ALH TDI engine transplant into '84 Vanagon

Andy, you may want to try a rheostat to find the correct value of resistance, depending on the current flow (I doubt that it is very much) you will need to also consider the wattage handling capabilites of the rehostat. Once you fine the best resistance value then focus on a fixed resistor. Radio Shack may have something.

AndyBees said:

So, what Ohms resistance would be best?

Click to expand...

When I set up my friends car I think I used a 5000, but it started kind of "hard", well, it fired up *right now* but those first few fuel injections while your finger was on the button were large injections, the engine had a notable "hard diesel sound" to it. Based on that I felt that something in the 2000-2500 range would be better and probably minimize that hard diesel sound.

Andy, I copied this from one of the hot start threads when I was reading about it. Seemed they had done quite a bit of testing. Your scan gauge should reflect the temp the ECU is seeing when the resistance is added. Keep me posted. One last thought, I don't know if those are standalone values where the resistor is grounded to the the ECU or added to the CTS value. They must be standalone values, because the CTS value would fluctuate based on actual coolant temp. mark


10k ohm = -15 degrees cel. Not necessary, too low and turns on glow cycle
4.7k ohm = 2.7 degrees cel. This is good, only starts short glow cycle
3.28k ohm = ~24 degrees cel. This is probably the best setting IMO. It doesn't cycle the glow plugs, but does modify injection timimg and starts real easy.
1.67k ohm = 56 degrees cel. on the warming up side.
Somewhere between 3.28 and 1.67 is where the ECU will stop considering the car cold and assist starting.

rsxsr said:

10k ohm = -15 degrees cel. Not necessary, too low and turns on glow cycle
4.7k ohm = 2.7 degrees cel. This is good, only starts short glow cycle
3.28k ohm = ~24 degrees cel. This is probably the best setting IMO. It doesn't cycle the glow plugs, but does modify injection timimg and starts real easy.
1.67k ohm = 56 degrees cel. on the warming up side.
Somewhere between 3.28 and 1.67 is where the ECU will stop considering the car cold and assist starting.

Click to expand...

A little of topic, but glow cycle changes significantly with altitude, even on a 90* summer day I get a few seconds of glow up here while at 32* at close to sea level I get only second or so.

Back in the fall on a warm day (50ish out and hot engine), with ScanGauge on, I noticed that the voltage dropped to about 10.8 when I turned on the ignition. And, interestingly, the voltage drop stayed down for several seconds after the GP light went out.

So, either the GPs cycle at a higher temp (both ambient and engine) than what the manuals say, or something wierd is going on with mine. My CTS shows correct engine temp, cold or hot. When the engine is cold, the EGT gauge and ScanGauge show the same temp within a couple of degrees or lesss. Also, out on the road at full operating temp, ScanGauge will show the engine temp to be anywhere from 188 to 192 but mostly right on 190. So, I feel that my CTS is just fine. And, I did install a new one that did not change anything.

But, my "hot start" problem is a fueling issue. Unplugging the Coolant Temp Sensor results in an instant fire-up every time (hot or cold). And, that works whether or not the ECU is programmed for Manaul or Auto tranny.

Well, I've almost got this thing ready for a road test, etc. I'll post results later today/night.

Okay, here is a very early update.

I did a "dry run" here in the garge using the electronics (key, hot-wiring, etc.).

The two Relays work just fine. But, here's the problem that I suspected would pop-up and show it's ugly face. With the coolant at 20.7 Cel (ambient in the building about the same) as shown on VCDS, when I run the current thru #2 Relay with the resistor on the circuit to the ECU, temp reading drops to -4.5 Cel which throws the CEL. I'm using a 4500 resistor.

Keep in mind, if the engine temp was at full operating range (87 cel), this would not be an issue. The probably is at lower temps, the resistor sends the resistance way below the ECU programmed parameter, thus the CEL.

