LPG injection: weird cold start problem

[kikkegek]

I pulled up the log file from my -5 Celsius cold morning start. Just as a reference to my 10 Celcius cold morning start today. So 15K warmer.

The only thing I have changed in between is the injectors. Same design. Both drilled 2.65mm nozzles. But one has 2 ohm coil (the -5 Celsius shot) and the other has 3 ohm coils (10 Celsius shot).

This is the -5 Celsius screenshot:

Cold starting at -5C outside temperature.png (154.81 KiB) Viewed 1139 times

And this is the 10 Celsius screenshot:

Cold starting at -5C outside temperature.png (154.81 KiB) Viewed 1139 times

isnt it just weird how crazy the AFR is after the wideband gets online. I had none of that during thos cold morning starts when it was freezing. The car would fire up. I even had some "WUE" around 110 or 115% active back then, cause it loved to run a little richer when the coolant was still cold. Now it runs pisrich when it first starts and I need to pull 30% fuel in stead of add 15%. Thats a 45% difference in startup behavior....WTF...cant get my head round this.... it should be easier to start on warm temperatures...not harder...

[Marek]

According to that graphic, only a fraction of the water looks to circulate through the reducer, regardless of whether the thermostat is shut, so the useful coolant volume is much lower than 23l. The path back to the pump is much shorter and easier than the path up to and around the heater matrix. Any crud settling in the heater matrix also needs to be flushed through and the heater matrix is higher up, so can have air pockets sitting in it. This is a bad design for lpg. A better design would be to have a higher percentage of the water come through the reducer. The heater matrix tap method is fine for smaller cars where not much heat is needed for the lpg side, but is not the basis for a good circuit for a large engine which uses much more lpg, particularly so at startup.
kind regards
Marek

[kikkegek]

According to that graphic, only a fraction of the water looks to circulate through the reducer, regardless of whether the thermostat is shut, so the useful coolant volume is much lower than 23l. The path back to the pump is much shorter and easier than the path up to and around the heater matrix. Any crud settling in the heater matrix also needs to be flushed through and the heater matrix is higher up, so can have air pockets sitting in it. This is a bad design for lpg. A better design would be to have a higher percentage of the water come through the reducer. The heater matrix tap method is fine for smaller cars where not much heat is needed for the lpg side, but is not the basis for a good circuit for a large engine which uses much more lpg, particularly so at startup.
kind regards
Marek

thanks for checking. Im not worried about the flow. Its how GM designed it to provide enough heat to warm up a big cabin in a huge car 40 years ago. And trust me...at startup when the thermostat is closed a lot of water passes through this hose. I know from checking circulation test after refreshing coolant to get the air out of the system.

And if you check my log files, the CLT temp rises from startup. So there is no lack of heat from the coolant. Even at the -5C startup. I do think the reducer might be contaminated on the LPG side. But if LPG would be able to pass through in liquid form at the first cold startup, then I'd expect the gas pressure in the gas hose to show higher pressures and it doesnt. So while I now only suspect the reducer for the rich startup, I dont see any leads to that in the log files. Do you?

[kikkegek]

what makes it even weirder is that later on the day, when the engine is cooled down a little (coolant around 25C) and I start it, the engine will run superlean. To get is started on the 1st cold start I need to pull about 20-30% fuel at that temperature to get it running. I will let it warm up to about 70C. And turned it off. Then when I start i later on day with the WUE numbers untouched it will run with AFRs of 20-22. Amazing how lean it will run.

this is the log file of the startup later that day. You can clearly see the lean AFR when the wideband comes online and then I have to reducer the WUE from 70 to 90% to get it back to an acceptable AFR.
https://drive.google.com/file/d/1Amkku7 ... sp=sharing

I really cant get my head around what might have changed/happened or be caused by the long trip...cruising 175km at 100 kph and 1800rpm and about 16AFR and 40C gas temperatures.

its doing it with the new Valtek injectors too. so its not the injectors.

Ill check my ignition system. Maybe some cilinders arent firing right when cold at first?

[Marek]

Perhaps it suggests the rubber diaphragm in your reducer is not very flexible when cold.

kind regards
Marek

[39ClevoUte]

Hi

I'm not a Chev guy. But with early fords (in the 60s heaters were sometimes options) engines had an external bypass hose. If you optioned a heater, this hose was replaced with the hoses to the heater and the heater did not have a tap, but had internal flaps to dirrect the cold air either via the core or not via the core.
Later engines had an internal bypass and the heater had a tap.
Your engine diagram shows you have no tap and two bypasses. This may or may not be adding to your problems. If it was me I would eliminate the factory bypass, one is enough. It might help a little.

