lpg feed pipe size on 500 bhp engine.

[robertXX]

well one major advantage to having two engines ..... you can fire it up on one.... and just keep changing the fueling on the front engine until it chimes in :)



now running on two , front engine needed 20% more fuel .

mot tomorrow.

[robertXX]

mot passed , even got 2.6 co and 250 ppm hc . one weird thing is , the car now starts on the rear engine ,then the magic jets fire up 5 seconds late , i wonder if the mj's need a full voltage to open ...which only starts once the alternator is charging ?

[robertXX]

WELL , i had a tiny window of opportunity to get the monster onto the dyno last friday .


i have not been able to set the fuelling very well at higher rpm , basically because its so bloody fast at accelerating ...being under a ton and over 350 lbs ft of trq , and big tall wobbly tyres ,means i keep ploughing into corners way too fast ...or going well over my winnebago 7.50x 16 tyres speed limits , i dont mind going over it on the dyno cos i can always run away.


so strapped her down ,and did a few minutes working on the low down tickover and below fueling , so it has a lot more low down trq for trickling in traffic , made a little progress ,still too rich , but its hard to hold steady state over 300 lbs ft at below 1500 rpm ...more work to do on that area it kept falling in a big rich hole at higher throttle opening .


then onto a couple of full throttle runs , with the ecu set on auto tune , so it would set the fuelling for me ,after a fashion .

now judging by these piccys ...



30 mph light throttle..







higher rpm full chat max trq..





i may have just found the trq limit of my rollers with little narrow tyres and no loading !

i am not sure if it shows very well ,but there was a churn load of blue smoke coming off the tyres , also you can see the grated tyre rubber being thrown across the cave as my gritted rollers hone down the treads a touch .


however i did manage to get some readings..



bhp is now up to 354..







that's a graph of my best venturi run with the 2 spinning injection runs , and my theoretical max with grip ...i reckon at around 380 bhp.




then torque at 419 lbs/ft ..







here i reckon on around 435 lbs/ft max theoretical ..however my 419 is still 58 lbs/ft up on the old fueling system...noice


inj run number 2 had more ignition advance , 1 degree more made a fair bit of difference .


now i am aiming for around 13.5:1 air fuel ratio , here you can see the fuelling logs for both runs on both engines ...

engine 1 , the 3.6 litres ..


run 1






run 2







engine 2, the 4.0 litre..




run 1





run 2









you can see how the autotune has modded the fuelling , sometimes over shooting and going too far the other way ,also i am well over 100% duty cycle with these injectors at this 15 psi fuel pressure ,which is a surprise ...and basically means they are wide open non stop ,hence the leanness in the places at the higher rpms .




also these power runs where done on a low load so the tyres had a chance to grip ,so the ecu only had a few seconds to adjust the fuelling before it was run over ...so i am pretty impressed the ecu was able to take a car that wouldn't rev ,to one that would , in just a few pre run rev ups .



also it revved to 6k with little vibration


i think with better fueling and timing i should reach 400 bhp and 500 lbs/ft ....but maybe nor be able to grip it lol. .... for all i know its there already ...it does explain why a blip of throttle in 3rd at 40 lights up the tyres like a volume control.





ok ok ,geekness over



regards
robert.

[robertXX]

running at 15 psi on both engines i got this ...

engine 1...




engine 2 ..









bit more fuel pressure needed ..

[LPGC]

Great write up as usual Rob!

Does fuel pressure stay at 15psi under load?

How does the ECU calculate dc (is it just straight electronic pulse time or does it take into account injector opening and closing time)?

You drilled nozzles to 2.9, 3.0 nozzles should flow up to about 7% more gas. I know the drill bit will leave a bit of clearance but when I drill a 3mm nozzle I mean using a 3mm bit.

Figures suggest that just a bit more pressure (with or even without drilling injectors to 3mm) should prevent having to run injectors full open (but this might depend on how dc is calculated). Better not to get too close to approaching electronic 100% dc, once gas injectors get to the point they don't get time to fully close before having to open again linearity and accuracy of fuelling suffers. Very soon after this point (pulse the injectors just a bit longer) the injectors will be fully open, at this point there's usually quite a hike in the quantity of gas delivered because the injector is fully flowing all the time when usually (when metering gas accurately) some of the time it is opening and closing.

