.

[AndreyB]

The problem is, if you only have one low side driver left... You'll need three since the governor, TCC, and OD are all low side solenoids.

The plan for right now is to use an external board for more io, but I agree that 0.4 needs more low side chips

[Sam]

This might sound crazy but I feel like, if the chip supports it and it isn't a big deal to support it software-wise, what would really be nice is if there was support for fully sequential injection and ignition on up to 12 cylinders, WITH room for extra IO. Or at the very least, fully sequential ignition and injection on a V8 with room for other IO, and maybe a V12 with no extra IO. I'm not familiar with the coding or our hardware limitations here since I'm very new to both coding and electronics. Most I know is some java stuff.

Regardless, if I could get my hands on even an older form of Frankenso, I could go ahead and start messing with my setup and trying to help out. I should have a lot of free time a couple weeks from now, and honestly I'd like to go ahead and start prototyping.

[kb1gtt]

I understand the software isn't the problem with 12 cyl's. It can currently technically do 12 cyl if you use wasted spark. However 8 cyls allow you to use those extra 4 low side drivers for other things. The problem has been partly that we drew a line in the sand to make sure we took manageable bites. In the early bits it felt more likely to succeed if we did 8 cyls and 12 seemed more intimidating. The issue we have now is that a change in scope like this would be hard to make physically. At some point I think we should spin a board that can do 12 staged injectors, with 12 native ignition channels, as well as 12 map sensors, and lots of extra IO for things like fan's, fuel pumps, NOX, ect. I also feel we should make a 2 cyl version for motor cycles and various small engines. However the short term goal and short term line in the sand is 8 cyl with 4 cyls wasted spark.

About OD and perhaps TCC, I understand them to be relay or push button options. I can add a relay to an unused hi/lo driver, and OD will always be on, as I never expect to not up shift. I plan for govt control on that last remaining PWM-able IO pin.

About Alternator control, the goal is to control this to help find and squash bugs. An external one would be fairly easy, however this isn't a track truck, it's a yard test mule. I thank you for the suggestions, but unfortunately the goal with this truck isn't to make it easy Well not easy for me at least, but my pain should make it easier for others.

[Sam]

Haha yeah, I'm not expecting you to go the easy route, and I admire seeing people do the hard things for the benefit of others. I'll probably go with an external setup for my alternator, MAYBE, or at least expedite it with some other way. As for TCC/OD, I'd imagine you wouldn't want OD to always be on, since that will mean you practically won't have 3rd gear. The hydraulic circuit allows OD to occur immediately when the transmission is in 3rd, so if you just leave it on, your shift pattern will be 1-2-4, honestly switches like you said would be the easiest route and really wouldn't be annoying to use for even daily driving. If it's possible, you could do some sort of logic or use another separate dev board like I mentioned ealier, and still let Frankenso control the trans, but have like, I dunno, signal pulses or something to the dev board? Sort of like serial communication, allowing you to use only one output on the Frankenso to control the TCC/OD solenoids..? I dunno.

Also, should I go with a four coil setup on mine and remove the distributor?

[kb1gtt]

I don't see much advantage of 4 coils vs single with a distro. Yeah you can blow a coil and still limp home, but you don't get much of an performance increase until you start COP and CDI kind of stuff. The CNP is basically just an EMI reduction. I would suggest keeping it to the distro as that simplifies the process.

[34blazer]

For simplicity the dizzy/cam sensor combo could be left alone, and can light off enough spark to surpass the limits of the OEM rotating assembly. Baby steps.

If the desire for a CnP or CoP system arises, an interface could be built. A lot of people also go for the ford EDIS-8 system as well, with universal code, the SAW signal can be made to work with this system.

Is there a parameter list and summary readily available? I'll help out wherever I can.

I like where this is going, cant wait to see trans controls!

Have you figured out the outputs for the gauges and other switches? Just curious...

[kb1gtt]

See top right section of schematic found here http://rusefi.com/wiki/index.php?title=Vehicle:Dodge_Ram_1500_1996 I expect I'll be missing Tach, and 2X mil lights. Long term I dream of a DIY display. I posted a kind of concept sketch here http://rusefi.com/forum/viewtopic.php?f=4&t=775 However I have not had time to pursue that concept.

I'm also thinking that to address the desire to have more IO, we would need an IO board that sits above those Wxx jumper things. AKA we can add a board to Frankenso which will offer additional features. There are at least 9 IO pins that aren't used, but have vias, which allows an IO board to be added.

[Sam]

Another question that 34blazer brought up, is barometric correction or knock detection/protection planed at the moment? If not, how hard would that be to do? With the OEM dizzy setup, do we have individual control over the ignition timing of all 8 cylinders?

[AndreyB]

knock detection/protection planed at the moment?

work in progress, not so trivial so it's WIP for a while already

With the OEM dizzy setup, do we have individual control over the ignition timing of all 8 cylinders?

please elaborate. what part of it would you consider 'individual'? different advance angle on different cylinders? why?

baro adjustment should be easy, but please tell me where does it fit into http://rusefi.com/wiki/index.php?title=Manual:Software:Fuel_Control ?