Cheers

[kikkegek]

thanks 39ClevoUte

I dont know what the "tap" or bypass""would be, but these chevy's have a valve that closes the water to the heater core when the air conditioning is active. I have A/C but disabled it and now always have a nice flow to the heather core.

Also remember that I did not have any problems for weeks. I started this setup at cold wintermornings at -5 Celsius and it would idle just fine. For some reason the long drive (175km) has changed or adjusted something (no laptop was connected, no settings were changed) that causes this strange pisrich cold start.

So Im thinking towards the reducer like Marek is also pointing out. Would be weird though since its only couple hundred km in use, so should be in mint condition. But I'll open it up and double check to be sure. Im currently looking for a rebuild kit, but can only find it in Ukraine so far and Im in the netherlands. LOL

[Marek]

There is a relationship between CLT corrections for general warmup, ASE corrections to keep the car running after cranking, but before it is safe from stalling, gas pressure corrections and gas temperature corrections.
It is my impression that you have wanted to use gas temperature corrections as part of the warmup process, rather than CLT. What I think a good working setup is, is to use gas pressure and gas temperature corrections only for a normally running fully warmed up engine and leave the entire warmup period after ~45 seconds to CLT, WUE corrections. This will be a proxy %age number adjustment unrelated to the actual gas temperature, but related to CLT only. The reason for this is that the entire engine bay is at below operating temperature, so the gas temperature will follow the reducer temperature, which if fed by the entire coolant flow, not a small proportion of the total flow, will mean the reducer temperature (and thus gas temperature) will accurately track CLT tempertature. Once the engine has fully warmed up and is under normal running conditions, gas temperature and gas pressure adjustments can take over. Their only job is to make up for an overheated engine bay versus a highway cruising engine bay temperature, the inevitable pressure spikes either way when the throttle is snapped open/closed and the slow but steady very slight drop in gas pressure as the tank empties. ASE are there to stop the engine from stalling immediately after cranking is over and the idle valve moves to the run position (6 seconds max) through to the period of perhaps less than a minute later when WUE alone can keep the engine running.

An engine running very lean at startup later in the day suggests the ECU is being fooled into thinking these corrections are needed when they are not, e.g. a heatsoaked engine where MAT is not well reflected because the MAT sensor is badly placed. The same might be true of the gas temperature sensor if it doesn't accurately reflect the temperature in the gas rail, but the general engine bay instead.

kind regards
Marek

[39ClevoUte]

thanks 39ClevoUte

I dont know what the "tap" or bypass""would be, but these chevy's have a valve that closes the water to the heater core when the air conditioning is active. I have A/C but disabled it and now always have a nice flow to the heather core.

Also remember that I did not have any problems for weeks. I started this setup at cold wintermornings at -5 Celsius and it would idle just fine. For some reason the long drive (175km) has changed or adjusted something (no laptop was connected, no settings were changed) that causes this strange pisrich cold start.

So Im thinking towards the reducer like Marek is also pointing out. Would be weird though since its only couple hundred km in use, so should be in mint condition. But I'll open it up and double check to be sure. Im currently looking for a rebuild kit, but can only find it in Ukraine so far and Im in the netherlands. LOL

Hi
With what you have done by removing the valve/"tap" is ok. But I would eliminate the small original bypass.
Cheers

[kikkegek]

How? And what bypass?

[kikkegek]

I have been looking for a repair/rebuild kit for my reducer. I have only been able to find it in the Ukraine and not been able to get one so far.
Its a Femitec FR400, but is originally being sold as a Gurnet Vapojec Luxe

Gurtner_Vapojet_Luxe.jpg (122.85 KiB) Viewed 1100 times

Since I cant get a rebuild kit, I ordered a new reducer to replace it. I decided on a Tomasetto AT13XP that should be able to put out 365hp and should be more then enough for my current setup without turbo.

Tomasetto AT13XP (02).png (830.85 KiB) Viewed 1100 times

this way I can be 100% sure that its not hardware related, because I have replaced injectors and reducer and made sure the harness and peak&hold drivers are in proper order.

Hope this fixes it and get me back on the road.

2b continued.

[39ClevoUte]

How? And what bypass?

The one circled.
The reason:
You said the original Chev cut the heater circuit when the A/C is on. This then means this circuit would most likely/normally be off in summer and on in winter.
In winter having two bypass circits is OK as two give good water circulation in the engine, with the thermostat restricting water to the radiator.
Since you have removed the tap/valve you now have two active bypass circuits in operation which could cause the cooling system to struggle in summer.
If your thermostat when opening cuts off the bypass, then it's not so bad.
Not sure if your thermostat is a special type, but eliminating the bypass may give you a lot more choice of thermostats, as a special type would no longer be required.
You may like to try it simply by clamping the the little bypass hose closed with vicegrips. From cold watch how your temp gauge and performace reacts on trips. Then do the same another day with it clamped off.
Cheers

[kikkegek]

well....the new reducer came in the mail this weekend. Replacde the 2nd hand reducer I had with it. And all weird problems are gone.

