..after a power upgrade - what do I do to the LPG system..?

[mrLongstaff]

Hi all.

as title suggests, my '99 Forester is due to get a new wrx ecu, td04 turbo and deCat exhaust fitted next month when I get it ecuTek re-Mapped.. My question is, what do I have to do to the LPG system or its ecu to compensate for the extra 80-100 expected bhp increase?

TIA..

LongI.

[sijones]

hope that the injectors and vap will keep up.... you might be best checking what vap you have and it's rated output to see if there will be any issues...

[rossko]

It's unlikely your existing mystery system has enough spare capacity in hand to handle another 100bhp. (Most systems struggle with off the shelf 4-cyl turbos).
Find out what you have, then we can assess the shortfall and what options there are to go forward.

[mrLongstaff]

It's unlikely your existing mystery system has enough spare capacity in hand to handle another 100bhp. (Most systems struggle with off the shelf 4-cyl turbos).
Find out what you have, then we can assess the shortfall and what options there are to go forward.

OK, thanks Rossko.

It was registered as an EayJet system in 2007. The vaporizer has ROMANO written on it.

pictures here..

http://carsinalbion.blogspot.co.uk

I hope this helps with the suggestions..

Steven

[mrLongstaff]

I was intending to get my local LPG man who recently serviced it to do a recheck and a retune if possible afterwards.

I've also read that most LPG injectors can cope upto about 280hp before new/additional injectors are needed. Is this true?

Regarding the main mapper, its worth asking him if he retune the LPG ecu.. but then again, its new software and perhaps outside of his domain..

In the end, I guess, I'll have to see if the LPG system actuallly works at all when the car has been upgraded..

L

[rossko]

I wouln't expect to find a Romano reducer on an Easyjet system, perhaps its a previous repair. I don't know what that one is rated at.

I was intending to get my local LPG man who recently serviced it to do a recheck and a retune if possible afterwards.

erm, if he is familiar with your system, and likely to be the man sorting it out, why not ask his opinion now? Although from your previous post, he wasn't happy about setting it up as it is.

I've also read that most LPG injectors can cope upto about 280hp before new/additional injectors are needed. Is this true?

No.
Injectors have a per-cylinder rating. 45bhp each (which you ought to have now on a Forester Turbo) is considered "big".
So you need to know the expected total output of your engine and divide by four to size required gas injectors. Let's guess you are aiming around 270bhp, your injectors ought to be uprated to ~70bhp each, which would be considered "unusually big".
Easyjet injectors come in a range of sizes up to 28bhp, can't tell what size you have there.
http://www.autogasitalia.it/db/prodotto ... ng&id=1685

Tommo may know more about your Micromise system, but I begin to think it is seriously underspecced anyway. From your last post, it is sick anyway.

In the end, I guess, I'll have to see if the LPG system actuallly works at all when the car has been upgraded..

That might be a bit cavalier. If it runs lean under power (which would be the most likely shortcoming after upping the bhp) then at worst it could melt your engine or new turbo.

deCat exhaust fitted

oh, and beware of MoT failure on removed cats

[markocosic]

Power handling of the injectors depends on boost though?

If they're screwing up the boost pressure to make the extra power then the flowrate (in kg/sec rather than litres/sec) will increase too. (in proportion less offset)

http://www.cosic.org.uk/MarkoStuff/lpg/LPG-calc.xls

52.5 kW/cylinder at 0.8 bar boost is about the same injector size as 38kW/cylinder naturally aspirated. 280 bhp is probably near 1.2 bar boost and you're down to the equivalent of 34 kW/cylinder naturally aspirated, which isn't that unusually big?

Then again, being a Subaru it'll be throwing in 400 bhp worth of fuel to make 250 bhp anyhow...

What injectors are fitted/being fitted? (petrol side)


"Seeing what happens" will chargrill pistons for sure. I'd get it mapped on petrol by the petrol guru then have the LPG guru look at that map and clone it on gas.

