MAP sensor discussion.

[JSMskater]

to avoid another hijack of the temp sensor FI issue thread-- lets continue the discussion here.

MAP sensors have always been a point of headaches for D-jet owners-- either because someone has screwed around with the adjustments not knowing what they were doing, or because they don't work right from engine to engine, or because they simply fail and you can't find a correct replacement.

here is the summary from the other thread:

[KTPhil]

They were subcontracted by Bosch, who I presume calibrated them using an induction meter while the unit was drawn to a certain vacuum. VW would not have duplicated this setup on the production line. I can imagine Bosch had some special "test plugs" and special nut/screw drivers that let them get at it and adjust, then seal it up before sending it to VW.

[JSMskater]

[KTPhil]

They didn't test it to an A/F ratio; that would require them to be on a running engine. Bosch just made the sensors as an end item. The MPS works by creating specified inductances at a varying vacuum levels. That is what they would have calibrated. It is a curve, not a single value.

So perhaps those that are adjusted to a given tolerance were then graded, and those that were towards one side of the spec or another got different paint codes.

There is always a manufacturing cost tradeoff between making them with precision (expensive) or testing and grading them. Perhaps the California models, with their stricter emissions regs, used those of a certain paint code. I don't know, because I have never seen those witrh paint codes. I presume they are for the E-systems?

As they age the metal diaphragms begin to react differently to vacuum. Many crack, and they also can stick. This will not change their static resistances in the coils, but it can strongly affect the inductance as the parts move inside.

[JSMskater]

[KTPhil]

In theory the inductance is best, since it is immune from other system problems that might change over time, like out of tolerance sensors or imperfect injectors. But indictance meters cost $100+ and Im not sure anyone has the curves to calibrate them anyway. At best you could match a known good/NOS sensor. There aren't many of those around.

If you are really adept at adjusting all three settings of the late MPS, and have a good A/F meter, and have all other components in great condition, you could calibrate them periodically. There are good web references on what to adjust for what effect, but yu would have to be really careful to eliminate test variations or you end up with a poor baseline as you sequentially make changes. Also, some are measured under severe load, and without a resistance dyno, you can't duplicate them in the garage.

How does Jim adjust them? Does he match static inductance to a known good MPS?

I'll have to check my cources, but I think the late units just have an extra contact that reacts when the unit bottoms out at full throttle, just liek the old pressure switch used to do. There were a lot of changes on those last couple of years as they chased Cali/Fed regs, so I'm not sure.

[Bobnotch]

Jeff Bowlsby (the guy who is doing my harness) also calibrates MPS's. Just so you know. I don't have a clue how he does it, but he's located in San Ramon, Cali.This might be close for some of you here.
_________________
Bob 65 Notch S with Sunroof
71 Notch ...aka Krunchy; build pics here;
http://www.thesamba.com/vw/forum/viewtopic.php?t=249390 -been busy working
64 T-34 Ghia...aka Wolfie, under construction... http://www.thesamba.com/vw/forum/viewtopic.php?t=412120

[JSMskater]

[raygreenwood]

I'm sorry I never got around to putting in my 2 cents. I just found a note in my mailbox about this thread from oct 29th. Been busy.

The MPS's from Bosch (as noted, they were subcontracted) were adjusted via induction with a set vacuum reference or reference(s). You have to remember that production started in 1967/68. Exhaust sniffers....much less repeatable and accurate ones did not grow on trees.

I did not start seeing paint marks on the MPS's until about 1971 models. I saw them mostly on MpS's out of 914's and a few of the late 1972 411's and the last of the D-jet 412's.
In other words...it seemed apparent that VW did not really realize where their error was until late in the game....at about the point the 412 went L-jet...wherin it became a moot point. I have a bit of cross referencing at the time....as some of the Renaults I came across also started seeing paint marks on the otherwise sealed MPS's.

So why did they not just identify them via part number?
Because that cannot be done.
You cannot just have a specific part number for say....a 72 411 with a 1.7L engine, auto tranny with "x" gearing, distributor number of "x" and emission equipment of "y".

