Digijet Idle speed too high when warm

[jimrat]

Hi

For the last few months I've noticed the engine would stall unless it was held at a fast idle for 30 seconds. Now it has got frosty, the engine needs 2 minutes to warm up, so I decided to investigate.

Last night I found that the throttle switch underneath the butterfly was out of adjustment: -the switch wasn't clicking when coming off idle. So I removed the throttle body, cleaned it all up inside and out, and adjusted the switch position. All good, and the engine now holds a steady idle from cold.

The problem now is that when warm, the idle is too high. There's no rev counter, but I'm guessing 1500 rpm+. What should I be adjusting to bring the idle speed down when warm? The big idle screw on the throttle body looks like it is wound right in.

I've checked for air leaks and all is OK. The throttle didnt have much play in the shaft and the butterfly is nice and snug in the bore and moves freely. I haven't adjusted the base throttle stop (but that doesn't mean the PO didnt - the CO mixture screw has been fiddled with by someone ).

Should I be looking at the temp probes? AFM? Anything else?

Cheers
Jim

[AdrianC]

If the idle valve's not been doing much for a while, I'd check it's moving freely - can you feel it buzzing? What happens if it's disconnected? Give it a good clean internally with carb cleaner, too.

[jimrat]

Hi Adrian

Thanks for the reply. Sorry I should have mentioned in my first post that I had checked over the Idle Stabilisation Valve and it is all working as it should: coil resistance is within spec, moves crisply when 12v on the terminals, buzzes when the ignition is on, and the engine runs carp if it is disconnected

Cheers
Jim

[ajsimmo]

Do you have the correct dual vac dizzy (adv and retard working correctly), or has PO put a single vac on? The retard is vital to bring idle speed down to within spec. Failing that, next place to check is temp II sender for higher than expected resistance, either from sender itself or corroded connection in the circuit.

[jimrat]

Spent some time today investigating

Re-checked the adjustment of the throttle switch - all OK: continuity (zero ohms) at idle and full throttle; open circuit at part throttle
Temp 2 sender resistance is within spec when hot/cold
Noticed that the previous owner had bypassed the idle stabiliser (the box next to hall generator) by joining the two connectors together.
So I reconnected the idle stabiliser, but it made no difference to the idle - still idling high 1500rpm + when warm
So I then did all the pin checks on the ISCU (behind O/S rear lights)

All ok EXCEPT:
At idle I get 0V between pins 5 and 8.
At part throttle I get 5V between pins 5 and 8.

This is the opposite of the Wiki on here (2.1 DJ bad idle), which says I should get 5V at idle. Posts on thesamba.com say '5V present when throttle switch closed with engine running'

How can that be Any suggestions. Is the signal at pin 5 an input to the ISCU or an output?

Thanks
Jim

[ajsimmo]

But does the vac retard work?

[jimrat]

Timing light is on my xmas list!

All the vac hoses are in good condition. Will have a closer look today.

[jpennington]

It might be a faulty ISCU. A few years ago, I experienced a similar problem with a high idle speed. It gradually got worse until it was around 2000 rpm (I do have a rev counter).
A power transistor in the ISCU had failed (apparently they are prone to overheating). I replaced the transistor and all has been well since.
I found the relevant info on a US vanagon site. I dont have the link now, but here is the text I saved.

Vanagons with the Digifant fuel injection system can develop a problem whereby the motor does not idle properly. In my case, the symptom was a high idle that could not be adjusted with the idle screw on the Digifant air filter box. After conducting the standard diagnostics according to your repair manual, if you find the idle stabilization control unit to be faulty, here is a cheap and easy repair.

Remove the right tail light by removing the 4 Phillips-head screws. Reach into the back of this compartment and lift upward on the small black box. This is the idle stabilization control unit. The wiring plug slips downward and hooks onto the car's body, and the upward motion should free it. The wires are long enough for you to pull the control unit, plug, and wires into the open. Remove the plug from the control unit. It is tight, and will require a bit of force to remove. Now you must remove the black case of the idle control unit. This is not easy since there is a lip on each of the four sides that you must simultaneously lift in order to pull the case off.

Once you have the case removed, you will see that the idle control unit is composed of two circuit boards. One of this boards hinges to give access to the other board. The culprit, in my case, was on the stationary board. It is a rectangular transistor with a hole in the middle, the BD438 PNP Power transistor. The board, and transistor can be seen in the figure shown below. As far as I can figure out, it is part of an amplification circuit that provides current to the idle stabilization valve (the 450 mA of current that you measure in one of the diagnostic tests) through the connectors labeled ST1 and ST2. This transistor should be heat-sinked (hence the hole in the middle), but it isn't. Apparently, it gets too warm and fails.

