GDI drivers

[busher06]

Hello all,
Looking at using MS3 to run a GM LLT 3.6L V6. What is the latest thinking on GDI drivers. The injectors are p/n 0 261 500 056 and measure 1 ohm with digital meter. The nearest match found in the Bosch catalog is the HDEV 5.2 (1500mOhm / 200 bar max ) which uses a booster supply of 65~ 90V, Booster current 13.2A and a hold current of 3A.
Short of a small welding plant, is there a stand alone circuit that could provide this power and be controlled by MS3X Drivers?

[piledriver]

Can you PWM an IGT? perhaps using one of Jeans injector drivers or a modified variant?
(no idea, I'm looking at the Orbital air assist stuff at least somewhat to to avoid the issue, it uses std high Z fuel injectors, haven't found info on the air pulse coils)

Here, Courtest of some Chinese students:
file:///home/greg/Downloads/09-IJEI8-082002JE_1_.pdf

...Looks like they used a good old 555 timer... Seriously.
I'd bet the PWM chip Jean used for low Z injectors could be adapted.

[Matt Cramer]

I've seen some speculation that this could work:

http://www.liferacing.com/?page_id=35

But to the best of my knowledge, nobody's tested it with MS3.

[racingmini_mtl]

I'm not sure that the chip I use could be used without generating way too much heat in some components. However, I'm looking at another chip that might be adaptable for something like this.

Jean

[dontz125]

Here, Courtest of some Chibese students:
file:///home/greg/Downloads/09-IJEI8-082002JE_1_.pdf

Those crazy Chibese students, storing stuff on your home computer!

[busher06]

Jean,
As the resident guru in this field, any help would be appreciated. As GDI systems become more commonplace, I am sure MS/JBPerf will have a say in what develops.
Thanks again.

[racingmini_mtl]

I am very far from being a guru about GDI and I still need to gather quite a bit of information before I can even start to make educated comments or provide help. I'd certainly be interested in providing some products but there are reasons why the only available solutions cost many thousands of $$$. And injector control is only part of the solution because you also need to control pressure and select the correct injection strategy.

Jean

[busher06]

I am also on a steep learning curve. Synchronous vs non- synchronous HP pumps; What happens when you move the cam; the vital importance of accurate inj timing in sync with cam position,....We have been here before, and I am sure that the collective wisdom will prevail. My car uses about 200 lbs of ballast so an inverter tig welder is not yet ruled out.First step would be a 2kW alternator.

[piledriver]

I am also on a steep learning curve. Synchronous vs non- synchronous HP pumps; What happens when you move the cam; the vital importance of accurate inj timing in sync with cam position,....We have been here before, and I am sure that the collective wisdom will prevail. My car uses about 200 lbs of ballast so an inverter tig welder is not yet ruled out.First step would be a 2kW alternator.

I know you were probably joking but I must point out the typical TIG welder puts out about 14-17v under load.
(The boat anchors like my old Miller Synchrowave will go much higher at no load, but inverters do not as a rule)

I do happen to have a 4.5KW alternator sitting in a box for a riding welder I have yet to get around to building
(350A Ford ambulance alternator, ludicrously heavy lump would go a long way towards that ballast ;-) )
Run the pre-diode AC off the alternator and run it through a transformer? Kinda high/wide frequency range, but it should be relatively trivial to make the required 90v DC. (some decent caps and a <~1000W 110 v inverter and bridge + regulator should suffice, you only need 13A for a few ms, then 3A PWM)
Opto-isolators required of course, but that sort of thing has been done many times for many other applications.

I'm still looking for reasonably sized orbital air injector bits//specs. at least the fuel side and air pulse triggering are relatively trivial.

As to fueling timing/strategy, AFAICT the only time it gets critical is when trying to run stratified charge or idle, under any reasonable load you start injection during the intake event immediately after EC to prevent fuel short circuiting out the tailpipe on overlap.

[MS2tester]

how handy are you with a soldering iron?
http://www.aicit.org/ijei/ppl/09-IJEI8-082002JE.pdf
http://www.abaxor.de/publications/icit/ ... e_ICIT.pdf
http://www.bosch-semiconductors.de/medi ... driver.pdf
http://freedatasheets.com/downloads/App ... AN2736.pdf
http://www.aicit.org/ijiip/ppl/07_IJIIP-073059JE.pdf

[piledriver]

A breif Googleing cannot find a source to buy any of the chips or the STM eval board (discontinued) datasheets still contain excellent info.

[busher06]

Thanks to you MS2tester and piledriver. I have a lot of reading to do before I fire up the branding iron. It all seemed so simple from a position of abject ignorance. A cursory glance at some of the info seems to indicate that max power can be generated at Lambda =1. Is this so? My immediate goal is performance, but if economy is also available, then I am even more intrigued.

[piledriver]

The primary performance advantage of GDi is that all fuel evaporative cooling effects occur in the cylinder, increasing VE.

This has the splendid effect of not only increasing VE but also DEcreasing the propensity to knock greatly, allowing higher CR... which further increases efficiency/power.

...It also can eliminate fuel short circuiting out the tailpipe on overlap, which allows ~wilder cams and emissions to coexist to a point, allowing MORE static CR (or boost), further cranking everything up another notch in a lovely upward power spiral.

Best power is still ~12.5:1-13.1:1 or thereabouts.
If all you want is performance, all you need are excellent knock-resistant combustion chambers and std pistons, as the fuel shot is ~all during the intake stroke, and the little "magic" dimple in the dome is ~only useful to contain the charge for stratified charge modes... It's nowhere near the injector during non-stratified mode use, so useless for performance only apps.