Why not?...
because every single engine built to that spec. will have a slight and subtle enough vacuum signature even off the assembly line.....to make an almost endless number of possible variations even within one part number of drivetrain package.
To be exact....most of the engines ran the same enough right off the assembly line...that a few simple adjustments elsewhere (a psi or two in fuel pressure, a tweek to idle or a degree or two of timing)...would be enough to make them all run generally the same in the eyes of the buyer.

Where this really came into problems is that depending upon how the engine is broken in, and then how its driven or tuned....meaning how it wears.....made potentially large differences in vacuum signature by somewhere around 30-50k miles. Thats when owners start noticing idle and emmission issues and start coming back to the dealer.....whose only option is to tune as closely as possible....and then start swapping MPS's off the shelf until they get one that makes a difference.

The first inkling of these difficulties in dealer tunability....was seen in the last of the D-jet with "E" spec MPS's and ECU's....that had the little gray knob for adjusting idle mixture.

Its also helpful to remember that every engine is different and one engine may have an "at idle" vacuum signature that is virtually identical to hundereds of other engines...but totally different through midrange and part throttle or at full load......and/or any combination in between.

The D-jet MPS....is a bit more than just a MAP sensor. There are a lot more dynamics going on than meet the eye. There are three main adjustments.
There are also two completely different but intertwined functions going on as well. In the late (copper diaphram style) MPS's....you have not only general engine load sensing (like a regular MAP) but you have full load and partial load extra enrichment for positive throttle movement (the copper diaphram). In earlier models (type 3 to about 1969/70) you had a separate diaphram unit to do this).

Think of the MPS as a very adjustable MAP sensor.
(1)The copper diaphram has a separate adjustment. It works by allowing the main armature shaft and teh aneroid can it is pushing against to unload FASTER as you open the throttle. In other words it moderates RATE or velocity of unload.

Though the aneroid cans are actually more sensitive to slight fluctuations in vacuum than the copper diaphram (more subtle and more of a control function)....the copper diaphrams reaction is quick ...and is additive to whatever reaction happens in the aneroid chamber. Its more reactive to big movements of throttle because it has larger area and is thinner.

(2) The inner stop adjustment. this is preset at the factory (Bosch not VW). It is this way...because every coil armature of every MPS.....even though they are very controlled in manufacture...has slightly differnt characteristics. As the iron armature rod is pushed into the coils (high vacuum/idle)....no matter how high the vacuum signature reference goes....there will come a point where the end of the armature rod is starting to push outside of the area where the maganetic induction fields can react with it and change the reading. This may be only a few thousandths of an inch difference versus inches of mercury from MPS to MPS....but its a lot. So...they set an inner stop for the vacuum range referenced...so teh armature rod cannot be pulled too far into the coil. If it does go to far in....the engine will stall or run very badly at idle vacuum.

(3) The main armature rod adjustment (basic fuel mixture baseline). This has to be adjusted differently for each MPS...as the vacuum cann aneroid chambers inside....with 1960's technology and copper sheet thickness and spring level variances....can only be so consistent from set to set. Since the chamber with its internal reference pressure can only be so consistent...the level of spring and expansion it provides can only be so consistent. Hence....each MPS main armature needs to be adjusted slightly differently to the same reference vacuum signature.


There are a zillion different reactions that can happen within a given engine when everything from wear, outside atmospherics, valve adjustment, timing etc etc etc ....are taken into account.

It is this way with the MPS....all of this adjustability....because the part had to do more...because the brain it was attached to....was capable of doing less.

Modern cars can go with just a simple single range MPS...because they have so many sensors....especially downstream 02 sensors.....to take up the slack.

