I pretty much assume that every MPS I see is either broken or on its way out (diaphragm damage). I'm I right or wrong?

I think the next couple years will be unreal for engine losses.




M

I have about 12 units, one of which will hold some vacuum. I was really surprised to say the least. These had been collected about 15 years ago. I did buy two new VW units just for the parts. It does look like alternative FI systems are the way to go unless you really need originality.

Speaking of MPS's-- Are there not suitable later style ones that operate within the parameters of the stock teener one?? You know the usual, 5v ref, signal, ground?? Would be nice to just ripoff a late model chrysler or gm and use it.

If there was, and we knew about it, we'd have all been running them since that day..

Like if time machines were ever to exist... they would suddenly have existed, since the beginning of time.




M

Yea, I kinda figured that, but I also kinda figured that some people thought of it, and just did not go the extra mile to investigate---with 10,000+ differnt MPS's out there on open market, ....would be daunting. But all it really comes down to is specs of it..I would think a comparison of operating specs would reveal quite a few that could / would work.

Anybody got "All Data" or other form of auto repair software that would be intersted in researching this--would prove highly beneficial to 1000's of teener owners....

How about a virtual MPS, that fakes the signal, driven by something else?


M

That is easily done my making a pot and hooking it up to the throttle plate--oh wait--that is the TPS----problem is it is solid state, will not allow variables in altitude and loading up of the engine--that is the "limp home" feature on all obd cars--

it sees three conditions, right?

1. Idle

2. POT (partially open)

3. WOT

I'll just feed mine a WOT signal, and I'll get there in a jiffy, right?


M

QUOTE

5v ref, signal, ground??

Nice wish....read Brad Anders D-Jet page on how the MPS works and interfaces with the whole FI system and you might understand why eveything has been thought of already...

What we only really need is a diaphragm replacement...everything else would be workable.

Except, I have been thinking about a recent post by Geoff Bleyseng...Geoff are you out there?

You said your Janbo tester actually emulates the MPS...is that true or was it a misprint? If so, could the circuit/s just be reverse engineered and Voila...MPS substitute. How is this off base...?

QUOTE (jwalters @ Mar 26 2005, 08:26 AM)

I would think a comparison of operating specs would reveal quite a few that could / would work.

sorry, but if you did, you would be wrong...

we have BTDT through this topic a zillion times. D-Jet is *nothing* like a modern system.

it's based on an analog trick. the trigger points make the square wave that drives the injectors. the MPS is a variable-reluctance transformer. it uses vacuum sensing to move an iron core to swallow up more or less of that initial square wave signal.

i suppose you *could* build a device that could simulate the characteristics of a variable-reluctance transducer, but it would be hard, and it would not give you the desired effect of working with off the shelf components.

D-Jet was a clever analog computing engineering trick for the mid''60's when it was designed. but we've learned a bunch of stuff in 40 years, and we'd never approach that problem that way doing it now.

when you approach it with what's available now, you wind up with a MegaSquirt...

somebody out there has plains or a how to make a new diaphram that you have to calibrate to, doing the copper ecting is easy. the calibration is not with out the proper tools. i have read the how to on making the diaphram and it would be easy but the other stuff that might be needed is not like the coil, and the contacts.

"You said your Janbo tester actually emulates the MPS...is that true or was it a misprint? If so, could the circuit/s just be reverse engineered and Voila...MPS substitute. How is this off base...? "

No, its tests the circuits on the MPS and wiring. It emulates the ECU with a peabrain so you can test the ECU for failure.

Diaphrams are what are needed to repair the MPS. The hard part is the connection of the diaphram and the center piece/stop. I have seen new diaphrams that are soldered in, yuck! They run like sh*t too as that effects the throttle response.

I have even had Mercedes guys ask me to fix MPS's...there is no supply out there for the '70's cars that used the Djet.

If anybody wants to make a product this is it. There are enough of the other parts to last. Calibration using a Wavetek works thanks to Brad Anders. course the best calibration is with a A/F meter....

Bowlsby, get the Janbo from Chris Reale after he is done with it to play with. It even comes with a manual!

Geoff

okay who has a known bad mps, only known bad part in it being the diaphram. if some when sends me one i will see if i can reverse engineer the diaphram. i dont own any other djet stuff but i can try, i know how and have done copper ecthing and can get all other equipment to do it, but i need a "blown" mps that is complete to try.

edit: i might even try a ss diaphram to cut down on long term work hardening

Thanks Geoff, I thought it was too good to be true...I guess it was a misstatement.

Scott...Its not just copper...its Beryllium. Think Haz Mat. Go for it...there would be a large market for this if it were to become reality.

Here's the problem - money. Tooling to make a new MPS diaphragm is $5K-$10K. Even with the tool, cost each would be ~$150 due to low volumes, material (BeCu), and the threaded boss in the middle. That means that you have to sell them for $250 each, and then sell 50 to 100 just to break even.

These are just ballpark figures - but if you can find 100 people to cough up $250 each just for the diaphragm, then you are in business. Keep in mind, this doesn't include dissassembly, reassembly, and re-calibration of the old MPS - which most people are not capable of doing.

For that kind of money, I prefer to just buy used ones that hold vacuum for $20 each and re-calibrate them.

Demick

I have seen two MPS's with a SS diaphram, doesn't work or I should say it works but the throttle response sucks! The SS is too stiff compared to the hmm, I thought it was copper?

I buy good used units when I find them and install good diaphrams into 037 and 043 cases cuz people want the proper numbers for their MPS. Rebuilding/calibrating them isn't too hard about two hrs total time. I have rebuild about 20 MPS's so far....

Should be a way to get the beryllium sheet to die cut the round shape and stamp the pleating in. Pressing the center assembly on with a press shouldn't be too hard either just have to have the correct parts made.

Geoff

I went down this path partway with Brad Anders.

Making new diaphrams is, in fact, being done, but the replacements don't work exactly like the originals. I've heard a lot of dissatisfaction with the rebuilt units.

It's been suggested that all of the D-Jet MPSes are are the same internally, and only the sealed adjustments are what makes each one different (with a different part number). The MPS on my '69 VW looks exactly like the MPS sitting on the shelf for my '71 914, but the part numbers are different. Brad has done a lot of measuring, and it looks like the curves are all within the range of the adjusters, if I remember correctly. His site has a lot more detail on this.

As for a solid-state replacement: as Rich said, it's not that easy. Simulating an MPS perfectly electrically would be possible, but would require about 90% of the electronics of a Megasquirt to do it. The MPS really is the heart of the system, taking engine speed, temp, and manifold pressure and integrating the signals together to produce the pulses for the injectors. It's not just a MAP signal generator.

Over the next 10-15 years, I'd expect most 914s still on the road to be using some aftermarket EFI setup, replacing the D-Jet ECU and MPS with modern bits. Or they'll be switched to carbs. The only cars still using the stock system will be show cars.