Progress on the Nomad...

[MGVWfan]

Good discussion. I like this sort of thing, it makes me think (my brain hurts!)

Most everyone other than a few GM cars, and most European makes seem to have implemented some version of a "modern" D-Jet (engine RPM and vacuum sensing only, no air flow/mass sensing). I know Chrysler has used a MAP sensor type system since 1984, and has never used air flow/mass sensing systems similar to L-or LH-Jet. In all of these systems, there's a digital equivalent of the good old analog WFM, except it's a set of numbers in memory that map RPM (and manifold pressure, too) to injector "on" time (and modify it with O2 sensor inputs), and those numbers are pretty important in getting the system to run right. When a tuner "tweaks" a computer, he's changing those numbers in memory (the three dimensional space you were referring to Ray). I think what you're saying is that the vacuum part of the equation is predominant, and affects the final injector "on" time more than the "RPM" part. You may be right.

Funny you mention using a MAP sensor. One of the hats I wear at work is "Instrumentation", and what you need to get the relatively linear MAP sensor response to be useful somehow with what the ECU is expecting from the MPS (including full-load compensation) is a little instrumentation trick called a Calibration Curve (Cal Curve). The other needed bit is a way to emulate the transformer and inductor action of the two MPS coils coupled together by the manifold-pressure-variable core. It can be done. We've got vastly more complex cal curves than the MPS one in use right now in the late 70's technology of the Shuttle's inst system, and implementing a cal curve in analog technology, while it can be tricky, is something my world does all the time. The real problem in trying to get a D-Jet MCU to think a modern MAP sensor is really the MPS is figuring out how to get the sawtooth pulse generated by the MCU (modified by the head temp sensor) back into the MCU with the manifold pressure variations correctly superimposed, as is done by the MPS. I'm working on a saturable core three-winding reactor that uses op amp technology (implementing the cal curve) to change the mutual inductance of the core of the two-winding transformer based on a solid-state MAP sensor's inputs. I've got the 30,000 foot-level design drawn out, but I've put off the details until I get the Nomad back on the road (with a NOS MPS). Nerdy or what?

As for the WFM compensation...it appears to me one could do lots of magic by changing some resistors and capacitors in one or more WFM's (just like the tuners changing memory numbers on "modern" systems)...but it only affects the RPM-vs-mixture part of the MCU's reponse, not the load (vacuum) dependant part. And you're right, the "load" part of the MCU's function does appear to be predominant in the final calculation of injector "on" time. I guess the real question is, how much to alter the WFM "cal curve" (that's really what the WFM's do, create a "cal curve" that changes the mixture based only on engine RPM), and if it'll do enough to be worth the effort. I'm planning on doing some bench experiments to see how much component changes in the WFM's change the injector "on" time with RPM input changes. Who knows, I might end up with the solution to the mid-range issue, and a solid-state replacement for the mechanical MPS

And, a 1/4 Watt 10K resistor is a 1/4 Watt, 10K resistor whether it's German or Taiwanese, so unsoldering and replacing a few resistors in the MCU is not an unrecoverable situation if I damage one of the originals (gotta love standardization of electronic components!). It couldn't hoit!

[MGVWfan]

Today's progress...due to the low humidity conditions, painting was in order. Today was cleaning, de-rusting, painting, and baking the intake manifolds and air distributor. Doesn't sound like much, but the rats had left their...mark, shall we say, on everything topside of the engine, and the cow's udder (what I took to calling the air distributor) was especially rusty and nasty. The udder took 4 hours by itself. Wire wheel on the grinder, and rotary cup brushes on the drill press, along with sanding and scraping. Damn, wish I had a media blaster and 5 hp compressor to run it

[raygreenwood]

Dang!...sorry..I somehow missedyour return post the other day on our EC/MPS discussion. Yep...you exactly got what I was trying to say (nice translation skills). The rpm vs enrichment rise...appears to "act" like a preset linear curve. It appears that the lead/predominant enrichment and variation mechanism is in the MpS load sensing armature. It has lot of adjusting that can be done...but some limited vacuum response ranges. BUt...there are also some out and out modifcations (and then subsequent recalibrations) that can be done.....I will yak about those only privately . Question.....is there anything that can be done....to augment the output of the MpS....by simply adding circuitry into the harness between the MPS and ECU? Just wondering. Thats one of the few things I have not played with. Also...you should check out Brad Anders D-jet web-site. He has spent years inside of the ECU.
I am starting on your centerlink in the morning. Unwrapped it tonight. Deeeeewd...I bet your steering was floppy! I know mi ne was last time I had one like that. Ray