Oh, by the way, the ECU sends current to the CTS on circuit #3 (5 volts). So, I installed the resistor on circuit #4 going back to the ECU. Since it's all about resistance, probably doesn't matter, but worth pointing out.

So, in one sense, I'm back to square #1.

BTW, Mark, those numbers are helpful and tend to be in the ballpark of what I've been seeing as well as what's in the PDF file from a Tech Document that Dan sent me .

So, now, any ideas?

AndyBees said:

the voltage drop stayed down for several seconds after the GP light went out.

Click to expand...

That is normal, the GPs do stay on longer than the dash light.

Okay, a temp of -4.3 Cel shouldn't throw a CEL.

So, I'm thinking the ECU compares temps with something else, such as the Fuel, the Air (in the InterCooler), and/or the Temp gauge side of the CTS.

More thought and ideas.................. of course, if it solves the starting issue, I can alway "clear code" with ScanGauge.

I changed the 4500 Ohms resistor to a 2200. I did a check with VCDS before going for the test drive. At roughly 21 Cel ambient temp and coolant temp, the 2200 Ohms resistor dropped the coolant temp reading to 6.5 Cel which did not throw the CEL. I also programmed the ECU back to a manual transmission.

The test run:

-Engine started fine
-Clutch worked absolutely great
-Trannny shifted great
-Engine seems quiter, especially at idle (remember I did a TB job and found the large roller to be bad,thus the engine rattle) (could have been some noise from the lighter flywheel)
-Speedometer worked fine, then lost power for some reason (probably my cobbled wiring which is mostly temporary for calibarating purposes).

Meanwhile, back at the ranch with hot engine, I shut it down. .............. waited about 10 minutes .......... then, had a rather long crank start, but not as long as previous.

So, maybe I should add another 500 to 1000 Ohms to see what happens.

I do have other ideas..... to be revealed if I try them.

Andy, just curious, where is your starter battery mounted? Stock location or under the back seat? The proper battery for a vanagon diesel starter is quite large and does not fit the stock battery box. The diesel vanagon had the battery in the engine compartment with a short cable to the starter.

I understand the mapping and the rpm, but I wonder if you could get the engine to crank faster if that might help. I did not really have a problem until my battery was on it's way out. There was no indication that the battery was having problems. It started fine the evening before it flat died. I can still sense the hot start delay, but nothing to make you gulp wondering if it is going to start. What starter did you use? Have you seen Karl's from Westy Venture's adapter for using a TDI starter on a vanagon bellhousing? I don't know if that would work with a diesel bellhousing though. I think the market is Subaru conversions. Worth an email. mark

Mark, the battery is under the passenger seat. I have the largest battery that will go in it and the lid still shut. Right now, it's raining hard outside or I'd go look at the CCA, etc.

I am using a Bosch Reman starter as come in the diesel Vanagon. Actually, it spins the engine pretty good on warm days and especially after it is hot. But, this morning at 22F the battery would barely rotate the engine. However, it popped to life in about two to three revolutions.

I have looked at options relating to a closer location for the battery. And, I've also considered purchasing a high torque starter.

Today, pulling an 8 foot trailer, I hauled 1200 pounds of drywall (maybe more since it was the green stuff). I also had on board two 5 gallon buckets of dry wall mud, several bags of pieces parts (nails, screws, etc.), 6 rolls of insulation, full tank of fuel, etc. The engine purred like a kitten up and down the hills.....no problem at all. That jerky clutch that I have not reported here, well, it's now seated and smooth as silk.

Then, we went back out to pick up some groceries (for the work hands tomorrow) (we live 14 miles from town). Well, about 2 miles from home, on the return trip, the engine went into limp mode and defaulted to 1200 rpms. Code P0226 (accelerator) and flashing GP. Wow! I could not have handled that with the load of dry wall.

Anyway, being I was on the rural county road I down shifted all the way to 2nd gear and let it take me home at about 15 mph (up and down hills). Don't you just love the torque of a TDI???

