SVX community

[flomulgator]

So I have a nagging issue with my SVX and need some advice. Everyone here points to this yahoo group:
http://autos.groups.yahoo.com/group/SVXVanagon/

And I'd love to join and bask in the communal knowledge. Only problem is I get shutdown hard with this message:

SVXVanagon: "Sorry, this group is available to members ONLY.
You are not allowed to access this group."

No way to request to join, contact the moderator....nothing. Anyone know what's up or who I should contact?

[wasserbox]

[flomulgator]

Ahhh, good to know!
I am a part of that group but had not seen much info on the EG33 in the SVX folder there.....I guess I'll need to search the general message archives for the knowledge I seek.

Kinda wish things were better organized on either this site or over on the Yahoo group...it takes a lot of searching through large general forums to find anything about the EG33.

[insyncro]

Not too many new converters are using the EG33 any longer.
The EZ engines are newer, available for short money and Subaru still stocks parts for them.

What are your EG issues?
I have completed a few EG conversions.

[flomulgator]

poor idle when hot and occasional associated stall. Found a large exhaust leak at the flex joint between header collectors, so I'm having a exhaust shop look at that today to see if the can replace in place (EGR fitting is really rusty so I don't want to open that can of worms if I can just leave it on). The leak is screwing up the D-side O2 sensor (throwing code).

There are, like, 3 other possible reasons and 2 other codes but that exhaust is the one I'm focusing on now, because fuel mapping gets all weird when an 02 sensor is borked.
Started doing a mid-freeway power-loss/surge yesterday but I'm hoping that's just due to an ECU reset that was performed, as it hadn't done that before. I've heard these ECUs need to relearn after conditions are fiddled with. Will report back with results and beg for help if necessary.

[flomulgator]

Okay looking for a little help now:
The power bucking thing has gotten worse; undriveable now. A mechanic unplugged the ECU while checking another problem, so could it just be the connectors? The other thing I've found on the yahoo group and the SVXworldforums.com site is that the Throttle Position Sensor can have issues. Another SVX owner fixed the problem by replacing a bad coil.

The shop manual isn't much help here. Anyone else ever solve a bucking issue? Or that loping/stalling issue? An SVX shop can't get me in for a few weeks and I want to go on a trip before that so I have to tackle this myself.

[Ryan Keating]

I had the same bucking problem, I had my O2 sensors reversed.

[flomulgator]

The O2 sensors themselves haven't been touched in years. But is it possible to plug them back into the ECU wrong? I don't know the harness config but I do know the ECU was unplugged right before this started.

[Howesight]

Hey Flomulgator:

Yes it's possible to mix up the connectors to the 1,3,5 bank Oxy sensor and the 2,4,6 bank sensor. It will in fact run, but poorly and unpredictably.

But until you have all your exhaust leaks sorted, you will still have issues.

The ECU learning actually happens pretty quickly and relates mostly to fueling and a bit to spark timing. The ECU compares the internal fueling map to the corrections the lambda system is making and if the corrections exceed a given parameter for a given number of engine revolutions, then the map is revised - - a bit. The system does not "forget" the "base map". It goes back to the base map when the ECU is unpowered.

Having now bored you and anyone else reading this, let me pass on the best tip I have seen anywhere regarding SVX vanagon conversion running/driveability problems: Clean up the four main engine connectors.

On most conversions, this is not easy (not on my Syncro, anyway) due to the tight spaces. But after chasing an annoying driveability problem (intermittent reduced power below 2,100 rpm), reading post after post on the old SVX-Vanagon board, and elsewhere, the only post I ever saw that spoke about a CURE was the one giving this tip. It worked for me.

I honestly do not recall what tiny tools I used to do this, but part of the drill was to use 3M scrubbing-type material on a tweezer, and blasting away with compressed air. I am sure there must be something better for cleaning those individual pins and sockets, but I jury-rigged my cleaner.

I think Dylan (InSyncro) mentioned a contact enhancer or protecter in one of his comments. I used to use a contact enhancer called "Stabilant 22" on finicky 1980's Audi electrics. Success was hit-and-miss with that.
_________________
'86 Syncro Westy SVX

[scubabrian]

Do you have a fault code on the ECU? You might check that all the factory terminals are pushed in tight from the backside, I've had that fault cause me some problems with my 2.2 that were intermittent. Over time the corrosion and vibration can cause them to loose contact. I would not automatically assume its conversion related just because its in a converted van especially if the conversion has been running fine since it was completed.
_________________
Brian
85 Westfalia with 2.2L Subaru
brian-mchugh-uwphoto.com

[flomulgator]

Thanks for all the ideas so far!

I forgot to mention that I had the exhaust leak repaired; bucking only continued to worsen after that.

