Author Topic: M42 VEMS PnP (Read 7317 times)

keflaman · « **on:** December 30, 2015, 06:19:30 AM »

I saw this thread and thought there may be some interest in it when I saw that the M42 engine was added.

http://www.e30tech.com/forum/showthread.php?t=120330

gstuning · « **Reply #1 on:** December 30, 2015, 09:56:32 AM »

Hi guys.

I am the organiser of the group buy, I joined so I can make sure to answer any and all questions there might be.

bmwman91 · « **Reply #2 on:** December 30, 2015, 08:14:28 PM »

Wow. OK. This makes me conflicted.

I have been building up plans to run a Megasquirt 3 Pro module in a motronic case and build the necessary adapter hardware. However, this is PnP, and it addresses my two main beefs with Megasquirt (namely that it does not have built-in WBO2 control or ignition coil drivers). So VEMS would eliminate my Spartan WBO2 controller and the need for BIP373 drivers.

1) However, my concern is regarding the age of the hardware. VEMS is well proven and all, but the hardware looks like it is at least 5 years old, whereas the MS3Pro uses a much faster and more current MCU. Is there going to be any actual difference in terms of response speed and performance between the two systems? I realize that this may feel like somewhat of an unfair question since the hardware is what it is, and both are probably better than the old Motronic 1.7.

2) Also, what sort of MAF support does VEMS have? MS3 had some issues with resonance and backflow, but because it is open source a solution was worked out on the forums and put into a later firmware release which made a MAF work flawlessly with it. For me, a MAF is really sort of mandatory since it means that the load calculation is not dependent on the volumetric efficiency curve (like in the case of a MAP sensor).

3) Regarding additional injector driver channels and whatnot, can a buyer modify the internal wiring so that fully sequential injector control is possible (the stock setup pairs up injectors 1 & 3 and 2 & 4)? I am planning to add a couple of extra wires into the harness so that I have 4 injector channels.

4) Aside from that, is the VEMS board inside there going to have all of the usual drivers for relays and whatnot so if an M42'er wanted to use some of the additional VEMS features they could wire them up to the inside of the 88 pin connector?

Thanks!

gstuning · « **Reply #3 on:** December 31, 2015, 02:01:52 AM »

Good questions

1. The VEMS is capable of tracking ignition timing to well above 12k rpm, there is no actual run time difference when it comes to timing sparks, timing fuel or any other calculations.

2. There is no MAF support, I have tuned 100+ engines without MAF and a few with and I cannot see the benefit at all when it comes to standalone tuning, also the ecu has wideband closed loop so it�ll keep your lambda on target. You can tune a VE table in less then 30 minutes on the road to almost perfection. The VE table output is sent to the Lambda target output table which modifies the output to correctly adjust the "load" for lambda targeting as in a load based system.

3. The auxiliary connector comes with 2 high power outputs and a relay output so possibly it could (I�ll verify). Sequential injection is good for emissions and some economy and a little better throttle response (meaning less fuel to get the best possible throttle response) which anyway is always poor on a single throttle body system anyway. The injection sequencing on a 4cyl semi sequential is actually good since the injection is always on a closed valve for optimum vaporization.

4. as in 3, http://www.vems.hu/files/M50Turbo/scn011.pdf , you have some inputs and outputs to play with already.

bmwman91 · « **Reply #4 on:** December 31, 2015, 02:47:53 AM »

Thanks for the info. This is actually starting to sound much more tempting than MS3Pro since it would in fact be much more cost-effective.

If I was to run a MAP setup, where would I be looking to tap into? Would the main plenum (on the vacuum side of the TB) work, or would it be better to have 4 small taps in each of the manifold runners near the intake ports? I'd like to get the best possible throttle response.

I am not super familiar with the connectors on the VEMS. Basically, is this PnP kit a full-featured VEMS unit (8 injector drivers, 8 coil drivers, etc) stuffed into a case and mated to an 88 pin connector? If so, is it connected via a custom PCB with the VEMS board and 88 pin connector footprint on one edge, or is it a stock VEMS board connected to the 88 pin connector with a bunch of hand-soldered wires? The latter (wires) is preferable in my case since then I can make the various customizations that I would prefer.