PS: we've totally hi-jacked a thread which has nothing to do with rusEfi plans.

[34blazer]

Hijacked? If so, should a new thread be started for calibration discussion?

For reference:

http://www.thirdgen.org/forums/diy-prom/256564-baro-da-calculations-730-a.html

An initial baro read will need to be accomplished for the BPC. A pseudo read for running correction. *edit* The VE multiplier is affected by BARO correction, AFAIK.


Individual ignition spark tuning is not needed, and would be a waste of writing and hardware.

[kb1gtt]

Knock has been problematic. In Frankenso R0.2, the TPIC XTAL get's into a no start condition, which has caused a bunch of headache. I changed a couple components and it seems to start reliably now. While mine is starting, who knows if that's just luck. The original design used a 4MHz XTAL and used a selection process for choosing several components relative to the XTAL. The fix was to do the same selection process, but use a 8MHz XTAL. It should have worked for both XTAL's, so I tend to think it's on the hairy edge of working with this 8MHz XTAL. R0.4 has been modified to will allow us to measure and see if we are on the hairy edge. For now, this questionable hardware has put the knock software development at a lower priority relative to other software development, as russian can make better progress on other items that have reliable hardware. Why fight hardware bugs while tracking down software bugs. Might as well isolate one at a time unless you run out of other things to focus on.

I suspect the inquiry about individual ignition adjustments are relative to knock. AKA if you sense knock on cyl 1, you can retard that timing a bit relative to the other cyl's and prevent knock for that cyl. As an engine gets carbon build up, or perhaps even from initial build tolerances ect, you can get different compression levels from one cyl to the next which changes the angle where it knocks. The first issue I see in developing such a feature is we need knock, then we need knock which can identify what cyl knocked, then we need individual cyl trimming.

I believe that is all physically possible with this platform, but has been only partly developed if developed at all.

About the Baro stuff. I generally expect you do not need baro compensation, as the sensors are typically absolute not gauge readings. When you get a gauge reading of 0, you still have 14.7 psi (ish) of pressure. If you used that for your predictions of O2 in the cyl, you would guess there is no O2 and would dump 0 fuel. So in reality, those systems assume 14.7psi. This is where a baro sensor comes into play. The baro will allow the ECU to know that it is 14.2psi or 15.3psi instead of assuming 14.7psi. I understand the baro compensation was far more common of an issue with carburated stuff.

However with most electric MAP sensors being absolute instead of gauge, a reading of 0 means 0psi not 14.7psi. Which means you don't have to worry about the baro compensation.

Is the OEM sensor absolute? What should we expect for the baro compensation? I have a hand held pump and needle style gauge. Can I remove the MAP sensor and validate the pressure vs measured values? I have no idea where I would find that sensor or if I can attach my hand pump to it some how. I expect it will output about 5V when just sitting there and it would produce 0V when in pure vacuum, with a reasonably linear transition in between those to end points. Is there a pressure vs voltage chart for this sensor I could review?

[kb1gtt]

Hmmm, how to get LSU4.9 into exhaust. I guess I'll have to remove the NB and installing this one. It has 2 pipes that go into one cat, so I can't put on a saddle or weld in a bung or anything like that. I'd really need 2x LSU's one for each bank. Perhaps I'm better off using the 2 NB sensors.

Does anyone know of a saddle or similar that makes it easy to install a bung? I can drill a hole and add a clamp on.

[34blazer]

I posted a couple of links earlier in the thread pertaining to volts vs. pressure. It's not about a gauge reading to determine differences in barometric pressure, the corrections are based on elevation. In the aircraft I fly the same applies, there needs to be correction, otherwise the fueling is way off. It may not be a big deal if the system is used in one place where there is little to no variance in elevation. My drives to work vary by 1500ft. When I was tuning my 7730 equipped mustang, at 2300ft elev the kpa readout from the MAP was ~96, at work it was ~89. I zeroed out the pseudo baro read table so that there would be no correction during transit, and the fueling was drastically different. BLM's climbed(added to BPW) from ~128 to ~140.

IMHO, as far as I'm concerned, there needs to be knock protection, doesn't necessarily need to be wideband to keep the configuration simple. These engines didn't have knock protection and it cost Dodge a lot of replacement engines, which their answer was the "death flash." MS incorporates GM external ESC for knock protection, and a fixed freq sensor...

http://www.megamanual.com/ms2/knock.htm

[AndreyB]

Hijacked? If so, should a new thread be started for calibration discussion?

For reference:

http://www.thirdgen.org/forums/diy-prom/256564-baro-da-calculations-730-a.html

An initial baro read will need to be accomplished for the BPC. A pseudo read for running correction. *edit* The VE multiplier is affected by BARO correction, AFAIK.