[marek]

Mark,
There is a small double counting error in your spreadsheet with regard to opening/closing times:-
If the injector takes 2.36mS to open and 1.2mS to close, then the total "dead" time to deduct (when you can't inject anything) is 1.16mS. Knowing that and your stated desire not to exceed an 80% duty cycle, you can recalculate your desired flow rate at full throttle, since the total non-injectable time per full cycle is 2.36mS plus the 20% of time that isn't the 80% - the closing time is a red herring because your injector is guaranteed to be open, so you don't need to deduct it from your calculation.

For example, if the ECU commands a 2.35mS pulse, no gas flows; if the ECU commands a 2.56mS pulse, 1.4mS worth of gas flows (0.2 plus continued flow for a further 1.2mS worth of gas) assuming it flows equally when open; if the ECU commands 3.36mS, you get 2.2mS of gas, etc.

If 20mS is the cycle time, then you can inject for 16mS and will get 13.64mS of gas. Since the lion's share of the closing time is mechanical (for the pintle to close its spring versus gas pressure) rather than electrical (for the magnetic field in the solenoid to collapse), you may find that the closing time can overlap the 20% to some degree.

Also be aware that the driver circuitry affects opening time:- using resistors will be slower than custom circuitry which opens the injector with a high current pulse and keeps it open with a lower holding current (so called "peak and hold"). The former is a compromise to make low impedance injectors work with circuitry designed for high impedance injectors.

The calculation primarily affects your ability to tune at idle, since you will never be able to deliver less than 1.2mS worth of gas per cylinder per cycle.

kind regards
Marek

[markocosic]

Hi Marek,

Sorry for the late reply! This makes a lot more sense now that I have had a chance to tune a system for the first time - thank you for the explanation.

I was not sure how to treat the open/close times ad the fuel delivered during that time, so assumed that it was zero to be "on the safe side" (the spreadsheet would specify injectors slightly larger than those needed)

I've started a new thread over on the AC forum ( I'm guessing that you're Marek.Bezubik @ AC Autogaz Support? ) with a couple more questions related to my install.

Cheers,

--
Marko

[tommo2009]

The Easyjet injectors, if they are 21's will quite easily cope with (They are rated to 44hp per cylinder)50hp per cylinder and i have them running more than that on a few cars.
The romano vap which as Rossko says, is an after fitment addition, will be your main limiting factor as they don't flow enough gas.

As it stands the system you have will quite easily cope with a standard 2.0 turbo forester, but put another 80 - 100 HP on it and you are going to struggle with what you have.

[Punx0r]

I don't know if this feature is common to other systems, but the OMVL software lets you supplement the LPG with petrol above a certain injector pulse time. I plan to use this when I exceed the ~65HP/cylinder ceiling on mine, so I hope it's effective

[dieseldog]

Hi Tommo,just looking back on Easyjet topics and i see you have fitted a lot of these.The problem i have is taking the manifold nozzles out of a plastic manifold without ripping it to pieces,are these nozzles available or will the Easyjet ecu run other injectors ie OMVL etc.Thanks.

[LPGC]

Power handling of the injectors depends on boost though?

If they're screwing up the boost pressure to make the extra power then the flowrate (in kg/sec rather than litres/sec) will increase too. (in proportion less offset)

Not on your (or most) systems, because the pressure from the reducer is relative to manifold pressure (as opposed to relative to atmospheric pressure).

Most normally aspirated vehicles (with sequential petrol systems) idle with about 3ms of Pinj duration and go up to about 16ms Pinj when the engine is making it's full torque. The 16ms limit allows the engine to do over 6000rpm with time for the petrol injectors to open/inject/close for each induction cycle, the 3ms at idle is simply the fraction of that full load flow that the engine needs at idle. Turbo engines can still only go up to 16ms at full torque (if that torque extends to high rpms anyway), so because turbos make a lot more power at high loads the injectors have to be able to flow more fuel. Since turbo injectors flow more fuel, but engine load at idle is still similar to an NA engine at idle, to flow the correct amount of fuel at idle the injectors will need to de less than 3ms at idle, usually about 2.5ms depending on the spec of the engine. Subarus tend to use turbo spec petrol injectors even in their NA engines, so most NA Subarus idle at 2.5ms Pinj and (because they don't have a turbo) only go up to around 12ms at torque peak.