Sorry if I am being a pain in the ass. Thanks again for the info.

gstuning · « **Reply #5 on:** December 31, 2015, 05:34:57 AM »

The map input is the same as the FPR, the throttle response is a function of fuelling against TPS rate of change at various rpm's, so MAP measurement point is not crucial.

I am working towards using a M50 non vanos PnP to run a M42, this is infact the best possible option and gives the most versatility possible for a M42 user.
This would mean that the remaining injector and ignition outputs are actually mapped to the 88pin connector. The M50 PnP caries a custom VEMS PCB made specifically for the 88pin connector so there are no wires inside to tap into.

So in short you can infact add wires to do sequential injection this way. i.e. 8 total injector outputs (2 used for std injection, 1 for idle control valve, leaving 5 available in the 88pin for other outputs, 3 when using sequential injection) also 2 extra IGBT outputs.
So all in all 7 high power outputs for whatever along with all the auxiliary inputs and outputs.

I do not mind the questions at all.

bmwman91 · « **Reply #6 on:** December 31, 2015, 03:48:48 PM »

Sweet. So it will come in PnP form, and a user can optionally install jumpers between the unused pins on the 88p connector and solder points on the VEMS board (and it'll all be one PCB anyway)? You know, I am starting to think that this might be less of a headache than MS3Pro, mainly because the ignition drivers are built in. Anyway, is it safe to assume that the throttle response with this will be leaps and bounds better than the Motronic? Even a 1995 US-spec M3 feels like lightning fast throttle response compared to the M42, and it has bugged the hell out of me.

How hard would it be for me to switch around the WBO2 connections so that they run through the stock harness? I'd need to add a 5th wire of course, and maybe run shielded cable for 3 of the 5 wires, but it would be super easy to remove the stock 4-position twist plug and crimp on the terminals for the LSU plug and install the LSU plug housing on there (since I happen to own an array of open barrel terminal tools).

If you want any help or input with tweaking this to work with the M42 harness feel free to email/PM me. I would be happy to help with some of the planning, with no expectation of a discount or anything...I am just stoked at the possibility of FINALLY having a nice PnP stand alone setup for the M42. I have been going over my spare engine harness for a while now in preparation for the MS3 Pro, as well as getting an oscilloscope on most of the DME's inputs and outputs to fully characterize them. At the very least I am happy to share that stuff since it will help everyone.

gstuning · « **Reply #7 on:** January 01, 2016, 08:31:07 AM »

1. No jumper wires, the outputs are in the 88pin already, you would need Junior timer connectors and wires to add outputs.

2. I can�t promise anything with throttle response, only that is can be as good as possible the manifold arrangement allows.

3. You could take the lambda wires coming out of the ECU and splice them into the M42 loom via. No internal options are there so you can route them internally into the 88pin unless you open the ecu yourself and start cutting PCB pins and routing wires that would normally go to the Wideband connector.

4. The M42 is a simple thing,
VR crank, VR cam, all outputs are grounding, all inputs are 0-5v and thats really all you need to know about it, no need to scope anything really.

bmwman91 · « **Reply #8 on:** January 02, 2016, 01:02:06 PM »

1) OK, I think I have a good idea of what all would be going on in it. Thanks. Now I just need to call a couple of the local tuners in my area and see how comfortable they are with VEMS since the whole reason I am going stand-alone is that none of the local guys can deal with M1.7 and my WOT maps need some work (and I want better throttle response, but that's secondary).

2) Understood. I think that throttle response HAS to improve with a more modern ECU. The 1995 US-spec M3 I drove has a similar plenum/TB setup and the response is like lightning in comparison.

3) That's what I figured. Since this sounds like it will be a custom PCB, would it be possible to run some WBO2 traces either to the factory pins or to 5 unused ones and have jumper headers inside for the customer to choose with?

4) Indeed, it is an old and simple system. No CANBUS nonsense, just analog signals really. I have scoped a bunch of other stuff including the Motronic's tach output signal, the fuel rate output signal and the vehicle speed input signal. One of my goals with whatever stand-alone I go with is to have the instrument cluster be fully functional (and yes, I am aware that the MPG gauge is sort of superfluous lol...but I am going for a very OEM++ type build). Since VEMS has lots of spare outputs, how hard would the following 2 signals be to generate for the instrument cluster?
--1) Tach signal: Signal is idle at ground, goes to 9.0V for 1.4ms twice per crank revolution. So, two 1.4m high-pulses per revolution. Can VEMS be programmed to do this? The voltage may or may not be important, so I will mes with a signal generator and see if 5V and/or 12V pulses work properly too.
--2) Fuel consumption rate: Signal is idle at 8.1V, goes to ground for 1.6ms after each injector firing event ends. So, one 1.6ms low-pulse per injector event. Can VEMS do something like this with a spare GPIO? Again, not sure how important the voltage is and I can confirm with a signal generator.

