Who's doing the physical testing?
Turbo Mini Forums

I have some injectors available from an MPi Mini.
These are known to be high-impedance, fast response, 30degree cone units that flow approx 480cc/min at 3bar.
I will make these available to whoever (Marcel?) on a 'return or purchase' approach after testing is completed...

I will be testing these on my injector flow rig within a month to see just how small a resolution they go down to (so we know when we're pushing the boundaries of idle duration) at different pressures from 2.5 up to 6bar - so anytime after that, they'll be available.

IF and thats a really fugging big if, i get some headway wiht current projects and get the turbomini finished and off my drive (its a mates) and find a job that actually gives me some free time i'll want to hack about with this a-sereis efi lark. even tho i no longer use the old lump its an interesting subject. in which case i'll be happy to help with testing and basic proof of concept, tho the only a-series(s) i have thats mine are sat on the floor, so no use under load :S (but easy enough to pop in a front subby with brakes sat in buckets of water i guess...)

anyways, just piping in :)

turbo 16v k-series 11.9@118.9 :)

Denis O'Brien.

Good to have you on board Evo...

I'm going EFi this year too :)

Alex

AlexF

Dave, Seven Magazine will be very interested in conducting feild trials and helping out with parts costs etc for those who want to develop and test this new MS code seriously.

Just to be clear, that does not mean we want any hold over the data or concepts, just to give back to the mini community and participate in the development and write up results to publish as we go along.

Going to email you now anyway about other magazine stuff.

The Seven Magazine turbo 850 will be running the new beta MS code in the trial as soon as I have had time to work it all out!

Peter

Editor

MAGA7INE

Edited by MAGA7INE on 5th Jan, 2006.

Peter

Editor

MAGA7INE
Seven Magazine
The brand new quality Classic Mini Tech and Journal
Issue 1 out late 2027.

I've been doing a lot of mental testing recently... Would that be of any use in the magazine Pete?

Edited by Vegard on 5th Jan, 2006.

Peter

Editor

MAGA7INE
Seven Magazine
The brand new quality Classic Mini Tech and Journal
Issue 1 out late 2027.

Peter,

Let me know if you have any question about the code or how to set up the new features.

Also, like I said in the other threads, I haven't been able to test the code due to my lack of access to an MSII so when you're ready to run the code (or anyone else is) , it's very possible there will some bugs. I'll do my best to help with debugging and/or explaining the code.

Regards,
Jean

http://www.jbperf.com/

Jean,

Can you be more specific about the MS hardware required to run the code.

I, and others no doubt, will be investing in MS kit soon. It would be a pity to buy the wrong version and add ons.

For example will we need v2.2 or v3.0? Do we also need MSII?

Has anyone also considered the physical installation of the injectors. Would it be practical to put injector bungs in the standard Metro Turbo inlet manifold or will it be necessary to manufactuer a special?

Saul Bellow - "A great deal of intelligence can be invested in ignorance when the need for illusion is deep."
Stephen Hawking - "The greatest enemy of knowledge is not ignorance, it is the illusion of knowledge."

Axel,

The code is written for MSII. The board you need depends mostly on your own physical setup, i.e., if you have very low-Z injectors, want to use the ignition control when the code supports it, have a VR sensor for the trigger or plan to have it when the wheel decoding is supported, or have a stepper motor for idle control, then the V3.0 makes more sense. If you already have a v2.2 or want the cheapest basic installation then the v2.2 board will do. You probably wouldn't even need the flyback board with v2.2 even with low-Z injectors if you only have 2 of them, i.e., no staging.

http://www.jbperf.com/

Jean,

Thanks

I was going to get v3.0 anyway as it's designed for greater reliability and more functionality.

The MSII is an extra £45 ish.

I'm thinking about using the Innovate LC-1 wideband lambda sensor. Looks like it will pay for itself if it is not necessary to use a rolling road.

Saul Bellow - "A great deal of intelligence can be invested in ignorance when the need for illusion is deep."
Stephen Hawking - "The greatest enemy of knowledge is not ignorance, it is the illusion of knowledge."

pff, vegard, you writing the story on my car? you sold it as the worst project in Norway

Yes I would agree Dave, analogue gauges are always superior for fluctuating readings, digital stuff either changes so rapidly it can be hard to tell in which direction the value is changing, or is damped to the effect that it updates so slowly that you don't truly know what is happening.

