CHT balance - ljet

[type11969]

Greetings. Does anyone have cylinder by cylinder CHT info for stock t4 cooling systems - specifically fuel injected, more specifically 2.0L ljet in a bus?

Does cylinder #3 run hotter on t4s like t1s?

I am seeing a 30degF difference between cylinder 1 and cylinder 3 in my 76 westy running a camper special with backdated exhaust. 390degF on cylinder 3 when lightly loaded at 60-65mph (too hot) and 360degF on cylinder 1 (much better). I'm curious if this discrepancy is expected with stock cooling systems or if something else is going on. I've confirmed that the flaps are working, all of the tin is in place and the engine seal looks good. I did notice that the pipe running up to the flapper valve on the pass side separated from the heater box - I need to correct/test - but in the meantime, if anyone has cylinder by cylinder test data for stock cooling systems, I'd be interested to hear it.

-Chris

[Piledriver]

Sounds about right, #3 a little higher than usual.
VW mostly fixed this with the Vanagon fan shroud, also injector balance, crud in the air channels in the heads, aluminum flash or otherwise, exhaust system/hx issues are all possible contributors.

IIRC Jake once posted data on stock vs his dtm setup here before he started his own forum.

If the fresh air to the hx is restricted or disconnected it can make that head run hotter.
Sounds like the "out" side in your case so should not be an issue, unless the hx are full of mud and don't flow air.

One advantage of MS3 is you can tweak fuel and spark for each hole.
May also be true for Speeduino systems, and fuel only for ms2E/sequential rigs.

[type11969]

I just had this engine apart (head rebuild and re-ring after 25k miles), so unless I left a towel in there, I don't think this is a crud issue . I had it apart for low compression issues that ended up being attributed to overheat heads, hammered seats, etc. I didn't catch it because I hooked up the TC wires for my DD gauge wrong, creating additional offsetting junctions (expensive mistake). I've since backdated my exhaust - likely some of the root cause of my elevated engine temps - and I'm trying to get a better idea about actual engine temps by monitoring more than one CHT. I was surprised to see this offset. AF readings with a wbo2 sensor put me in the 14-15 range during light loading, 12.5-13.0 full throttle, which I believe is okay (throttle response is good). Full advance is in the 28-32 range, and running with the vac advance disconnected results in the same CHTs, so that does not seem to be a factor. Interestingly enough, the #1 plug, when pulled, looks a bit rich (black/sooty) whereas #3 looks pretty good (tan). Perhaps I need to run a compression test again, but when I checked post rebuild I was in the 120 psi range all around when cold so I don't think that is an issue.

I need to take a look at #4 and #2 to see how they compare. I also need to swap CHT sensors to make sure the offset doesn't follow the sensor (the ol' hot water test shows that they are the same, but physically they are a bit different so I don't know if cooling air is a factor).

Considering that ljet is batch fire FI (right?), I can't explain why the plugs would look this different. But I think there is a separate driver for each cylinder (can anyone confirm? Pretty sure there are separate wires from the ECU). I'm interested in looking at injector pulse duration to see if there is a difference from cylinder to cylinder - I have a "stretch" theory that these old ECUs aren't as bulletproof as we might think considering all of the (what I think are) electrolytic caps in the design that undoubtedly have long since dried up. But it is probably something else - maybe the rings haven't seated well, etc.

Any other things I should be checking? I really don't want to make it more rich just to cool this one cylinder.

Speeduino is pretty interesting - I didn't know about this. I have a megasquirt v1 thats been sitting on a shelf for, jeez, 14 years now. Original plan was to fuel inject my t4'ed beetle, but my dells run well enough for how frequently I drive that car. My westy is a different story - FI is a must. I'm leaning more and more towards some more of DIY injection vs. ljet, but considering my familiarity with the arduino IDE and C vs. assembly (megasquirt), this might be enough motivation to make the plunge.

On the flip side, free time is at a premium, so if I can get the stock setup working reliably with enough instrumentation to give me a heads up that something is going south, it is probably preferable.

-Chris

[Piledriver]

It can also be due to injector balance/dirt.
You should consider having them cleaned and flow checked, they are after all old enough to have grandchildren.

The old MS1 is relatively easy to upgrade, there is a $12 "open/Free" design adapter board that basically allows adding and MS2 and MS3X for full sequential in a ~plug and play manner, but for another ~$100 it could be a full MS3+MS3X setup.
(It can all fit in the Ms1 box if you make new ends and flip the MAP sensor to the top of the board)

I want to play with the Speeduino badly but I have way too many projects going at the moment.

[type11969]

I actually have a MS3X from more recent times along with crank fire ignition parts (trigger wheel, coil packs, etc). But alas, life got in the way.