The D-jet MPS's in general...were only very generally adjusted to cover a wide range of engine characteristics and still run "relatively" normal. If the dealers had been allowed to remove the adjusting plug and put a car of say 40k miles age....onto a basic dyno at the dealer to make a few mid life tweeks.....the reputation of the D-jet system would have been so much better.
Of course....by the time even a well cared for engine started to wear and need MPS adjustment (even if the dealer could have done it).....the weaknesses of the original harness and connector system usually reared its ugly head.....(and these weaknesses were unknown to the world until at least 1975)......it was all but impossible with existing test equopment of the time....to focus on a maladjusted MPS as teh culprit because of the "noise" created by the harness.

I hope this info helps. Sorry for the delay and the length. Ray

As an edit.....the best way to adjust a totally unknown MPS...is weth the induction method. But.....it cannot be a adjusted perfectly by induction. To believe that adjusting an MPS on the bench...to a known factory spec.....and then slap it into a car and have it run perfectly.....is to succumb to the same belief that the factory originally did.....which failed them.

You must first know what range of vauum your engine produces at temperature....at idle...a midrange reference...and at WOT. The idle vacuum is most important. You need to know this to set the inner stop.

The bench setting is MOST important for getting the inner stop set correctly. You should not set the inner stop so close to absolute maximum....that a strange spike in vacuum (which can be caused by a timing change)....causes an extra amount of vacuum....maybe 1-2hg...to draw the armature in far enough that the car cannot idle.

The bench method should get you close enough that the car will run....maybe not well.....but well enough to adjust on the run.

To do this, you need a few instruments. Oil and head temp...and a wideband AF sensor will do.
I pull the CHT wire loose and attach it to a fixed resistor so that I can mimic the hottest average correct CHT...so that short spikes of too lean do not put me outside adjusting parameters.

You drive while watching driving style over a fixed course and watching the AF gauge. In a very short period you can get the main armature adjusted. Then re-connect the CHT and drive for a while. Its at this point that you learn that the CHT sensor should be ballasted to keep from dropping below 125 ohms or so.

The outer full load stop (copper diaphram) can only be adjusted in teh car while driving and running.

If you have lost the outer screw plug...you cannot adjust the full load stop at all. Ray

[JSMskater]

MANY thanks Ray-- thats a lot of info to process but exactly what we were looking for.

I just picked up a box of about 15 MAP sensors. I'll see if any are better than what I've got.
_________________
71 Squareback-FI -- 73 Bay (subaru powered)
TOOB Member #3
I make D-jet FI connectors

[raygreenwood]

One of the many projects that I hope to get back on the rails is the die for the replacement copper diaphrams. I know they work well, but I have not been able to make a permanent die yet. I have to make a new polymer die for each plate I press. I have made about 7 plates and all but two worked and tested out extremely well. Maybe by January I will have a couple to give out for trial.
The delay also was that I have to either give them to people who are crafty with written and photo instructions.....or not at all. Getting the four screw holes and the center hole into teh plate without screwing it up is not for the un-crafty. Getting the new plate onto your old center is not hard...but again....not for the un-crafty. Ray

[KTPhil]

Ray, I am glad to hear that you are still pursuing this!

Broken MPSs are probably the single largest threat to continued use of D-jet VWs. The ECUs don't wear out, wiring can be replaced, temp sensors are available, or even could have modern replacements or fixed resistors inserted. TPS and trigger points are a worry, but seem to have long lives and there are many good and NOS units out there still. But MPSs are hard to come by, hard to test when you DO find them used, and no prospects (other than your efforts) at replacement of the wearout parts.

Are there diaphragm differences in the units over the years? If so, are you working on diaphragms for all years, or certain generations of MPS?

Thanks for your efforts, and PLEASE keep the Samba community apprised of your progress since so many of us are interested in keeping these cars on the road!