So, I'll have to figure out what popped loose under the back seat where all those circuits come together........

Andy, we had the same problem on our first real trip. It had been raining all morning and the van running on the highway, went into limp mode. We limped to the exit, dragged out the laptop. It ended up being moisture on top of the pedal. I blew it off, and covered it with a glad bag. It was fine ever since. I eventually did tackle the water leaks. Hope it is something simple. mark

Andy.... what size battery cables are you running ?

jimbote said:

Andy.... what size battery cables are you running ?

Click to expand...

The cables are whatever size come as OE on the vehicle.

We're hanging drywall today................ I'm in for short break

This evening I'll do some measuring and determine the size. Also, I do have the complete cable set from that junk yard find. I know current follows the path of least resistance, but I wonder what it would do to install the other cable along side of the existing cable?

Mark, the rain hit as we were coming home from the grocery ...... you may be onto something ....mine doesn't leak up front. But, the "sealing" I did under the seat is not the best....could be there!

Hmmm.... that run is pretty long...not sure what size you need for the diesel starter but I would think the stock size gasser cables are not big enough....I'm considering a battery relocation on my westy to under the seat and I would probably use 2/0 cables minimum ...

Okay, yesterday evening when I got back with the load of drywall (post 971), I absent mindedly, drove past the driveway. So, I turned up a drive-way at a church about 100 yards past where I suppose to turn (about 30 degree incline) and the Van began to spin in the gravels. Due to circumstances, I had to go on. The Van never did stop and just spin in the gravels. So, I let up and it grabbed traction and went on. We unloaded the drywall and headed back to town for groceries.

Well, as we were approaching our drive-way on the return trip from the grocery last night, I heard a clack, knock, thump, etc.

So, this morning I drove the Van to the work house ....more thumping.... At lunch, I brought it back to my garage ...more thumping, knocking, etc.

Well, I ended up having to "come-along" it into the garage.......... Busted left rear outer CV Joint ..............

Wow! Last summer, I cleaned, inspected, greased and installed new boots on all four of the CV joints. I guess the weight and the torque was too much for that joint. Both wheels were spinning because there are two nice long streaks in the gravels showing that both wheels were spinning............

Andy, what happened to fix the "limp mode" problem? I use RedLine CV2 grease for my CVs. Started using it on the racecar years ago. It does not "run" like the regular CV grease will do over time.

I recall the gasser positive battery cable being sort of chinsey. Not sure why I have that memory. markw

Mark, the "limp mode" was gone the next morning. I'm really not sure what happened.

Due to lots of things going on (hint, the drywall), I've not taken the time to look into the problems. The Van is in the garage for now.

I borrowed a friend's Dodge duelly pickup with 5.9 cummins and 5-speed tranny to make three more runs hauling buiding materials, bathroom stuff, supplies, etc., over the weekend........... dang, that thing likes Diesel. But, it can get the job done in short order.

Well, I'm thinking about pulling the Half-shaft and remove the busted CV joint tonight. I need it to be mobile so I can test run while trying to isolate the limp mode issue. I'm thinking it may be a wire connection there under the back seat. As, I stated previously, the front of the Van has never leaked water, so, hopefully the problem is not with the accelerator connection as you suggested earlier.

A bit of an update:

This evening I pulled the left Half Shaft (Axle) to see what the damage looked like. Well, the inside CV joint cage busted to bits. Oddly, that joint shows a lot of wear that wasn't there last summer.

I feel certain the tranny/engine is level and equi-distant (centered) in the mounts. Maybe I just didn't remember. Or, maybe the trip back to town (26 miles round trip) allowed the CV joint to do some grinding.

