Measuring the vacuum? New VW Type1 installation& misfire

bon_corno · Post by **bon_corno** » Wed Sep 26, 2007 2:21 pm

Hi,

I installed my MJLJ few weeks ago. This is my first experimental installation. These will go to my turbo boosted 1600 engine which I am planning to complete in 2-3 months. Current engine is a dual port 1285cc Type1 with dual Weber 36 IDFs, custom ss headers, external oil filter.

Here are some images

Vacuum canister. I'm using MAP option in my setup.

This is a custom trigger wheel bolted to the crank pulley. It is 8mm thick at outer part and 4 mm at inner part. I removed some part of the inner part with lathe. What do you think about this trigger wheel? Does it need any modification to perform better? I also want to lighten it as much as possible. What can be the minimum wheel thickness? Sensor mount is solid, there's no sensible shaking. The gap between the sensor and wheel was more before and it ran perfect. After having this misfiring problem, I decreased the gap. I didn't measure it, remember like 2 mm.

MEASURING THE VACUUM & MAPPING
I had my first challenge with the vacuum signal. It was not a surprise. I only have one vaccum port on my carbs. I tried many ways, but couldn't achieve to dampen the signal. Then I decided to pick the vacuum signal from synchronization tubes. These tubes open to the carb barrels just below the butterflies. There's a continuous signal between 65 and 70 kpa at idel depending on the rpm. I collected all of them in the canister, the signal is now smooth. Vacuum decreases when I push the throttle and at WOT I read 102 kpa. According to what is happening, I prepared the following igntion table. What would you tell about the vacuum measuring method and following ignition table?

* The configuration file is also attached to this post.

MISFIRING
It doesn't happen at the same rpm but 95%of all misfiring occurs at 3200 rpm. It is not easy to tell it with words but I will try. First of all, there is no problem when there's no load. Under load, the engine pulls normally up to 3000 rpm. After 3000, there is a noticeable performance drop and backfiring occurs in the exhaust. If I don't pull the throttle back, rev increases slowly if the load is low and I don't hear the backfiring sound sometimes but the performance is dead. I tried to check what is happening after 3000 rms with timing light (without load, in the garage). It flashes perfectly up to 3500 rpm then it flashes lke randomly. The problem is I don't trust my timing light. It is a very simple and old one. I also checked the voltage supply, car's regulator limits the charging voltage at 14,5 V correctly. Is this an acceptable voltage for MJ, Edis and coil?

What would be the cause of this problem?

I have a spare set of Edis, coil and sensor. If I can spare some time tonight, I will replace them one by one and see if any of them is faulty.

Thanks in advance for all comments and advises!

Ibrahim

rbalmford · Post by **rbalmford** » Wed Sep 26, 2007 3:58 pm

Hi, an initial scan reveals your map is effectively 'upside-down'!! The top row with the lowest pressure is for closed-throttle (low load) and the bottom row with the highest pressure is for open-throttle (high load). Conventionally, you can increase the advance at low load/part throttle, so the numbers in each colum should increase towards the top (apologies if i'm labouring the point too much!!). Try inverting your map (easiest with a text editor on the .mjlj file). One approach is to set the full-throttle row first, and when happy with that (road tests) work on the part-throttle map.

Rich.

Post by **brentp** » Wed Sep 26, 2007 7:24 pm

rbalmford wrote:Hi, an initial scan reveals your map is effectively 'upside-down'!!
Rich.

Correction - Ibrahim's load axis is indeed correct for a MAP setup. You can verify this with the factory default map as well. Plus, the configuration software should block you from inverting the direction of the bins.

Forcing the map to be inverted will cause problems in the controller as it will be forever stuck on the first load bin!

Regards,

MartinM · Post by **MartinM** » Wed Sep 26, 2007 7:37 pm

The axes are the correct way round - it's the advance bins that are "upside down"

The easy way to remember is the old-fashioned phrase "vacuum advance" - more vacuum (lower kPa) = more advance (higher degrees)

So in Ibrahim's map, the higher advance values should be closer to the (physical) top of the map e.g. 800rpm @ 40kPa should be 13BTDC, 800rpm @ 102kPa should be 8BTDC etc etc

I'm sure that what Rich means...