A photo of the idle control opened up and identifying the replacement transister is on the Westy Site Tech Drawings link. See "IdleContFig" in Tech Diagrams folder. The red arrow points at the BD438 transistor.

To change the transistor, note its orientation on the board (make a sketch), and carefully clip it off with a pair of wire cutters. In using a solder iron, be careful not to overheat the circuit board. Warm up the left over legs of the transistor on the underside of the board one at a time, and push them out of the board. Carefully remove the excess solder with a desoldering tool, and clean the holes so the new transistor will fit in. You will have to bend the legs for the correct positioning. Do not clip the legs till after you have soldered the transistor in place. Now solder the transistor being careful not to overheat it.

Note that if the idle stabilization valve has an internal short, it will draw more current than it should, causing the transistor to burn out again. If this happens, you may need to replace the whole unit. This repair worked fine for me, and it was cheaper than the $250 a new idle stabilization control unit typically costs.

Good Luck!!

John Neumeier
'91 Westfalia and '87 Syncro

Required parts: 1 BD438 PNP Power Transistor

ISCUboard.jpg

Also, re your tests on the ISCU; according to my Haynes wiring diagram, pin 5 goes to ground and pin 8 to the throttle position switch. So I would expect 0v with TPS closed (idle) and something elsewhen TPS is open - consistent with your measurements.

[AdrianC]

I found the relevant info on a US vanagon site. I dont have the link now, but here is the text I saved.

Vanagons with the Digifant fuel injection system

Whilst I'm sure that the concepts and quite probably symptoms apply, it's worth remembering that DJs don't use Digifant, but Digijet. Digifant was on US 2.1s, whilst they used Digijet on 1.9s (albeit with a cat/lambda). So, if you're translating US fault-finding and diagnosis to DJs, look for 1.9 not 2.1.

[jpennington]

AdrianC is partially correct.
European DJs use Digijet, but they also use an Idle Stabilizer Control Unit (ISCU) and Idle stabilization valve on top of the motor just like the US Digifant models.
The US Digijet systems do not have an idle stabilisation valve.

So you need to be careful when using information from US sources. The specific detail in my previous post is correct, at least for my 1986 DJ - purchased new in Holland.

[jimrat]

Thanks for the replies

The vac advance seems to be working when I suck on the vac pipe. Will confirm timing when strobe arrives. Thanks for the excellent info on iscu - will also look into that

Cheers Jim

[ajsimmo]

The vac advance seems to be working when I suck on the vac pipe.

The use of the singular suggests there is only advance, not retard i.e. it has the wrong dizzy on for a dj, and will tickover too fast as it's too advanced.
If you meant "pipes" and you've tried both then check throttle body base setting. The vac port on this is, well, ported so there should be no vac at idle. It overcomes the retard signal when the throttle butterfly starts to open.

[jimrat]

Hi ajsimmo

The correct dizzy is fitted - part # 025905205N - and has 2 vac pipes connected: one to throttle body and one to inlet manifold

Cheers
Jim

[jimrat]

Update:

Bought myself a timing light and fitted an aftermarket rev counter. I checked the timing as described in Haynes / wiki, disconnecting both vac hoses and idle circuit wiring.

The static idle for my engine (2.1 DJ running 98 Ron) should be 10 deg BTDC @ 800 rpm. I cannot get the engine to idle below 1000 rpm; to get 1000 rpm the static advance is nearer 18 deg BTDC and it idles very badly. At 10 deg BTDC it holds a steady 1250 rpm. I have wound the big idle speed screw completely shut, and even fiddled with the idle stop screw, but it made no difference to the idle speed.

I checked the centrifugal advance and it all seemed to be working OK - max total advance of 26 deg @ 3500 rpm
With the vac advance hose connected I get an instant 10 deg advance when the throttle opens, up to a max total advance of 40 deg @ 3400 rpm.
With the vac retard connected, the timing is retarded by 10 deg at 1250 rpm (to TDC)

Any suggestions as to why the idle is high? Worn throttle? All seemed tight when I looked a few weeks ago. Running out of ideas now
Cheers
Jim




Just to recap

[AdrianC]

Any suggestions as to why the idle is high?

The idle circuit IS all kicking in, right? TempII sensor telling the truth?