Analogue tells it as it is, and is much easier to "process" what is happening at a quick glance.

I think it's a mental thing aswell - you can look at a 'relative position' of an analogue output and see the movement. On a digital, your brain has to view the display and constantly de-code it. Of course, it shouldn't move that much! The AEM with the analogue and digital display just seems more 'ideal'...

Definately Dave, that was what I was trying to get at with the "process" in the sentence below.

Actually, the LC-1 does not have a display, although you can drive one from it.

I was thinking that for set up using Megatune, it would give significantly more information than a narrow band sensor.

If you can check the A/F ratio through a whole range of driving conditions and make adjustments on the fly, do you need a RR setup? If you can get 12.5:1 on WOT, then you know your fueled OK. Does ignition tuning alone warrant a RR session?

Megatune is configured to give an analogue display on screen, so no prob there.

Have we established if MS can deal with two lambda sensors yet?

Saul Bellow - "A great deal of intelligence can be invested in ignorance when the need for illusion is deep."
Stephen Hawking - "The greatest enemy of knowledge is not ignorance, it is the illusion of knowledge."

MSII does support 2 O2 sensors and MSnS-extra does too.

http://www.jbperf.com/

Edited by TurboDave16V on 8th Jan, 2006.

I'm a bit confused with what you say. You say that the best way would be to have 2 narrow band sensors and then you go on to say that you will use 2 wide band sensors.

The problem with using narrow band for setting up the injection events is that unless you're working with stoïc AFR, you're not going to be able to trust what the sensors tell you. Narrow band sensors are precise only at 14.7 AFR and only tell you you're lean or rich otherwise. There could be quite a difference between actual AFR for the same measured voltage between 2 sensors.

In addition, you have to know that the pressure before the turbo will affect the measurement made by the O2 sensor (narrow or wide band). Of course if you only want relative measurement, that's OK, as long as you're not relying on a narrow band for rich or lean AFRs.

By the way, the wide band sensors are not that expensive and can be had for less than US$40. But the issue about high temperature and the pressure I mention above would make calibration non-trivial.

Regards,
Jean

http://www.jbperf.com/

I don't know why i had 'narrow' in that sentence !

Now corrected!

I didn't know that exhaust back-pressure affected the sampling of a Lambda sensor. If you're sure (?) about this then fair enough! At least i was right in saying that the one in the downpipe is what the tuning should be done with...


I realise WBO2 sensors themselves might not be too expensive, but i got the impression reading through the AEM blurb that each sensor is calibrated to the gauge (or AEM control unit)?
Certainly, there seems to be an awfull lot more to the WB circuitry than the NB equivalent.

Yes, I'm sure about the backpressure affecting calibration. You can have a look at http://www.megasquirt.info/PWC/index.htm

This covers a lot on wideband O2 sensors (probably more than you want to know :) )

There is a need to calibrate the sensor but I believe that most controllers have a built-in way for the calibration. So it is possible to change the sensor and keep the controller. I don't know if the AEM unit is different.

http://www.jbperf.com/

With regards ignition timing, i did mine without a rolling road and used a spreadsheet that i made that seemed accurate enough for me to find the minimum advance required for best torque, with repeatable results. it required a flat road and no wheelpin. that and a set of home made det-cans (32p piezo crystal, some insulating tape to insulate it electrically from the block and an amplifier, i.e. mic input on the laptop jobs a goodun) and you can map away. there is a very obvious change in the sound when det is about to start. it analysed the raw megasquirt datalogs. if i get some time (not enough hours in teh day) i'll write some instructions and try host it somewhere. if anyone wants a copy the now to try drop me a pm and il ge tit off the laptop. btw little disclaimer, just cause it worked on my engine doesnt meen itl work on yours, purely depends how well the engien responds to ignition advance/retard. I also used this to tune my AFR as my narowband was a bit vague when it came to fine tuneing. cant afford a wideband just yet :)

turbo 16v k-series 11.9@118.9 :)

Denis O'Brien.

turbo 16v k-series 11.9@118.9 :)

Denis O'Brien.