Howesight: I checked the wiring manual and then the actual harness, it seems impossible to switch the O2 Sensor inputs except during initial wiring. Fore future reference they are in ECU connector B59, pins 5 & 6, with shared 17. The pins looked fine and the connector had been previously fully seated. I checked to make sure the muffler shop didn't remove and then accidentally swap the O2 sensors and it doesn't appear so, the patina of rust on one connection is acting like a custody seal. The thing I can't verify off-hand is if there is some connectors between the O2 sensors and the harness.

While I was rooting around I noticed some of the general engine compartment grounds and positive bus stations looked slightly corroded. Not bad enough to say "aha!" but I think I'll give them brass brush anyways. As I noted above the pins for at least B59 looked good but if you think it's worth it I'll clean the whole harness. Worth it? Would MAF cleaner work for this (essentially a contact cleaner).

Are there other connectors/harness areas I should be checking?

Also yet another question: My fan started running all the time. No bueno. Where the hell is the fan switch?

[insyncro]

[flomulgator]

Unfortunately, it has no AC. I'm looking for an engine temp switch/fan control.

[Howesight]

Hi Flomulgator:

I have a harness by Tom Shiels and a SmallCar header. With that combination, the harness allows you to connect the oxy sensors to the wrong location quite easily, which is what happened for me. I recognized the running condition immediately, checked and reversed the connectors at the header end, not the ECU end.

Regarding bad wiring contacts, after posting last night, I checked on line to see if Stabilant 22 was still being sold. It is. Google it. There are other products that pop up when you google it, many with great reviews by techs in the field. One thing to remember in all this is that the newest SVX harness connectors are now 16 years old and the oldest is 21 years old(mine is 19 years old). Not so for the late-model 2.5's everybody uses in four-cylinder swaps. The SVX also has a lot more connections, so more to go wrong!

Also, the SVX coils, (at least from what I have read) can go intermittent (although complete failure is more common), which then causes the lambda system to "correct" for the high oxygen content by increasing fueling for the affected bank of cylinders. EFI relies on no misfires being present, so an intermittent coil can really cause havoc. There's a test procedure for the coils in the factory service manual, but I doubt it would do much to reveal an intermittent coil - - only a failed one would test positive. Do you have the manual?

The reason I have pinpointed the items I have in yesterday's post and today's is that the SVX sensors seem, for the most part, to be pretty robust and the OBD feature helps find those faults fairly well. Sometimes a DTC for a sensor is really caused by a conversion harness problem or a poor connection at a connector.

Your rad fan switch is on the bottom of the rad, driver's side, front of the rad. Remove the connector. If that stops the rad fan from running, the switch is bad and needs replacement. If it doesn't stop the fan from running, then your wiring has been altered and you are on your own in sussing it out.
_________________
'86 Syncro Westy SVX

[flomulgator]

I didn't have much time before sunset last night but I did what I could.

Most importantly, I feel I've eliminated the possibility that the O2 wiring was switched. That was complicated as hell but i got it figured. I also performed some electrical maintenance, cleaning the harness w/ MAF cleaner (it's what I had in the garage) and brush cleaning some chassis ground points that had corroded a bit. Honestly the harness looked pretty good before I touched it, clean pins and connectors, nothing mangled.

Could only test drive a little after dark, still appears to be hesitating but never got fully warmed.

Next up is coil test; I have all SVX manuals so I should be able to figure out that test. Thing I lack is anything related to the van including simple owners manual.

My codes are 23, 32, 37. Of course 32 & 37 are the O2 sensor but it ran fine with the codes earlier, and the MAF also appears to work fine but is still throwing code 23 so rather than throwing expensive parts at it I feel my issues are electric and the codes could just be false flags.

Or the O2 sensors (particularly the downstream) got burned out by the exhaust leak. Is there any way to test the actual sensor for function if I pull it out?

[Howesight]

Hi Flomulgator:

Yes, there is a way to test the Oxy sensor(s) when removed from the vehicle. Here is the method described by TenCentLife:

http://www.thesamba.com/vw/forum/viewtopic.php?p=6142101&highlight=propane+sensor#6142101

In case you don't already know, "TenCentLife" is Chris Corkins, who has already forgotten more than I will ever know about engines and engine control systems, meaning that his information is always 100% reliable.

I can already tell you that there is a strong likelihood that both oxygen sensors are toast. If your exhaust leak was as large as you describe, the lambda circuit was "correcting" for the overly-high perceived oxygen by adding a large amount of fuel to achieve stoich. mixture. That rich running will usually ruin an Oxy sensor fairly quickly. I have used the NTK aftermarket oxy sensors on my SVX with no complaints and less than half the cost of the Subaru stealership price. They use the correct connectors.

Be very careful not to introduce any silicone products into the exhaust system nor anywhere near the oxygen sensor. Silicone products ruin oxygen sensors. This does not mean that cured silicone sealer used in engine assembly will somehow ruin the oxygen sensor, but it should not be used anywhere in the exhaust tract, including the various attachment points/sealing surfaces in the EGR system.