If the injectors are full open the mixture will be richest when airflow is lowest, so for example if you're going up the rev range and hit the point injectors are fully open the mixture will suddenly become rich just at that point but then if you continue increasing the engine rpm (if the engine is still on cam and drawing even more air) the mixture will start going a bit leaner again. The amount of hike is in a way inverse to the speed of opening and closing of the injectors so there's usually a bigger hike with poor injectors than there is with good injectors at the point they start not having enough time to close before having to open again. You get twice the hike when running group injection than with sequential injection because when the injectors are metering normally (not full open) there are twice as many injection cycles.

[robertXX]

i don't monitor the gas pressure in the data Simon , so no idea...

i do know i was getting a higher rpm map signal drop to 97.2 kpa . ...which with the reducer vac connected will drop the gas pressure , however , whether there is another restriction in the line from the tank to the rail causing a high load /rpm reduction in flow i don't know , i must rig up a camera to watch those pressures on a run .

i can drill out the stubs to 3mm , i was wondering how to work out what amount of gas pressure increase would work ?the reason i am running 15 psi is that with the inj at 24 psi the car would start on the rear engine ,then the front inj would start opening after the batt volts rose with the alternator working . dropping the pressure reduced the volts needed to open the magic jets , and so the car would fire up on both engines together .. i suppose it may be that ,with a bit of use the MJ's will open with a lower voltage ?


i appear to be hitting 100% dc at 4500 rpm odd , with that engine , with its 265 Degree inlet cam it may well be making a touch more trq , above 4500.


i think the dc is simply pulse length .

as you say the inj open all the time means as rpm's rise the engnie gets leaner you can see this on the afr logs where it goes to 15.4:1 at 5500 rpm.


regards
robert

[LPGC]

Thought I'd replied last, sorry!

Read back a few pages to try to find what reducers you have fitted and how you plumbed them in but didn't find that info... I know you have the Prins reducer fuelling one of the engines (which one?), what make/model reducer is driving the other engine? Or are both reducer outlets plumbed together?

These injectors will usually open when cold with 12v peak pulse at 1.6 bar (24/14.7), they will open under same conditions but with quite a bit more pressure when warmed up a bit. On some installs (where I need to run high pressure) I set the ECU to warm up injectors before switching to gas (the injectors are pulsed before gas solenoids are opened so the injectors are electrically heated), I realise this isn't an option on your install but I wouldn't expect you to need it. One set of injectors opening while the other don't suggests either they're fed at different pressure (my questions above related to this) or something going on with the ECU driver stage.. If there's a problem with injectors not opening when cold is it ECU injector driver stage related? Do you mean literally not opening or perhaps opening but performance so far effected that fuelling is drastically wrong? Like you've said, bigger nozzles could offset the need for more pressure.

If battery voltage falling during cranking is causing a problem you could fit a second (smaller) battery that powers the LPG system (and why not the ignition too), the second battery automatically connected in parallel with the main battery when main battery voltage rises above 13v... the kind of voltage sensing relay circuit often fitted with dual towbar electrics (12s socket) to prevent running 12v lights or fridge in a caravan from flattening the vehicle battery when the alternator isn't running but allow the car to charge the caravan battery. Seems an easy way to see injectors etc get 12v when cranking the 2 x engines

[robertXX]

hi simon ,

i have had it on the dyno gain , now at 21 psi front inj gas pressure and 20 psi rear inj gas pressure , it now wheelspins too much , but i have read while spinning 423 lbs/ft and 354 bhp .

both regulators are prins , ones a older vsi red on the front engine, and a newer type vsi on the rear .



my duty cycle is down to 119 with a rich mix , so getting there .


re the starting . if i run 15 psi on the front MJ injectors , they will open on a cold start . if i run 21 psi , they won't .on a hot start they open on cranking .

it is noticeable that , as the car runs up when cold as the voltage comes up with the alternator kicking in the inj then open . bear in mind my cranking voltage is 9v .very low.