Marek has explained some of the constraints regarding LPG injectors. I'll add that some LPG injectors are far quicker to open/close than others, some have better peak flow rates than others, etc, and add that the 16ms total 'window' that the petrol injectors have to open/inject/close applies equally to LPG injectors. Now, if we have a (theoretical) LPG injector that takes 4ms to open and close then a 4ms pulse to that injector will result in a very tiny pulse of gas, maybe equivalent in fuel terms as a 1ms pulse from the petrol injectors... If we subtract the opening/closing time of this LPG injector from the 16ms window we are left with only a 12ms window for this LPG injector to actually inject gas (at full load and high rpms).. So while the injector is actually flowing gas, it must be flowing gas at a higher flow rate than the petrol injector (it is putting in the same amount of fuel in 12ms as the petrol injector did at 16ms). Already we can see that the flow 'resolution' (for want of a better word) for this theoretical LPG injector is only 3/4 (12/16) of the resolution of the petrol injector (it must be controlled maybe 30% more accurately in terms of pulse duration than the petrol injectors to offer the same level of accurate fuel dosing). LPG systems multiply the Pinj duration figure (by figures set by the installer in the calibration / map) to arrive at the Ginj duration figure. At full loads we want the Ginj to be very similar to the Pinj figure (because we don't want to go beyond the duration of the window), whilst at idle we expect the Ginj figure to be a little higher than the Pinj figure (we expect a bit of a multiplier at idle but we don't want a multiplier at high loads). One property/aspect of LPG injectors is their linearity at low pulse durations (at low durations, if you increase the duration by 10% will it inject 10% more gas or will it inject 20% more gas? I mentioned above that the theoretical injector might inject about 1ms worth of gas if repeatedly pulsed for 4ms, but at this 4ms pulse duration the injector won't be metering gas properly - sometimes it might inject 0.7ms worth of gas, sometimes it might inject 1.3ms worth of gas... it isn't accurate at such low pulse durations). It follows that usually the faster responding the LPG injector is then the lower the duration it will start to work linearly from (but an LPG fuel map that starts high, goes low at two thirds load and goes back up again might show that the injectors are reaching their peak flow level - not usually seen except on powerful engines with inadequate flowing injectors). Bearing in mind that fueling precision at idle is more critical than at high loads - If we have an average 1.6 NA engine that idles at 3ms and goes up to 16ms at torque peak at higher rpms, then these theoretical fairly poor injectors might have a map with a multiplier of 1.4 at the 3ms idle and a multiplier of 1 at the 16ms full load. Since these figures mean that the LPG injectors would be working at say 4.4ms at idle and 16ms at full load, and since this is engine only makes about 25bhp per cylinder, then the LPG injectors will be working within their linear range at all times (except maybe on the over-run) and will not be approaching their maximum flow limit... we would have lost some of the fueling linearity but this setup should work OK and not cause any problems from a calibration / drive-ability point of view. If we fit these injectors on a Turbo engine that idles at 2.5ms and goes up to 16ms, then (since the injectors can still only work linearly from say 4.2ms we can still only go down to 4.2ms at idle and we can still only go up to 16ms at full load), we have lost a lot more in terms of resolution (less accurate fueling). In addition these poor theoretical injectors may approach their maximum flow rate at higher engine loads. This kind of injector saturation can mean the working pressure needs to be increased (it would anyway for the turbo power levels but even more so with the saturation effect to get the injectors to flow enough gas for top end power), and this can mean that at idle the injectors would have to work below the minimum pulse duration at which they start to work anything like linearly... Imagine the calibration on these injectors that take 4ms to open/inject a little/close with a Pinj at idle of 2.5ms. Do you set the multiplier at 1.8 or at 1.7 for a Pinj of 2,5ms? One may be too rich, the other too lean (sure we can set it in-between 1.7 and 1.8 but here I am highlighting the fussy setup)... The situation gets worse when we consider that LPG systems need to compensate for the temperature of the gas (cold gas being more dense and vice/versa). So with poor injectors you might manage to calibrate the system to work OK when the engine is warm at both idle and at full loads, but then when system first switches to gas on a cold day (when the gas temp is also cold), the LPG system will remove a % of the Ginj to allow for the denser gas... which can again mean the poor injectors are trying to meter gas into the engine with a lower Ginj than at which they will work linearly - This all adds up to making for a poor performing LPG system - Slow responding / un-linear / approaching maximum flow rates, in injector terms, is a bad idea on an engine that has a wide dynamic working range (the dynamic range here being fuel requirements at idle and over-run compared to fuel requirements at full load at high rpms). If this turbo engine had been fitted with better theoretical LPG injectors which only take 2ms to open/inject/close, were linear from 2.5ms and didn't approach saturation at the engine's full power, then we would have more resolution on the turbo setup than we had on the NA engine with the poor injectors, at idle the multiplier could be maybe 1.2 / 1.3, at 16ms full loads the multiplier could still be at 1. With the better injectors the turbo engine would be a lot easier (not to mention possible) to calibrate properly and the fueling would be more precise throughout the range of the engine.