Regarding your code 23, I am willing to bet that contact cleaning, followed by contact enhancer, will clear that code. The ECU self-diagnosis looks for "abnormal voltage" from the AFM, which is exactly what the added resistance from compromised connections can do. The Subaru AFMs ( more correctly, a Mass Air Flow sensor - - who translates from the Japanese FSM?)seldom actually fail - -doing so mostly when some dufus pulls out the screen in the name of "performance".

BTW, whose exhaust system do you use?
_________________
'86 Syncro Westy SVX

[flomulgator]

Great info.

In case I forgot to previously declare, I bought this van recently, used, and all conversion work was done by PO. That being said, the exhaust looks like a Small Car system that has had some later reinforcing added to the weld areas. RMW didn't exist when this conversion was done.

I was looking up O2 sensors in the SVX world car forums and the SVX owners seem to also be now going with the NTK sensors with stable results.
For future reference: http://svxworldforums.com/forum4/2141.html

Given your description of the situation (I know zilch about O2), and in the interest of getting this van up and running sooner (I got a trip next weekend I REALLY want to take this on) I think I'm just going to order the sensors now and then play with the (admittedly cool) bench test later.

Thanks!

[flomulgator]

Update: my O2 sensors finally arrived, but I haven't been able to swap them on yet because the rust welding is SO BAD. Interestingly the SS header is rust free except for the O2 sensor bungs and the connecting ends. I'm working through various techniques, next up is the blow torch. Of course the more important driver's side I can't get a O2 sensor socket around due to clearance

Anyways, way more important is my coils. I went through a lot of effort this morning to pull them all, pull and check the spark plugs, and bench test the ignition coils. Below is a poor cut/paste of an excel spreadsheet I made:

resistance of each ignition coil on shitty DT830 multimeter
specified resistance: 0.68-0.83 ohms "extremely low indicates a short-circuit"
Subaru VW # cylinder? value
DR PF F-644 0.9
DM PM F-645 1.0
DF PR F-644 1.0
PF DR F-614 0.9-0.3 has occasional short at wires where sheathing is stripped.
PM DM F-645 0.9 stripped
PM DF F-6 0.3 stripped

As can clearly be seen, the VW driver's side is totally F'ed. I'll get a picture up in a bit but all three had the wire sheathing pulled back from the coil, exposing copper that was then shorting out. Although it looks like it got yanked I'm pretty sure this happened due to excessive heat because the (rusty-ass, PITA-to-remove-stuck-bolts) muffler sits rather close...close enough that I had to pull it to get proper access. The other reason I can tell it's too hot is that the normally cream colored epoxy backing is dark brown on all three D-side coils, but still cream colored on the P-side.

So I'm going to eventually buy three new coils (ouch...$300) and get some heat wrap on the backside of the muffler to insulate it. For this coming weekend as a temporary fix I'll probably get fancy with heat shrink and electrical tape to isolate the shorts.

Although I still have yet to repair it and verify the fix, I think there's already a lesson learned. Engine is ~202,000 miles on it, and is 21 years old. Conversion puts a lot of heat right next to those coils. I'd suggest others with problems or others doing the conversion or maintenance check those D-side coil wires (where the join the coils) and get some heat protection between the muffler and the electrics.


Plugs look fine, no wetness or carbon. Whiteness supposedly indicates running lean, but not surprising given I had an exhaust leak.

Also I LOVE that I can use the Westy for bench testing in this way!

Also, does anyone know which cylinder is which? between service manuals, the internet, and physical inspection I STILL can't figure it out.

[flomulgator]

Well this thing's F*ed.

Replaced all 3 D-side coils, and it drove better for.....one trip. And that was probably just luck. So basically the bucking/fuel cutoff is still with me, and the local shops don't want to deal with my van, so I'm on my own (or take it to SmallCar).

I'm now considering the VSS, fuel delivery or the ECU, but any ideas are appreciated. Just to rehash my two issues are:

1. Once sufficiently hot, the engine randomly cuts all fuel, then delivers it again. Unless I am careful to kill throttle this gives a dangerously massive "buck". Occasionally, pushing in clutch (changing RPM) will restart fuel. Essentially undrivable.

2. Once warm, the vehicle has a low (2-400 rpm) and unstable idle. Occasionally dies.

These two issues may be linked, but they showed up at different times. Replacing 02 sensors, fixing an exhaust leak, replacing worn D-side coils, heat shielding the muffler and seafoaming the engine have all done....nothing.

HEEEEEEEEELP!

[edgood1]

what year SVX engine is it?

Did you check for vacuum leaks, both on the intake and in/under the manifold?

any codes?
_________________
1987 Syncro Westfalia powered by Subaru
1963 Panel to Deluxe Bus project :::: (photo album)
'65 Westfalia