[MGVWfan]

Centerlink...I see you've met my leetle friend, Mr. Floppy Leenk
Yep, it had about 3" play at the sterring wheel rim, and most of that was the centerlink pin moving about. Hope you can do something with it. It had me looking for a new one already, and not finding

On Brad Anders site...already mined for gems (and it has a whole motherlode full of good D-Jet stuff )...and no, unfortunately, with the shape of the waveforms involved, adding a passive component (resistance, inductance, capacitance) probably wouldn't give a predictable result. Adding some sort of active circuit that changes the height of the sawtooth waveform might, but simulating the MPS coils and the variable transformer action driven by a solid state (easily obtainable) MAP sensor should be much better, and end up using the same circuitry, anyway...it's a matter of finding a way to do it. I've got a thought, but it'll require getting my 412 running well enough to use it as a test mule, so first get the Nomad going, then play with the FI toys

BTW, I used some of the data from Brad's site to confirm your suspicion. The total variation in the injector "on" times from RPM alone (Wfm Generator effect) is about 1.5 milliseconds, and the total change from idle to WOT vacuum from MPS inputs alone is about 5 milliseconds, almost 4 times the effect. Therefore, messing with MPS calibrations can have a greater total effect than messing with Wfm generator constants. However, the MPS does not allow shaping the response versus RPM, while messing with the Wfm Gen constants does, and that's exactly what you need to do when changing intake or exhaust restriction values (more fuel needed at higher RPM's with better flow than before, I think), or CR changes (lower CR = ? change to fuel requirements, but still RPM dependant, I think). With my Heathkit CO Meter to help me out, I think I can map out exhaust CO vs RPM and see what I need to change related to flow improvements, displacement increase, and CR decrease without having to load the engine down on a chassis dyno (since the load-dependant part of the response is very small when the engine has only the friction and windage losses to drive while revving in neutral/park)...BTW, any problem with revving a 003 VW tranny in neutral/park? I understand certain BMW's have serious problems with that sort of thing, it makes them fly apart, no personal experience with this tranny to know.

[ubercrap]

It had me looking for a new one already, and not finding

Ask and ye shall receive...

http://cgi.ebay.com/ebaymotors/ws/eBayI ... egory=6763

[MGVWfan]

Thanks herr Uber! It WILL be mine, so I have backup steering parts for contingencies

Does this make me a card-carrying member of the type 4 Parts Whore Club?

[ubercrap]

Thanks herr Uber! It WILL be mine, so I have backup steering parts for contingencies

Does this make me a card-carrying member of the type 4 Parts Whore Club?

Mmmmm....Maybe, a few more duplicate parts and we'll see...official status is bestowed by Ray, though, I think!

[raygreenwood]

No....No....I am just a slut!. I used to be a whore, but now that I only have one car...and lots of parts, I have been demoted. Dr crap (uh...uber i mean)....we will simply remove your legs to make "access" to even more car bodies...easier.

MGVW....yeees....I met your leeeetle floppy friend (STOP that UBER....I'm talking about the centerlink and nothing else!)!

I took pictures while disceting it. The nylon parts had turned totally to powder! . I showed side by side, the pin assembly from yours and one with the nylon parts still intact from another unit I have (hhhe hhhhe snort...he said "unit").
By the way...this is from a very late model 412...no? There are at least two different variations, perhaps more. Yours has machined in grooves to allow the use of a standard tie-rod end, outer boot. Thats nice. I will set up to cast you a silicone pair of boots. The joint head itself is slightly smaller, and has a different material on the central bearing ring. You wil see from teh photos on teh disc I send you.
The earlier model like I have, uses a urtheane cup boot that slips over the top form the outside. Rebuilding both of theses is approximately the same. What this mainlt entails, is replacing all of the nylon packings...with hand lapped, tight fitting bronze thrust washer, each grooved so grease can flow from end to end. On teh pin end...I will press on a thick, flat bushing to lock the whole assembly upright in the rod cup. On teh back end...it finishes with another pair of shims on the step of teh tapered pin end. That goes against a thick, hand lapped forged washer...and the whole mess is held in by a snap ring...set into the very same swaged groove that the cap on the back was in . Everything is hand lapped for a very tight fit. Then seal it with a smear of silicone on the back...and itwill last forever and be at least ten times stronger than stock...and tighter...since al of that nylon is being replaced with bronze. Ray

[raygreenwood]

Oh...almost forgot to mention. Since the rpm signal in D-jet has a lot to do with the injection window...mainly because its timed from the dizzy rotation, I have found that making a modification to the distributor and adding extra triggerswill allow making a sequential set-up...even with the same two original injection chnannels.
One of the very best ways to augment fuel enrichment for larger engines..especially when all else is the same, including cam and stroke....is simply to bring fuel pressure up across the board.