On Subaru engines that idle at 2.5ms, even the NA (none turbo) versions, it is a good idea to fit linear injectors by default. On Subaru turbo engines this can be completely necessary - At least if you want the possibility of a smooth idle AND the ability to run on gas at full power.

Simon

[Gilbertd]

What you've written is probably very interesting and must have taken you quite a while to type it. I say probably as I can't be bothered to read a lengthy reply to a 2 year old thread......

[rossko]

dieseldog - Tommo does not appear much here of late. You might try private messaging him. Unscrewing injector fittings is usually just a case of breaking the loctite's grip.

Not on your (or most) systems, because the pressure from the reducer is relative to manifold pressure (as opposed to relative to atmospheric pressure).

Markocosic's point is that boost raises manifold pressure under load (else there would be no benefit from having boost i. e. extra air mass) and so indeed the absolute gas pressure also rises in sympathy. The gas density rises, and so a given injector (limited as to the volume it can handle, not the mass) can flow more mass of fuel at the same durations under boost conditions.

[LPGC]

What you've written is probably very interesting and must have taken you quite a while to type it. I say probably as I can't be bothered to read a lengthy reply to a 2 year old thread......

Oops! I didn't look at how old the thread was, I only took notice of when the last post was

[dieseldog]

Thanks for reply Rossko,and apoligies to LPGC for the mislead.I tried loosening one of the nozzles but it feels like something is going to break.I thought it might be easier to blank the old nozzles of and fit new but cant get a lot of information on Easyjet/Mistral.What i wanted to know was would the Easyjet ecu run another type of injector ie Regfast etc.Thanks.

[rossko]

What i wanted to know was would the Easyjet ecu run another type of injector ie Regfast etc.Thanks.

It's not likely to work well. Manufacturers would not generally include the fine tuning parameters for other makes in their firmware (e.g. will these BMW injectors work in my Honda).
You would need a fair match in electrical and performance characteristics between two types, and that info cam be hard to obtain.

[dieseldog]

Thanks.

[classicswede]

What i wanted to know was would the Easyjet ecu run another type of injector ie Regfast etc.Thanks.

It's not likely to work well. Manufacturers would not generally include the fine tuning parameters for other makes in their firmware (e.g. will these BMW injectors work in my Honda).
You would need a fair match in electrical and performance characteristics between two types, and that info cam be hard to obtain.

Bad example as petrol injectors are mostly the same other than flow rate. Match the resistance and close on flow rate and the petrol injectors should run ok.

What injector options does the software have. Whatever you fit you will need to re calibrate the ECU