While it is the garage I'm going to go over the circuits to make sure all connections are okay. Remember the sudden limp mode......idle power brought it home over the last mile or two? (comments on this are in recent previous post(s))

Andy, any updates? I just noticed your build thread was at the bottom and ready to disappear. Bump.

edit, Andy, don't hold me to this, but I recall a thread on thesamba that showed pictures of the input shaft and it is not centered in the bellhousing. I noticed my crank pulley was not centered between the frame rails when I added the Saab mounts. I don't recall the thread, but I know you have some other transmissions on the floor. If the input shaft is off center, then the crank would also need to be off center between the frame rails. Worth looking at. mark

Mark, I have no updates. Nothing has been done to the Van since my last post. I've not re-installed the CV joint/axle either.

I've been out of town for almost a week (Washington, D.C.). I got back a couple of hours ago. I was on-line very little while gone.

I've never measured to see if the engine is centered on the body at the TB end. However, I did center the transmission (axle stubs) in the body. I also made sure it was level (parallel with the floor of the garage) when I had the motor mount brackets fabricated.

Of course, one thing to keep in mind is "torque" of the engine and how it affects the angle of the transmission and CV Joints. Most likely, I may need to install a torque limiter as you did. Also, maybe the left side should be slightly lower (1/4 to 1/2 inch) to compensate for torque as well as a limiter!

Update on CV Joint and Axle

Well, I finally made time to install the CV joint and put the axle back in place.

Gees! Seems there may be more damage. The axle stub on the tranny seems to have about double the 'play' compared to the other side. So, I'm wondering if the flopping of the axle/cv joint has damaged the splines in the axle stub or on the shaft coming off the diff?

Well, I have parts to sift thru for replacement(s). This just means more work that shouldn't have been necessary.

Although those CV Joints appeared to be fine when I had them out for cleaning over a year ago, obviously one of them was less than par. The strain I put them thru shouldn't have ended with these results.

So, I guess I'll be pulling the tranny for a close inspection. And, I'm going to go ahead and install NEW cv joints too.

Andy, before you pull the trans, try swapping the flanges side to side. This should give you an idea if one of the diff stubs is worn. I also can't recommend Red Line CV2 grease for CV joints enough. I have tested it thoroughly on my VW Scirocco road race car. It is really up to the loads and temp. One tube should be enough to do all 4 joints. mark

rsxsr said:

Andy, before you pull the trans, try swapping the flanges side to side. This should give you an idea if one of the diff stubs is worn. I also can't recommend Red Line CV2 grease for CV joints enough. I have tested it thoroughly on my VW Scirocco road race car. It is really up to the loads and temp. One tube should be enough to do all 4 joints. mark

Click to expand...

Mark, I went ahead and removed both axles from the flanges to get a better feel by holding both of them at the same time while giving the "wigglement" test... It does seem that the left one is worn more than the other. Also, the "play" may have been there all along and just never noticed it.

There seems to be no grinding, etc.

Also, I have four 091 trannys in my pile. I checked all of them. Seems no two of them are alike. I may pull the flanges off one of those to see what the difference in play might be.

So, while the tranny oil is out, I guess it would be best to go ahead and change it. There was the typical fuzzy amount of metal on the oil drain plug.....nothing unsual!

Okay, I pulled an axle flange from one of the diesel trannys (most likely lower mileage). It was really loaded with crud. Anyway, after a good cleaning, I rolled under the Van with it in hand..... because it fit so tight, I had to hammer it on!

Wow! It's amazing how tight it fits.

So, I also think I found why the other flange(s) is worn. There is a wave washer (concave) that goes in after the flange and then a snap ring. Well, on my tranny, the wave washer was turned the wrong way. Thus, it did not apply pressure on the flange which "may" have allowed it to wobble under the weight of the axle. That same wave washer on the diesel tranny was turned the other way. Seems there's no mention in the Bentley as to which direction to place the wave washer. However, considering it is a wave washer, logic would/should prevail ....Duh!

Well, I'll finish the job tomorrow.

As mentioned in a previous post, the Van went into limp mode and defaulted to 1200 rpms about 1 mile from my house a few weeks ago. Well, it still wants to idle at about 1200 but does accept accelerator pedal..... any ideas? No CEL. However, I have not ran VCDS yet.