The real trick on engines of a different configuration , is to make fuel pressure adjustable at the whim of the MpS...and not general manifold vacuum. I have some ideas on that that just bolt on. But the main and overriding problem, is that the MPS...in its stock tuned form...has some defficiencies in the part throttle midrange. Its very sensitive to over-enriching at low rpm and not having enough range at high rpm. The high rpm is easy....simply by adding about 2-4 psi to fuel pressure over a certain rpm. The other part throttle issues, require: (a) a mainspring recalibration of the MPS. (b) an added inner and outer stop change to both keep the augmented spring pressure for keeping the idle enrichment level too lean ...meaning the rod will be too far out of the armarure at max vacuum. (c) The addition of a stroke stop for the armature rod to keep if from passing out of influential range....and also to keep its throw from potentially damaging the copper plate and the aneroid system.
Simply put....what is needed.....is an additional amount of force to be added to the mainspring to make it harder for vacuum in the part throttle range to return the armature to a less rich range. In other words, the armature needs more hang time...if you get my drift. What this does though...is make the armature where it will not protrude as far into the armature at max vacuum when the aneroids are most extended, becasue there is now too much spring pressure. What to do?....and this is the first time I have mentioned this publicly.....First map the armatures minimum and maximum range within the coil. Then...remove the pin that runs through the cylinder of the armature. You will find....that the outside is steel....and the inside is aluminum. The metallic mass the coils are reading...is the outside. That means...You candrill the center, thread the pin...and make the position the armature is in on the pin....variable. Next...you need to install a stop pin at the max stroke area...to keep the armature from moving any closer to the aneroids than is useful.
The spring tension is tguned by cutting a slight in the horn at the back end of the horn unit. From there...you canslip shims under teh springs. You seal it, by slipping a vacuum cap and clamp over the end.
You now have three more sperate adjustments to the MPS...that make the part throttle settings more adjustable.

This makes the "tool"....the MPS....that measures the load on the engine....more accurate...and sensitive. Only After doing this.....and then adding a way to augment fuel pressure in one specific rpm related range....like with with a frequency valve injector....would I then change the wave pattern if more was needed. Ray

[MGVWfan]

Interesting thoughts on modding the MPS. Sounds quite doable to me, having disassembled one of those MPS thingys myself

All the mods you describe to the MPS can be done in analog electronics, with a solid state MAP sensor (once the biggest hurdle, emulation of the MPS transformer action, is overcome). It's like the "old days" of mechanical system simulation...engineers used analog computers to simulate mass-spring-damper systems, now they do it in a digital application on a PC. But the analog computer still works for solving those problems, even to this day...same thing with the D-Jet MCU and electronic mods. The MCU is basically an analog computer, and modifying it requires adding analog circuitry, or modifying its guts with analog (passive) devices, even though nowadays FI computers are digital, and recalibration is done digitally. It can be done.

Adding fuel at any specific RPM is pretty easy, you add a pot in the approriate Wfm Gen and you adjust it to make Tinj go up or down at the RPM range affected by the Wfm gen in question. It's also easy to add Wfn gens (as Bosch did) to re-shape the curve, or change a capacitor if that gets you where you need to go. The only limitation is the maximum injector "on" time allowed at max engine speed and load. As you point out, you can increase fuel pressure to fix that up to a limit, or increase the opening on the injector itself ("bigger" injectors). It is possible to rework the MCU to do what you need, especially within the limits of the T4 engine's capabilities (normally aspirated...turbos and blowers are probably more than can be compensated for with the MCU's circuitry).

Once the Noamd's back on the road, I'll mess around with it a bit and let you guys know what comes of it.

Powder, eh? Doesn't surprise me. That was one floppy dingus (calm yourselves !) The link came off an August 1973 production 412, so it wasn't the last year of production. And I know it was never replaced, the original owner and PO kept all bills, and there was no front end work in there, other than a tie rod end.

[ubercrap]

The center link rebuild sounds kickass Ray, makes stock sound like a pile of puke. How much ya be chargin' for that thar rebuild?