Unlikely to solve the 3200rpm problem, but worth putting right first and seeing if it still occurs

EDIT: What happens if you disconnect the MAP sensor so that you are fixed on or around the 102kPa row and you just have 2D ignition? There's nothing special going on in the map at about 3200rpm so it might not be optimum, but it would put the MAP sensor/vacuum in or out of the equation...

Martin

david jenkins · Post by **david jenkins** » Wed Sep 26, 2007 7:50 pm

Just a couple of things to consider:

1. Is your VR sensor's bracket really rigid? It's hard to tell from the picture, but it does look a bit 'extended'. Wobbly sensors can cause problems at certain engine revs.

2. Most people find that the sensor works best at somewhere around 1mm distance from the wheel, sometime even closer.

Neither of these things may be the cause of the problem, but they're worth checking.

cheers,
David

rbalmford · Post by **rbalmford** » Wed Sep 26, 2007 8:50 pm

MartinM wrote:The axes are the correct way round - it's the advance bins that are "upside down"

I'm sure that what Rich means...

Indeed it is, I was referring to the columns of advance values, not the load axis itself.

bon_corno · Post by **bon_corno** » Sat Sep 29, 2007 11:05 am

I changed the EDIS and coil. No change.

I know this temporary sensor mount is not handsome but believe me it is rigid. It is attached to three studs and the metal is thick. I will decrease the gap to 1 mm.

I was also in doubt when I measured the vacuum for the first time. I'm measuring the vacuum from the manifold like it is described in vehicle installation guide. When idling I read 68 kpa on average and it is between 68 and 102 kpa under load. During deceleration it reaches 50 to 40 kpa. If I don't understand the logic wrong, I should increase the advance value as the load increases. That's why I added advance in the opposite direction. Is this a reliable way of measuring the load? If the method is wrong, then I will remove the carb with factory blocked vac port, open it and pray to have a smooth vac signal from only two barrels. I hate to remove and install these carbs!

As an alternative to disconnecting the vac hose I used the advance curve of Bosch 009 distributor before which is a mechanical only disty. I tried so much things that I can't remember the exact result now. I will try it once more. If it won't help, I will remove all MJLJ and EDIS components and install the stock disty to see if it is really an ignition problem. I will post the result.

Ibrahim

MartinM · Post by **MartinM** » Sat Sep 29, 2007 12:30 pm

bon_corno wrote:I was also in doubt when I measured the vacuum for the first time. I'm measuring the vacuum from the manifold like it is described in vehicle installation guide. When idling I read 68 kpa on average and it is between 68 and 102 kpa under load. During deceleration it reaches 50 to 40 kpa. If I don't understand the logic wrong, I should increase the advance value as the load increases. That's why I added advance in the opposite direction. Is this a reliable way of measuring the load?
Ibrahim

No, you should decrease the advance as the load increases/MAP kPa goes up. You advance as the vacuum increases/MAP kPA goes down.

Just think about the engine (or you!) trying to suck in air through the carb butterfly...

...On low/no throttle/deceleration the butterfly is closed and the engine is sucking as much air in as it can, but the butterfly is closed so it can't get much air and you get a vacuum (low kPa) in the manifold.

......On open/full throttle/acceleration the butterfly is open and the engine is sucking as much air in as it can, but the butterfly is open so it can get a lot of air so there is less vacuum (higher kPa) in the manifold.

Also, you'll remember pinking under load with a dizzy system - and this was caused by over-advanced ignition under load. The solution was to reduce advance ie as the load increases, the kPa number gets bigger and the advance needs to decrease - that is why the advance figures in your map are "upside down"

MartinM · Post by **MartinM** » Sat Sep 29, 2007 12:36 pm

...and "vacuum advance" is only required at lower rpm - up to 2000-3000rpm or so - which is what you, correctly, have in your map.