[JSMskater]

[Tram]

[KTPhil]

In a few years, we will be scrounging for viable parts, and have to make those we can no longer find. What is unique about the MPS as far as long-term viability is that it really is an analog device that cannot easily be replicated when the diaphragm cracks. Even the TVS can be resurrected with conventional circuit board technology, if one really needed to. It's a digital device that can be remanufactured in a home shop. The MPS windings could be redone by a competent and creative electical shop. But the MPS lives or dies by the peculiar mechanical response of that diaphragm to vacuum. So any variation in thickness, spring rate, etc, will totally mess up the calibration, and it may not be recoverable with the other limited adjustments. So if Ray can duplicate the mechanical properties with his copies, we have duplicated a wearout part. That is why I feel it is so important.

[Tram]

[KTPhil]

No doubt a LOT of MPSs and ECUs were replaced wrongly!

Since they are not refundable after installation, I bet a lot of shops bought and installed them, found the car still didn't work right (or that something simple like a loose wire was really to blame), then fixed the REAL problem elsewhere, and STILL charged the customer for the unnecessary new MPS/ECU. Then to cover their asses, they claim that the "EFI is unreliable," an adjective more appropriately applied to themselves.

[JSMskater]

I really hope that my issue is mis-diagnosis or my ineptitude, because I feel like I've tried every damn thing in the book and every time I drive it it gets worse and worse no matter what I do.
_________________
71 Squareback-FI -- 73 Bay (subaru powered)
TOOB Member #3
I make D-jet FI connectors

[raygreenwood]

Unless you have corrosion or shock damage...its exceedingly rare to find the coils of an MPS going bad. When you get readings of high resistance that fail an MPS....about 90% of the time its high resistance solder joints at the four contacts inside of the MPS. This is very easy to fix....if you are willing to carefully and methodically disassemble the MPS...and then you will have to re tune it upon re assembly. That last part is no big deal because usually its out of tune for your particular engine vacuum signature anyway.
All of this is especially so....if you have changed any major part in teh system...cam...compression...valve sizes....ignition...fuel pressure...TB size etc. Aside from the fact that MPS's were loosely calibrated (compared to how tightly they could have been calibrated to any given engine)....any of these changes means you need to reset the reaction characteristics of the unit if you want to get the unit to do more than just run average well.

Th fix is usually to open of the MPS, record all of the physical starting parts of parts like inner stop, armature and outer plug.....then snip off the old wires to the plug and resolder new low resistance wires.
I found this out by first just reflowing the solder on a few and cleaning up most of the high resistance issue. I say most....because sometimes its just old wires with high internal resistance.

Other issues are the aneroid chambers. I have never seen one fail due to flexing wear. They jsut do not move that much for their thickness and level of flexibility.....but sooner or later they will start failing with age as teh copper gets brittle. The main method of failure is a spot of corrosion or two on a seam that starts it leaking. If the cans don't leak....any changes can be calibrated around.

I have also made new trigger points...one set so far. I stripped a points plate and started from scratch.
The trigger points are a real weak point in getting the system well balanced. You can have wear of rubbing blocks issues.....and high and/or unbalanced resistance from side to side.

Ther biggest problem with D-jet is the harness....and its connector system to be precise. The connector system was actually a defective design as far as fuel injection is concerned. No one ever again used this connector system with primary fuel injection (resistance sensitive) components again once the L-jet style dual cantilevered terminal system came out. All competent fuel injection systems to this day use a variant of the L-jet terminal. I have switched all of my D-jet terminals to L-jet terminals. You get a huge jump in tunability and reliability this way.

Before you start any serious resistance based tuning....sort out the harness. Ray

[KTPhil]

Which connectors are the most problematic? If it is connector resistance that is the problem, then I can see how the temp sensor connectors would be an issue. But how about other connectors to devices that are binary sensing? This would include the TPS, the trigger points, and on the output side, the injectors themselves. Or are those problematic, too? And is the resistance of the wires themselves ever an issue? Some have talked about silver and alternative solder. Is that really important? And how about the board-edge connector at the ECU?

One reason I am asking all these harness/connector questions is that as our cars age, we end up repairing individual wires and connectors that have become brittle. Knowing which are critical would help us know how anal to be about our repairs. Until a complete repro harness is available for less than a few hundies, this is our sandbox to play in.