I actually suspect the stratified charge mode would be ~hard to pull off with most current aftermarket EMS,as it almost requires throttle-by-wire, as it runs pretty much like a diesel with the throttle wide open, controlling power only via fuel volume, at ~100 KPA manifold (or possibly under mild boost)

Hmmm...I suppose one could begin with a common engineering mind excersise:How to do (something) with what exists.
Assume engine runs in AN mode ONLY ran in stratified charge mode up to a certain load,(WOT on computer driven I/O)"
... then close the computer controlled throttle, swicthing to SD (near stoich or richer) and making the secondaries (now fairly wide open) manual throttle butterflies effective as the AFR drops towards stoich and richer. (sort of like vacuum secondaries, only different )

You could even have the computer controlled throttle reopen at (input) WOT above a certain rpm to change the manifolds tuning, but hitting the brakes or zero tps input would drop fuel flow back to idle or cut.(interlock)
Those last bits are already in the firmware or relatively trivial. A manual slamshut for the computer controlled throttle of some fashion would also be prudent, have to work that out, redundant TPS might be mandatory. (they exist)

I'm still not sure I could convince an MS3 to do that without some minor firmware fiddling, but I'd like to try.

Mmmm I wonder if any of the new FSI VW heads will fit on my old 1.8L 86 Cabby or 2L ABA block...
(Yes, it will. apparently one of the oil returns in the head needs welded shut. That and a different intake manifold and probably relocate the alternator to the 16v location w/stock 16v brackets, and ~10:1 with the 1.8L flattops.)

Good old VW never actually redesigns anything they don't absolutely have to, and even then as little as possible.

[slow_hemi6]

Some Mazda 6 mps had a separate direct injection module. There is a bit of info on those out there. I think I recall seeing pdf's on the motec site about it.
Yes, here is the thread with links that I remember. http://www.msextra.com/forums/viewtopic ... 09#p276309

[piledriver]

Here, Courtest of some Chibese students:
file:///home/greg/Downloads/09-IJEI8-082002JE_1_.pdf

Those crazy Chibese students, storing stuff on your home computer!

Sorry, ddin't catch my misteak: Too big to upload
https://redmine.fe.up.pt/attachments/do ... 2JE_1_.pdf

[piledriver]

Some Mazda 6 mps had a separate direct injection module. There is a bit of info on those out there. I think I recall seeing pdf's on the motec site about it.
Yes, here is the thread with links that I remember. http://www.msextra.com/forums/viewtopic ... 09#p276309

Dont get too exited (yet) about the alleged picture in the Motec link, the part# is clearly visible and is listed as an HVAC/entertainment module. I cannot find one that is obviously the proper unit online by name, or cheap.
OTOH ~every module seems to be an HVAC/entertainment module in the jim ellis mazda parts catalog, so it may be legit.

Mazda part# is (AFAICT) L3k9 18 941, early? 2006 Mazda 3 TC (not USA, perhaps Canada, Europe (Mazdaspeed 3) other designations for some locales)

OTOH a diagram in the docs shows a dead simple way to cascade a 12 and 100v driver with an isolation diode... the 100v pulse out would be a fixed triggered PW one shot spike on inj in, then the 12v hold driver takes over and finishes the inj drive time... (the one I'm refering to is not actually a schematic, simply provides an idea how it works, but that might very well actually work with only minor supporting circuitry. That triggerred initial 100v pulse is likely also what the Chibese students were up to with the 555 timer...
The resulting current profile would work out nicely with all(?) the common high pressure GDi injectors out there...

It's likely to be a dead simple little driver box, I'b be fine with a dead one to reverse engineer... Although I bet Jean could have something suitable whipped out in no time using parts dug out of his sofa

Does dry firing Bosch GDi injectors damage them?

[piledriver]

This is the manual for an (unfortunately) obsolete GDI driver eval board from ST, includes schematics.
Have had no luck finding the eval board or the chips at retail (at all, actually)

http://tec.icbuy.com/uploads/2010/6/30/AN2736.pdf

It contains actual full schematics with components and values, so the GDi drivers and flyback PS design bits have a man on base or three.

Freescale makes a solenoid driver chip that is mentioned as a suitable controller for GDi output drivers.
MC33816
http://www.freescale.com/webapp/sps/sit ... on%20Notes
Might be paired as the input/controller and driver with circuits similar to the above.
Some Assembler required.

I'm kinda leaning toward wanting dumb, discrete GDi drivers, can be fixed and troubleshot, by many humans.

[piledriver]

A lister on my "home" forum pointed out that ~all the 1st gen GDI Mitsubishis had a ~dumb stand alone injector driver box...and Delphi injectors. These go for ~$50 on ebay, at least until I actually went to buy one, they all seemed to vanish.
I guess Ill risk finding out what that Mazdaspeed 6 turbo box actually is (L3k9 18 941 from the motec forums link) for $43 shipped.
It IS the GDI injector control module, now I just need connectors and pigtails...

A search on ebay (with some further verification) show this ECM>stand alone DI driver box config was used by Ford (2000-2003 Powerstroke 7.3 turbodiesel) Lexus (IS350) and others until farirly recently (2011 for the Lexus), some models may still be that way.

A reman PS DI driver box can be had for $125. I'm curious if that would have any problems driving lowz GDI injectors... Researching.
(Piezo type injectors, not sure of the voltage/current waveform so blow that off for now)

The later model versions may prove problematic if they happen to be too "smart", still a fine example of something.

[MS2tester]

if you have deep pockets you can get the driver box
http://sine.ni.com/nips/cds/view/p/lang/nl/nid/210063
but at least the manual might give some more information about injector requirements
http://www.ni.com/pdf/manuals/375969a.pdf

[piledriver]

No deep pockets around, I suspect I just got what I needed for $43 BUY IT NOW...
I may have something for really interesting bench testing real soon now...