5) What sort of Check Engine Light control ability does VEMS have? In CA, they check to see if the CEL turns on in accessory mode, and that it is off when the car is running, so I need it to appear to be working if I want to be able to register the car lol.

At this point I know that I am being a pain, so no worries if all of this is superfluous. It's obviously not required to make an M42 run well with VEMS, and it also means a lot more PCB layout effort for you. I'm sort of anal about wiring and that sort of thing, and it is highly likely that I will just need to build my own "PnP" stand-alone ECU from either a VEMS board, or MS3Pro.

Here are some general ideas for nice-to-have features (in my OCD mind anyway):
a) 2 additional injector outputs connected to the 5mm JP pins adjacent to the 2 factory ones so that people that want to run sequential just need to crimp some JPT terminals and run a couple of wires. Then it just becomes a software setting to use those.
b) Fuel rate output pin on the connector hooked up to a spare GPIO for those that want to mess with it.
c) Tach output pin on the connector hooked up to a spare GPIO for those that want to mess with it.
d) Vehicle speed input is connected (web searches indicate that the factory E30 signal is no problem)
d) Traces and jumpers for the option of running the WBO2 through the factory harness (and the user has to install the extra JPT terminal and wire).
e) All other spare Inputs / GPIO's / driver outputs connected to unused 2.8mm and 5mm (drivers) JPT pins.

Thanks again for all of the time to deal with my nagging, and more so for actually working to bring M42'ers a PnP solution. As it is now, it sounds like what you are planning will be great for 95% of M42 guys that want stand-alone solutions!

gstuning · « **Reply #9 on:** January 02, 2016, 03:11:30 PM »

A M42 PnP will be made entirely from a M50 PnP PCB , not a new PCB design.
It will allow more outputs but not all the features you have mentioned due to PCB alterations etc.

3. Thats for one person and not everybody else so in reality no.

4. Tacho works naturally,
the fuel consumption will not work, and 1.6ms fixed per event would never give any fuel amount rate change right?? I suspect if you measure injector opening time and the fuel consumption you will find a strong relationship.

5. Download the software and you will see what options there are
a) to add injectors you would simply take JP connectors, crimp the wires and press them into the M50 PNP injector outputs then route to your new injectors.
b) tacho already works perfectly
c) VSS already works
d) won�t happen
e) they are already along with the 10 pin auxiliary connector.

Its a PNP not a development setup, that would be best done with a regular VEMS.

bmwman91 · « **Reply #10 on:** January 02, 2016, 04:57:36 PM »

Sounds good, thanks.

I rechecked a few different logs and it does seem like the fuel consumption pulses are actually ~43% the duration of the actual injector pulse. That makes more sense, and agrees with what you said. I guess that if it is that important to me, I can just make a little microcontroller daughterboard that does what I want.

So, that's it for me trying to tell you how to do your job. I will play around with VEMSTune since I bet that I'll answer a lot of my own questions that way.

gstuning · « **Reply #11 on:** January 03, 2016, 07:52:39 AM »

Excellent.

I will keep looking at this thread to monitor for questions from you or anyone else. I thoroughly believe this is the ideal solution for most M42 projects.

Author Topic: M42 VEMS PnP (Read 7317 times)

keflaman

M42 VEMS PnP

gstuning

Re: M42 VEMS PnP

bmwman91

Re: M42 VEMS PnP

gstuning

Re: M42 VEMS PnP

bmwman91

Re: M42 VEMS PnP

gstuning

Re: M42 VEMS PnP

bmwman91

Re: M42 VEMS PnP

gstuning

Re: M42 VEMS PnP

bmwman91

Re: M42 VEMS PnP

gstuning

Re: M42 VEMS PnP

bmwman91

Re: M42 VEMS PnP

gstuning

Re: M42 VEMS PnP