After spending the last 9-10 months of free time working on this engine rebuild and a bunch of rust repair, I'm not super inclined to sideline my Westy for another good stretch while I get DIY FI up and going, but perhaps that is what I will need to do.

-Chris

[Piledriver]

Once you get the hang of it its much easier than carbs, but the initial learning curve is kinda steep.
99% of it is learning how to use Tuner Studio.

[raygreenwood]

The biggest temp problem between the 3/4 bank and the 1/2 bank is from cooling air static pressure reduction on the 3/4 side.

Its all in the cooling air manifold. I do believe Pile is correct that the Vanagon had improvements.

Bluntly put there is a longer path from the fan exit area to the 3/4 side....AND.....about 2.5 times the volume inside the shroud on the 3/4 side.....so the air exits the fan into a larger area dropling some velocity and static pressure. You then get the air being squeezed over the oil cooler ramp....and blowing all the way to the back at #3 where the sheet metal reduces volume and height. Static pressure builds there and the air turbulence makes it hard for the air to make a 90° turn downward to flow through the head fins.

Long ago Jake noted some of this and there was conversation that noted that possibly a vent hole back by,#3 could allow venting some of the static pressure to allow better flow. This problem is mostly at higher rpms. At lower rpms...its a static pressure issue because volume is down.. At higher rpms the volume is up and the static pressure and sheet metal volume now work against you a little.

For stockish engines.....the differnece is not a real problem. When its a higher performance engine.....this balance issue can be a problem and is one of the reasons Jake worked sp hard on vanes and cooling air steering in the DTM.

Also I have found that yes.....due to the overly simple steel tube ring main of D-jet and L-jet....on D-jet especially....with the regulator on the left side and fuel moving right ro left.....the left side injectors have less volume related stability. Part of this is the harmonic from the fuel pressure regulator and part is from pressure drop from the 1/2 injector side.

A better pump....higher volume helps a little. But.....in later years with L-jet on the beetles....probably vanagon....and Digifant for sure....they got better fuel pressure volume and distribution with the small inline fuel plenums you see on L-jet on the beetle. Those are the gray plastic tubes about 20mm ish in diameter. They have accentral tube inside that has holes that flow fuel into the chanber in the opposite direction from the injector outlets. They give the whole chamber pressure and volume to keep the first injector outlet from robbing pressure too quickly and dropping the pressure enough to cause a spike. Ray

[type11969]

Ray-

Thanks for the detailed feedback. I took a look for Jake's data but it looks like the images have been pulled. Any recollection or firsthand experience indicating if the 30degF difference I am seeing between #1 and 3 is normal or not?

Thanks,

Chris

[sideshow]

Well FWIW I remember similar numbers ~30 at the peaks...nothing I would cringe over

[type11969]

Thanks, unfortunately this is a sustained, not peak, offset. I hope to have time this weekend to dig in to it more.

Thanks,

Chris

[falcor]

On my 2276 T¤ I have around a 30 degC difference between cyl1 and cyl3 but on mine the #1 is the hotter one.
I'm going to install CHT's on the #2 and #4 cylinders too this winter to get the full picture.

[type11969]

Managed to get a bit of testing in today. Again, 60-65mph, light to medium loading, 14-15:1. Adjusted the valves first thing, confirmed max advance at 28deg, vac can disabled. I hooked up the heat exchanger flapper valve and blocked off one hole in the tin around #3 (mech fuel pump cars?). Unfortunately no change. 390 up to 400 on #3, 370-380 on #4, 370-380 on #1 and 350-360 on #2. Bummer. Swapped injector harnesses on #3 and #4 - so much for my stretch theory, no change. So this seems like a real, pretty miserable, cooling imbalance.

I think the only thing I can do now is richen up the mix a hair. So tomorrow I'll mess with the AFM.

In other news, I have a speeduino on the way. Not because I necessarily think it will do better than the stock system, but I do like the idea of more control and think it should be an interesting project.

-Chris

[Piledriver]

I would have swapped the injectors rather than the harness.

[type11969]

Good point. I'm operating under the assumption that they are still fairly balanced since I had them rebuilt when I first built the CS, but that was a long time ago now . . .

[Piledriver]

I suspect the Speeduino will run tight little circles around LJet while patting it on the head, if its anything close to MS2-extra functionality...

The ignition control is probably more important than fuel side.
I suggest either a set of LS2 truck coils or one of the VW 4 tower waste spark coils (106B or later rev)
The former allows per cylinder timing tweaks with a RPM/load map.(sequential)
Sequential fuel allows the same (with fuel volume and perhaps injection timing, not sure on Speeduino)

The Tunerstudio/Megalogviewer (tuning software and log analysis) software is excellent if registered...
It's actually quite excellent even not registered, but the few missing "registered" features can really help.