LPG Discussion Forum

hello all ,
im new to the site ,but thought id share a bit of info on my history ,and this project .

i built my first lpg conversion on my mk3 astra van in 2003..,with a 2 litre 8 valve ,fairly tuned .it dynoed at 145 lb/ft of torque and 140 odd bhp if i remember correctly ,and used a omvl vaporiser and the plastic cone insert omvl mixer ring .with a 42mm diameter ring .i used to get from 25 to 35 mpg using it as a long distance courier van ,with a 90 litre tank .low vacuum due to the cam caused the occasional problem on tickover when the fan came on ,but it was a great engine and cruised at high speed and low rpm effortlessly .

sold that one and its still going strong ,and baught a astra estate mk 3 . used the 16v 1600 in a highly tuned state with lpg ,again a omvl vap ,and a wtv type ring 43mm,which i dropped to 36mm to get it to come in a bit earlier .this went well 23/33 mpg ..but i thought id try a 1.4 and lower the consumption and rd tax in one fell swoop.



heres a random misceleany(spelling) of pics ... and a link to the build thread ..

http://www.turbominis.co.uk/forums/inde ... tid=378624












































i built it all in a 2 week period ,in november ,and im now setting up .i started at 16.9 mpg ,and im up to 21 lol so pretty poor , but the design has yet to be really finalised ,im playing with pressure take off points for the vap .

i had a fairly small intercooler ,and that created more pressure in the pre ic pipe ,so i used that to give a higher flow from the vap since the differential from there to the mixer was higher ..but it was a bit rich , so i decided to fit a charge cooler and this gave less restriction in air flow ,but iv lost my higher pressure take off ..

latest step , is to create a dynamic take off at the inlet 4 inches before the 32mm ring ,and also to feed this to the omvl primary gas presssure diaphragm ,and block the bleed .this is ok but lean at 3500... work continues .

i m curious to see how well i can get it to run ,but i do have a set of 6 keihin injectors off a jeep to play with for free if i want ,and then just need to make up a ms ecu . i have a fair bit of mapping experience on turbo minis with these .


regards
robert

You managed to get modern engines running with a lpg pipe into the inlet manifold, love it. Great stuff.

thank you billy ....its not perfect ,,,,but it is cheap ..lol.

well intersting ,and maybe totally dead end development ,just been donated a new brc ecu and 4 keihin inj and a loom...um ..researching suggests the ecu is a piggy bank ecu ,so needs a petrol ecu to feed it data so thas no good ,,and the inj are green spot so tiny 90 bhp max ,loom looks nice though !!!

Some (very) old LPG injection systems did their own mapping, but none of those ECUs would be capable of driving Keihin injectors anyway. I've never seen green spots - Smart car?
I think you have just a curiousity there !

yes i think your right rossko , i supppose i can allways use them as boost pressure bleed valves .

quick update .i now have the water/air intercooler filled with water and pumping ,definately a bit more pokey .

also edging the advance up bit by bit ,and bringing the vaccum advance in sooner ,this will make it more economical,provided i dont go too far and detonate the life out of it .

have a much better fuel mix now ,just a problem around 3500 where it drops of the lambda guage (narrow band),and if i open the tap for more its too rich elsewhere .may put a rpm triggered bleed on it ,to bypass the tap .
im also not happy with how much it leans off on cruise ... its not going lean enough ...may have to bleed some pressure from the vapouriser pressure feed .

regards
robert

well some success's have occured ,

with the water/air intercooler working well, its going better and better .still has the huge hole at 3500 to 4000..im thinking more and more this is a harmonic pulse effect at those rpm with the long tube inlet ..reversing flow from the ring .

another bit of progress ..on the ignitoin mapping front , i decided to mimic the vacuum advance of the old 1.3s distributor ..this brings in a lot more advance just off full throttle at around .8 bar .so did this and now on 165 miles with one green light showing ,this would previously have been at around 140 miles .will fill today and work out if im now above 22 mpg .blimmen hope so .

no detonation noticeable yet .i also advanced on boost a couple of degrees ..looks like this 8 valve will take a bit more advance than the 16v ..slower burn .

regards
robert .

hurrah .25.54 mpg ...getting there bit by bit !