On race engines with dizzy's they often don't bother with vacuum advance pipes - at race rpms it's not needed and you can throw away the vacuum advance diaphragm...

Post by **brentp** » Sat Sep 29, 2007 5:34 pm

Reading back on the thread, I'm surprised we didn't ask our usual reflexive question in response to mis-fires in mid-rpm solutions!

So - Try reproducing this with the MJLJ disconnected, running the engine at the default 10 degrees. Does it still happen?

If the EDIS module loses synch with the timing wheel, it will stop firing until it re-synchronizes, and this should reflect on your timing light.

If you are suspecting your timing light, get rid of that doubt and fetch another one- However- it's seems highly coincidental that the missing flashes coincide with engine behavior, so be prepared to have your new timing light agree with the old!

DannyP · Post by **DannyP** » Sun Sep 30, 2007 4:27 pm

Ibrahim, be very careful with preignition/pinging. Aircooled engines are very easily destroyed by detonation. You should have advance set at 28-30 degrees around 3000 rpm or so for full throttle. After 4500, you can actually decrease it about 1 degree every 250 rpm, or 4 degrees per 1000. You can add more advance up to 3000 under part load/throttle conditions. This is what Jake Raby told me type1s like to see.

Many others on here have had resonant/harmonic problems with sensor vibration. Sounds like either that or a shielding problem on the wiring. Good luck and report back what you discover.

BTW, trigger wheel is very,very nice! Good job! Now you've given me an idea for mine, thanks.

bon_corno · Post by **bon_corno** » Sun Sep 30, 2007 11:47 pm

The problem disappeared after installing the old DVDA disty setup. Advance goes up to 32 degrees with vaccum hose disconnected. At least, I know where to focus now. Next step will be building a new sensor mount or reinforcing present one and trying different sensor to trigger wheel gaps. Does the material (steel/alu) of the mount make any difference?

And also my timing light worked properly with the disty setup. It is working. I can not confirm if EDIS is trying to re-synchronize since I don't know what to see. All I see is random flashes of the timing light but advance values seems to be at correct degree during these random flashes. Can the electromagnetical interference coming from the coil affect timing light's operation? (One of my tachometers doesn't work correctly when it is near the coil and also it disturbs the serial communication between MJLJ and computer!)

Before installing the disty, I disconnected MJLJ. There was no problem when idling but when I opened the butterflies it started to fire through the butterflies. Probably around 2500 rpm. I was alone and couldn't read tachometer. I also tried disconnecting the MAP sensor. Nothing changed.

Lastly, I will prepare a new table soon and post it here. I think I got the idea now.

Ibrahim

bon_corno · Post by **bon_corno** » Mon Oct 01, 2007 6:43 am

I forgot to write.

When EDIS was disconnected, static 10º advance was very solid. There was no change however I couldn't try upper rpms since rpm acceleration was slow and the engine started to backfire.

Ibrahim

bon_corno · Post by **bon_corno** » Wed Oct 03, 2007 6:11 pm

I read Patriq's recent Default cable to VR sensor from Ford not shielded !? post. My VR sensor cable is quite long. I collected the excess part of the cable by making some rounds behind the generator stand where two of the plug leads go.

I can't test it at the moment since the engine is running with disty setup again

Is signal of a VR sensor with shielded cable senstive to interference coming plug leads? Can this cause a problem like mine?

Post by **brentp** » Wed Oct 03, 2007 8:23 pm

Hi Ibrahim,

How long is long? and does it run near the plug leads at all?

Shielding comes in various forms with respective effectiveness- typically the best is a foil shield with full coverage. What degree of shielding do you have?

On another topic- looking at your VR sensor bracket, the first thing I would do is reinforce the part where you've split and bent the tab down for the sensor. I would cut a small triangular piece, fit it in the gap and weld it up. That should dramatically improve the rigidity of the mount. Should be a quick bit of work to do.

Regards,