1200 rpm usually means it lost the signal from the foot pedal.

If one of the wires from the foot pedal to the ECU is grounding out it will idle at 1200 also.

Andy, 1200 rpm seems to high regardless. When I first got ours going, I increased the idle to about 920 rpm using adaption and VCDS. Seems stock it idled high 800s. Anyhow as suggested, while the pedal does seem to work, your high idle would indicate a problem as far as the ECU is concerned. Perhaps as you said, drag out the laptop, clear any codes and maybe try unplugging the accel pedal to throw a code.

Also, a little vague on this at the moment, but the ECU on the ALH needs to relearn, "Readiness" I think it is called, Bentley and the Vag Com software address this. And one last thing, if you have been going back and forth between manual and automatic, this may also be affecting the ECU somehow. A torque convertor car would expect a different load at idle vs a manual. mark

Update

Okay, a bit of an update on a number of things.

Mechanical
First, I installed a used CV joint and an axle flange from a low mileage diesel transmission as I mentioned in the last post. It fit really tight and actually had to be pounded on. That tightened up the assembly. The wave washer and clip fit just fine....pretty easy to install.

Anyway, I went ahead and ordered four new Lobro brand German made CV Joints. For the time being, I put the used tranny oil back in which only has about 3500 miles on it. So, when the new joints come, I will install the other diesel tranny flange and change the oil at that time. Even though my tranny was very clean, the oil came out somewhat dark but with a green hue to it. I guess the new 3rd and 4th gears, as well as new bearings, are broke-in by now, thus, explains the discoloration of the oil. I have new GL-5 ready to install.

Electrical (note: the pics below are from gallery files, nothing new)
I've also ordered a new 12 pin connector to replace the "cobbled" mess I did trying to use a connector from the wiring harness (modified FCM assembly and 14-pin connector).

The pic below shows how I used the 14-pin side of a Fan Control Module as a connector. However, I did not use the CAT wire as shown in this pic. But, there are 14 wires soldered to the connector that go to the Blue and Black OEM connectors that are typically located on the Relay panel of the Jetta. The white plastic receiptable side is on the left in pic. You can see in the next pic where the black 14-pin connector plugs into it.



The pic below shows the other side where the 14-pin connector actually plugs into the modified FCM housing. All those wiring splices are okay.



In the pic below, top left, you can see how it looks now.



The pic below shows how the wires come out of the 14-pin side of the modified FCM assembly. All those circuits go to the Blue and Black connectors for the accelerator and cruise control headed to the ECU.



Anyway, those circuits involve the accelerator and the cruise control as mentioned above. I have all but ruled out any problems up front with the splices there. So, the issue of defaulting to 1200 RPMs at idle must be in this cobbled up connector.

I can use a 12-pin connector and add a 2-pin to cover those connections.

One other thing I have never experienced previously. Last night and today, I was out test driving and the BRAKE light began to flash but the CEL never came on. After about 15 seconds it quit flashing. I ran VCDS and the only thing that showed up was related to the accelerator which is probably those connections mentioned above. It's possible the wire going to the cluster for the emergency brake grounded....... just not sure!

EDIT: I could splice the wires directly to the wires coming off the Blue and Black connectors. However, those connectors have at least one wire each that goes somewhere else. So, in order to keep it as much 'plug & play' as possible, I need the 14-pin connector (soon to be a 12-pin and 2-pin) ............... now, it's clear as mud, right!

Andy, I am using the vanagon cluster, but in the van's life with TDI power, I recall a couple times where the glow plug light would flash. Both times, it was related to my brake pedal combo switch. Seems there are two degrees of faults. Maybe you are expriencing something similar with your cluster. For connecting harnesses, I like to use weather pack connectors. I have used up to T6 sizes before. They are snug and fairly weather tight. mark