That's an amazing link and a nice summary, Forked into http://rusefi.com/forum/viewtopic.php?f=5&t=870

[AndreyB]

MS incorporates GM external ESC for knock protection, and a fixed freq sensor...

http://www.megamanual.com/ms2/knock.htm

We are currently trying to figure out hip9011 - there is an opinion that having knock detection react to RPM change and be aware of current shaft angle could or should improve the efficiency, There is hope that we are pretty close, but life is a bit in the way. If hip9011 would not happen, then external knock detecting module would be plan-B. Let's discuss this further @ http://rusefi.com/forum/viewtopic.php?f=5&t=778

[Sam]

Another really neat thing to integrate would be two step launch control. And yeah with what Jared said, that's where fully sequential ignition would be nice, with knock protection.

[34blazer]

Another really neat thing to integrate would be two step launch control. And yeah with what Jared said, that's where fully sequential ignition would be nice, with knock protection.

That could probably be accomplished by holding the EST signal high, and alternating the sequence to involve all the cylinders instead of just four. Its been done with code 59. Only uncertainty would be loss of sync between the crank and cam sensor?

[kb1gtt]

About MAP sensors, I had read the prior post differently. I read it as these are common sensors, not these are common for this sensor. I understand the sensor to be a 1bar 5V linear absolute sensor.

About knock, I'll have to add it some how. I guess I'll find a sensor, then find a place to bolt it on. Seems they are typically in the top center-ish of the engine. I don't like that GM sensor as it uses the engines GND as part of the signal path. Using the GND will create a poor Common Mode Rejection (CMR) I'll plan to use a more standard sensor which usually offers both wires.

About Baro compensation, I'll migrate the conversation to that other thread, and I understand the VE will change based on baro pressure. AKA if you have a baro of -14.7 PSI (AKA space vacuum) your VE will be 0% as there is no air. I'm however not sure I believe the math posted in that thirdgen forum post. I'll discuss that in this other thread.

[AndreyB]

Another really neat thing to integrate would be two step launch control. And yeah with what Jared said, that's where fully sequential ignition would be nice, with knock protection.

http://rusefi.com/forum/viewtopic.php?f=5&t=871

[Sam]

Oh, I actually meant to say anti-lag. I sometimes accidentally get them mixed up. You know, retarding the timing so you practically turn your exhaust system into a pulse jet, spinning up the turbo without actually making power, so there's no lag.

[kb1gtt]

Pinch of progress today. I soldered in wires for MAP, TPS, IAT, CLT and alternator. I also got the dev console to make basic readings of these values. However my next hurdle is to figure out what the fuel pulses should be per the MAP readings. AKA what is the flow rate of my injectors and what is the scale of the MAP.

For MAP I plan to take a reading before the engine is cranking, and I'll call this close enough to 1bar. Then I'll start it and get it idling and call this close enough to .2bar. AKA I just need this to run, I don't need it to run well. I would bet in reality that curve is not linear, but for basic testing I guess it will have to do. I have no complaints if anyone happens to know where I can find the voltage vs pressure graph, but for now I guess that's what I'll have to do for a MAP reading.

The next issue is I don't know how much fuel I'm dumping per pulse width. Does anyone know the flow rate for these injectors? It's kind of a pain to pull one and put it in a beaker. Not to mention I don't have a beaker. I could probably weight before and after to get a measurement. However that's kind of a pain. It would be handy if I simply found it on a datasheet somewhere.

If anyone knows the flow rate or the curve for this MAP, please let me know. In the prior message I don't see a curve, just basic comments about 1 bar and 5V. I need to know the curve for the intermediate points. These sensor are often not linear.

[Sam]

http://dodgeforum.com/forum/2nd-gen-ram-tech/177015-questions-about-the-map-sensor.html

That might help. I dunno. Flow rate for yours should be 23.2lb/hr injectors.

-Sam

[kb1gtt]

Perfect. From that forum post I see this

The Ram was cold and had sat overnight in Durham when I hooked up the ScanTool
and did a key-on but engine off MAP reading: 29.4 inches Hg/ 4.53 volts. My
cheap pocket hiker's altimeter also said that the reading should be 29.4.
Battery temp sensor read 73F.

With engine on, at idle the MAP read 11.7 inches Hg/1.55 volts.

From that forum post, and the various data point noted I get this

Dodge_MAP.PNG (10.64 KiB) Viewed 18467 times

I believe I should be good with your noted 23.2lb/hr injector.

[AndreyB]

http://dodgeforum.com/forum/2nd-gen-ram-tech/177015-questions-about-the-map-sensor.html

That might help. I dunno. Flow rate for yours should be 23.2lb/hr injectors.

243.6 cc/min, thank you!

[AndreyB]

My 2003 Dodge Neon MAP sensor is 0kPa/0inHg at 0v 100kPa/29.5inHg at 4.5v, that would make 25kPa/7.38inHg at 1.5v if my calculations are right. This is kind of close to your chart but not really the same. I wonder if these are the same and somewhere we have low precision, or if those are different.

Wonder if it's the same part number between your truck and my test mule?

[Sam]

I can't guarantee that's your injector size, but I'm 99% sure it is. If you have any trouble with it, let me know and I'll tell you some other possible sizes, since they varied from year to year. I'm willing to be 23.2lb is correct though.