with the old injectors brc i did not have this problem all the way up to 25 psi gas pr .

they are definitely not opening at all , if i switch on the gas and pressurise the rail , then turn off the gas , and crank , the rear engine goes to 0 psi rail , the front stays stubbornly on 21 psi .


a new bit of data.... i just did an experiment ..

with the megasquirt type of injector control , called pulse width modulation PWM , you can adjust four things .

injector dead time ..Ms

PWM current limit ..%

PWM time threshold...ms

injector PWM period .uS




now , the rear engine injectors , the 5 OHM matrix ones , i can run :

dead time 1ms
limit....30%
threshold .3ms
period 66 uS


with the front engine 2 ohm MJ's.


it would not run right on those settings and needed the current limit up to 40 % with a warm engine.



so it looks like the MJ's ,although they have less resistance ,need a higher voltage and it applied for longer to open .


WEIRD !



regards
robert


now up to 12.1 mpg !

[robertXX]

i wonder if the majic jets are polarity sensitive ? i did not take note to connect them any specific way....

[LPGC]

Can't see how they could be polarity sensitive.

Strange why one set of injectors will open while the other doesn't if temperature of injectors and pressure to injectors is the same on all injectors and electrical pulse is the same on all injectors. Are you reading the pressure to injectors on the front and rear engines constantly? Could pressure to rear injectors be higher than to front injectors when the reducers are cold?

I think I mentioned that some LPG system ECUs are capable of pulsing injectors before switching to gas, while the engine is running on petrol and before the gas solenoids open, thus pre-heating the injectors. The pulsing starts off slowly so as not to richen fuelling on petrol but gets quicker when gas pressure has fallen away (injectors pulsing while gas solenoids are closed)... Not really an option if you're running monofuel! But a separate battery that could supply 12v even during cold cranking could be an option.

Is the MJ injector output stage up to driving 2Ohm injectors? Thinking most petrol injectors are much higher resistance.

[robertXX]

i am reading the pressure off the guages simon , cold front engine reading 15 psi on the front inj and they open fine , reading 21 psi and they delay till the alternator brings up the voltage. , with the rear engine ..matrix inj they open whatever the pressure, engine temp, no problem .

the mj is using pwm with the numbers shown above in a previous post ....they inj do work fine once openning .

regards
robert

[LPGC]

Does the peak and hold compensate for voltage?

[robertXX]

there is this setting Simon ..


"Battery Voltage Correction (ms/V) (BatFac) is the number of milliseconds that MegaSquirt-II adds to each fuel injection pulse to compensate for the slower opening of the injectors with lower supply voltages. Generally 0.10 ms/V or 0.2 ms/V is about right. So, with a 0.20 Battery voltage correction factor and supply voltage of 14.5 Volts, a 1.0 millisecond 'opening time' is adjusted to 1.0 - (14.5-13.2)*0.20 = 1.0 - 0.26 = 0.7 milliseconds."

[robertXX]

https://youtu.be/qg5o0_tJxYM

long winded vid of my startup and dyno research .

[LPGC]

Still to watch the vid Rob, looking forward to it

[robertXX]

simon , we were talking about nozzles the other page , and you were saying you have to have the nozzle at the injector , if i run with no nozzle i wont get good flow control ..reading this ,

https://www.lpgshop.co.uk/dymco-single- ... rail-type/


does that not say no nozzle on one of the options ?

regards
robert

[LPGC]