This is the cold startup log from this morning:
https://drive.google.com/file/d/1yCnpxO ... sp=sharing

OK, its not cold anymore, but I can crank it, it fires and behaves perfectly normal. Also I did a couple full throttle runs and the LPG pressure now remains at around the set pressure ~150 kPa (above MAP).

I attached the MSQ tune for anybody that wants to take a look.

This is a picture of the old Femitec FR400 (alsno known as Gurtner Vapojet Luxe) on the left and the new Tomasetto AT13XP reducer on the right.

Gurtner_Vapojet_Luxe_-_Tomasetto_AT13XP.jpg (167.28 KiB) Viewed 1074 times

Im looking for a rebuild/service kit for the old reducer. I want to open it up and check why it is causing these weird problems. It was supposed to be only 300km old when I got it and I maybe put another 500-1000km on it before it started acting weird.

But for now...problem solved and Im happy I can cruise again and go back to tuning this thing.

thanks everybody for helping out

[kikkegek]

well...................its not gone

From about 10C LPG gas temperature and up it seems fine. But the first cold start below that keep having this strange going rich problem.

I have upgraded my ignition system to CNP with 8 LS D585 coil and it seems to have improved its capability to handle running rich. It has a really nice spark:
https://www.youtube.com/watch?v=G-U1Ta5EM3k

I have made a couple log files:
cold starts seems fine > 10C LPG temp
https://drive.google.com/file/d/1_M73v6 ... sp=sharing

cold start seems fine, little lean > 10C LPG temp:
https://drive.google.com/file/d/1hFU_OQ ... sp=sharing

cold start, gets rich, stalls about 5C LPG temperature
https://drive.google.com/file/d/1OfCNlv ... sp=sharing

On that last start I'd let it sit for 2min...started and it ran fine

I'm currently replacing a leaking radiator, so Im down and cant do any testing. But Im scratchin my head here. I have replaced just about everythiung now:
- reducer
- injectors
- plug wires
- coils

[Marek]

If everything "works" then consider you may have a tuning issue.

The injector pulsewidth starts from a cranking table (so far so good). This then tapers away and is replaced with the main table and after start enrichment ("ASE"). There is more than one way to deliver ASE and you may be having a transition problem as the warmup enrichment ("WUE") takes over. That is temperature controlled, but the temperature sensor may not be in the place needed to record accurtately the temperature of the coolant which drives WUE percentage.

Whilst none of this appears to be an issue at normal temperatures, it may be there is a set of operating conditions at low temperature where these factors interfere and are being fooled into delivering too much fuel. Perhaps this because the temperature of the water circulating to the reducer is out of sync to that coming through the sensor.

kind regards
Marek

[django]

hoi
je kunt een praxis test doen en de megasquirt bedriegen bij een koude motor ,en hem voorspiegelen dat de motor heet is
door een weerstand 2.7k parallel aan de bestaande clt te maken .en met een wisselschakelaar te switsen
[of kijk eerst wat de weerstand is bij 85graden van de sensor]

[kikkegek]

hoi
je kunt een praxis test doen en de megasquirt bedriegen bij een koude motor ,en hem voorspiegelen dat de motor heet is
door een weerstand 2.7k parallel aan de bestaande clt te maken .en met een wisselschakelaar te switsen
[of kijk eerst wat de weerstand is bij 85graden van de sensor]

lets stick to English for all other on this forum. Why would I want to trick the MEgasquirt in letting it think the engine is hot? What would be the benefit?

[django]

you have the problem that he injects cold too rich..
with the resistance in between, the mega thinks it is warm and injects less.
only for testing,,,whether the warm-up system is working too slowly

[django]

the most famous LPG injection systems first heat up to 30 degrees and only then switch from petrol to LPG
you want directly on LPG
I think that in the beginning phase the LPG is still too cold / liquid.
you can install a heat exchanger between the evaporator and the injector valves,this prevents liquid / cold gas from reaching the injectors in the initial phase
such a thing exists ,,plate heat exchanger vailant ,,you have these in different sizes.
with that you increase the evaporation capacity
what do you think about that

[kikkegek]

django, yes a heater will help vaporisation at low temps. And will give you full performance quicker. But its not neccesary.
I have started my car on LPG injection at -4C freezings temps and it warmed upjust fine as you can see in the video.
I'm 99% sure its some kind of tuning issue I'm missing that causes the engine to go rich at a certain setpoint.
temps are to high currently to do any testing.

video of cold morning start at -4C
https://www.youtube.com/watch?v=pC5JrA6js4Y