[raygreenwood]

"0".....I need a few parts out there for samples . My brother had the ultimate suggestion though the other day. He looked at the centerlink while I was bead blasting it....and said....why don't you just make a new one. He proceeded to show me how the racers in the other part of the shop make theirs. A piece of 1" bar stock, thread both ends, drill the two tapered bores, simply order the two taper pins and two thread on taper pin sockets. Neat....but...its about $100 worth of parts, and unlike other cars, this centerlink exists in a very narrow space. The ends agle up, and I would have to find a way to angle up the ends or else scrape on the sub frame. But....I'm thinking of ways...... Ray

[MGVWfan]

Oooh, I like that thought...totally new and rebuildable! $100 isn't really that bad considering TRW's part was in that range, and had the nylon bushing problem same as the original part

What I've wondered is, why did VW use nylon, why not use bronze from the beginning? Is nylon really that much better in this application with a "sealed for life" part? I think not. Cost maybe? Probably...

[raygreenwood]

Well...$100 does not sound bad...but there is a bit of machining. If I start to build these things...they could get more expensive. But...since most of the centerlinks are going for between $125-179 these days...and are defective like we have seen...its not too bad.

Just like anything else, these parts were probably designed to have a life...and simply be replaced, but also....it may be that the auto industry did not really know enough about nylon and its reaction to water in those days.

Its a strange part. I had type 4's for years...all with sh*tty steering. Replace everything.....but the centerlink. Why? becasue unless its as bad as the one you sent me....you cannot tell by wiggling or prying or anything....unless you remove it. If its not floppy, it does not look like its bad..so you never remove one.
On the car I have...6 years ago, I took the whole front suspension off, just to get everything clean. The centerlink looked tight...on the car. When I got it off....there was movement.
The reason you cannot see it, is that there is just enough nylon still crushed between meal, when its installed, that the movement can't be "felt" by hand. But put some force on it..like driving...and its really soupy. Changing the centerlink is an almost 100% improvement in driving.

I did not know about the defects, until three years ago. I had a NEW one...with 35K on it. It got floppy. I took it out....and had no more good parts available....at least close by. I was at work on a saturday wih access to the maintenance shop. The joints had vertical movement (axial). So I took it out and ground the swaged rim off the cap on one end. In retrospect....i figured out that I didn't need to do that. In fact, I use the swaged rim itself....as the groove for the snapring.
Anyway, I took it apart...looked at the misery inside. It was immediately obvious that I could replicate the nylon with bronze thrust shims.
Since I had ground the cap structure off....I thought...what to do? Well for the moment...since its clamped from the other side (the pin..clamped with the castle nut)....there is no way to pull the pin through the housing. Theoretically...you could run it with an open back.....and I did.....for 6 months with no adverse affects. It was bushed tight....no problem. In the meantime I disected another more carefully...and learned the snap ring cover method.
The only thing that will need to be done when you get your link....is the actual filing of the outer bushing. Since every arm you pres these into is different...just slightly..and it needs to be a tight dead fit...no slack...no interference, either the outer bushing will fit perfect...and the tapered pin will lock into the pitman and idler arms...and the sterring will turn...or it will be too tight. If so...just file off a bout .003 0n the outer bushing. Its simply a 10mm bronze tube that slips around the taper on the outside of the joint. Or I may use thrust washers there. I will decide tommorrow. The best thing actually...is to install the link without the outer bushings, and measure the gap between the top of the joint on each end of the link, and the face of the arm the taper pins go into, and then adjust the tube or washer stack before you start. Should be able to just do it with a file. Ray

[MGVWfan]

And since the idler arm and steering box are out, I can bolt them up and measure, then pull them apart and adjust the clearance if needed, all before putting the whole mess back up in the car. Good deal Herr Doktor!

I keep getting pulled into office work after normal hours, hopefully I can get the idler bushing ordered and replaced next week, and order two more tie rod ends to match the two I've already got, and that's the steering part of the front end done. The suspension part...still need to order Rabbit control arm bushings and fab up a set of centering rings, and pull the old front strut cartridges out and replace them. Oh, yeah, I need to modify the ball joints to put grease fittings on them, and check for play (didn't appear to be any when last I checked, but I didn't do the full-up levering the joint test, just the "shake the wheel" test). I've done three sets of MacPherson struts to date, so spring compressing is old hat (but always induces pucker, all that energy, one slip and the spring bounds out and cuts off something vital ).