Hi Rob,

Without re-capping on what I said in an earlier post, the metering nozzle does have to be at the injector - Suppose an application needs a 2mm metering nozzle, if we fit the 2mm nozzle at the injector and fit a 4mm nozzle in the manifold, the manifold nozzle won't be any restriction on the metered fuel from the injector and the fuel from any duration injector pulse will quickly flow from the injector through the pipe and into the manifold, this is the correct way of doing it. If we try fitting a 4mm nozzle in the injector and a 2mm nozzle in the manifold, instead what happens is that for any given injector pulse duration the injector delivers 4 times as much fuel into the injector pipe (the difference between PI Rsquared for the 2mm nozzle to PI R squared for the 4mm nozzle) and the restriction at the manifold end of the pipe will only mean the fuel is delivered slower rate over a longer pulse. The latter situation would ultimately for a long duration pulse only flow as much as the former (correct) situation but the latter would mess up fuelling by making the mixture far more difficult to accurately control at low to medium injector pulse duration. There is very little volume in an injector between the opening that the piston(plunger) closes against and the metering nozzle but there is a lot more volume (in the pipe) between the end of the injector metering nozzle and the manifold nozzle. On none demanding applications it can be possible to get away with fitting too large nozzles on injectors by fitting restrictive nozzles in the pipe close to injectors but it becomes less possible as the restrictive nozzles are moved further away from injectors and closer to the manifold.

I've had a look at the injectors your linked to, I'm not overly familiar with this type injectors but I expect Silver/White correspond to a mechanical difference (so effectively slightly different spec injectors irrespective of nozzle size). You're probably talking about Silver with 3.8mm nozzles rated at 33-47hp each while White with 3.2mm nozzles are rated at 40-60hp each? To me this could mean the White will open and work linearly at higher pressure than the Silver or, more likely, that the White has a longer piston(plunger) stroke and/or diameter piston(plunger). Bit late for me to be doing the maths but they'd only be fairly complicated rather than very complicated and I'm sure you'll know the gist of what I mean - if you've got a wide hole for gas to flow through but only lift the lid on the hole a bit it's flow will be restricted compared to if you lifted the lid further, or if you've got an even wider hole for gas to flow through and have the same lift of the lid it will flow more.

In the context of this item, I expect '3.8mm (without nozzle)' to equate to 'effectively no nozzle', i.e. the internal diameter of the hole that the injector piston(plunger) covers is no greater than 3.8mm and the nozzle is 3.8mm. Since 'White big flow' is rated at higher hp than 'Silver 3.8mm (without nozzle)' I expect the White to have higher piston(plunger) lift than the Silver. Again it's a bit late for me to be doing the math but I reckon you'll understand?

I have fitted injectors without nozzles on some of my big power installs. The thing is, nozzles on injectors are quite an equaliser in case injectors have slightly different specs (which may be due to manufacturing tolerances and/or slight differences in design over the years)... it's far more important than it usually is to make sure injectors all flow the same if no nozzles (or nozzles equally as wide or wider than the diameter of the hole in the injector that the piston[plunger] covers) are fitted because injector spec (especially hole diameter and piston stroke and especially open and close responses) become the only factor involved in the metering.

Still haven't watched your December vid all the way through, watched around half of it back in December before having to break off... I've been meaning to watch it from start to finish but there's never seemed a great time. And to be frank, the first half I've watched hasn't been as entertaining as some of your other vids!

[robertXX]

yes ,video second half is more exciting to be sure . lol , i suppose ,with my ability to change the inj opening with a custom map , rather than a ecu taking its signal from a oem ecu , i have more lee way to make a nozzle less system work ...at least after a fashion ..

[LPGC]

I'll be sure to watch the vid!

A lot of modern 'slave' type systems feature individual channel offsets (e.g. plus/minus a set number of 10ths of a millisecond) and individual channel percentage pulse length adjustment but such adjustments apply across the whole map; While to totally correct for effectively different spec injectors on the same engine each channel would need to be mapped individually, which would be quite complicated considering there isn't a lambda sensor per cylinder and on the monofuel engine there's no way of switching fuels on individual cylinders to see how it effects overall mixture.

Should be able to map it well with nozzles even running group injection (which I believe you're running?).. Where are you at now with nozzles and pressure etc? Is a pressure sensor fitted and is the ECU compensating for pressure?

[robertXX]

still the same setup simon

,on the front engine batch fire old type MJ...2 ohm.. 3mm nozzles , gas pr at 17 so it runs sooner after start , lean above 4900 ,but i never get there driving it .no pressure sensor ,just gas temp like i do on the astra ,

back engine is rich enough at top , matrix inj ..5 ohm , always fire up from cold . ,don't know what size nozzles in there of even if there are any ! if not it's using the 3.8mm stubs at inlets .