That's good mpg. My Zafira with 1600xe standard engine in only does 26mpg average with a modern multipoint lpg kit.

billynoband wrote:That's good mpg. My Zafira with 1600xe standard engine in only does 26mpg average with a modern multipoint lpg kit.

thanks billy ,hopefully a bit better this time just wacked in another load of vac advance ,...the next fill up will tell.

well , i now have the chance to fit a single point feedback valve doofah off a saab , i think its a 1999 leonardo electronic power valve lambada feedback one ,should this be fairly easy to fit ? bear in mind , i have no ecu functions and the car is purely lpg no petrol . so no tps voltage as it is .

thanks
robert .

The Leo will require a lambda sensor ; I don't think a wideband will do, but read the book and see.
http://www.aeb.it/products/feedback-systems/leonardo

You should get away with no TPS, but I think the Leo will more effective if you can arrange a simple throttle microswitch so it can detect off-idle at least. You can always add that later.

thank you rossko ! luckily i just took my wideband out and put in a simple narrow band ,so that should be ok .

looks like i may need to buy some software and interfacing gubbins ?

http://www.ebay.co.uk/itm/AEB-Leonardo- ... 5646765f2a

Or download the software from http://www.vogelsautogas.nl/documenten/ ... .3.1.0.zip and build yourself one of these.....

thank you gilbert ,most kind of you , not sure about my making electronics skills but i might have a go .

new direction ish ,a megasquirt and a second hand load of injection bits , now to research flow rates and injector specs ...
















not really sure if the one temp sesor on the vap is enough for good economy ,no pressure aor gas temp input ,but i suppose the pressure and gas temp are pretty relative to the water temp in the vap ? ah ,well time will tell.

good thing is i already have the engine on megajolt ,so a simple swap to squirt on the ignition front .

injector specs....

2 POWER OF COIL 48 W – 144 W

3 PILOTING TYPE PEAK and HOLD

4 TIME FOR PEAK CURRENT 4,5 ms

5 ELECTRICAL CONNECTION Integrated waterproof connection two-ways

Amp/Delphi compatible 1,5 mm

6 RESISTANCE 3 Ohm +-4% - 1 Ohm +-4%

9 MAX WORKING PRESSURE 0,45 Mpa At 12V D.C.

11 NUMBER VALVE SEAT AVAILABLE 4,3,2 and single

12 OPENING TIME 3,3 ms

13 CLOSING TIME 2,2 ms

14 WORKING TIME 100%

15 MAX AMBIENT WORKING TEMPERATURE -20
16 MEDIUM TEMPERATURE -3018 INLET/OUTLET Connection for rubber hose D 10 mm IN and D 5 mm OUT
22 VALVE SEAT DIAMETER 4.0 mm

vap specs ..


apparently good for 190 hp ,whcih is fine fro my little 1.4 .oh yes .so looking promising .actually i just realized , i don't know if the ''intergrated temp sensor'' is reading coolant or gas ! hmmm.
• R67-01 & CSA homologated
• Compact design: Ø125*95mm
• Solid design: tested with 60 bar burst pressure
• Black anodised aluminium
• Adjustable pressure: 1,7 –2,6
• System pressure control with single stage valve.
• Optional MAP connection. (for turbo charged engines).
• Very Stable dynamic pressure behaviour compared to other single stage reducers.
• Lock off valve.
• Integrated temperature sensor.

robertXX wrote:no pressure aor gas temp input

Hint; every commercial gas injection system ever made, however cheap and nasty, has a gas temperature sensor. There'll be a reason.

The early ones did get away without a pressure sensor, relying on regulator stability. They lose something on transient throttle response because of that.

Commercial systems use these sensors to compensate for variable gas density. Liquid systems (like petrol, e.g. megasquirt) don't need/have such subtleties built in.

If you try to get away without pressure sensing, you must reference the pressure regulation to MAP - which isn't normally done with Prins vaps, but you can modify it.

robertXX wrote:but i suppose the pressure and gas temp are pretty relative to the water temp in the vap ?

Nope. Gas temperature (and hence density) varies by recent (not current) demand, it would be difficult to guess.

Take note that the gas injectors are peak-n-hold type, you must use a version of MS or driver add-on that copes.

Note that it may be difficult to get certification (if you want it) if you use a non-E67R marked gas ECU.