Honda Cr-v project

[Dave Johns]

Looks like you are doing a nice job and I've only got a couple of comments. You might want to consider some sort of grommet where the polypipe enters the alloy tube so it can't be chaffed on the inside edges of the tube. The other point regards getting it inspected and certified if you need to. CoP11 states that the feed pipe should be protected against damage, which running it inside alloy tube does very nicely, however, it also says that it should be visible along its length so it can be inspected which means it would fail with the pipe run inside alloy tube. I have always thought the two requirements contradict each other but it obviously wasn't written by an engineer.....

Thanks for the feedback exactly what's needed. And arghhh! I have been in two minds as to whether to remove the pipe and just tap to the chassis and will change it, but I think I'd like to leave it until I've commisioned the system. So that will be changed. The car is insured for LPG ( no extra cost and no charge for the change either ) so I'll commision it, shake it down and then make changes.

Thanks

[Dave Johns]

OK this is driving me crackers.
Cylinders longtidudal number left to right 1 2 3 4. Picture on AEB shows the injector rail facing the viewer ( and I suppose the front of the car) with nozzles to the front. Yep all good. It then labels the injector electrical connections as A B C D Right to Left! So A corresponds with cylinder 4 and D cylinder 1 etc. I'm quite happy to follow this but it seems arse back'ards to me.

I'm considering buying a secondhand injector loom for the car and adding this to the ecu Loom. This gives me any easy reversal should things go awry. Considering....

[Dave Johns]

So the injector rail and valves into the manifold done. I've also plumbed in a Map take point and the valve saver spigot. Well nearly the valve saver spigot which would've been plumbed in if I hadn't lost it . Currently awaiting another and the hole has a machine screw in it.. Ho Hum.
https://drive.google.com/drive/folders/ ... sp=sharing

Hated drilling and tapping that manifold but all done now. Within 50mm of the petrol injectors so all good. Guessing the soldering iron will be out next.
Thanks for your feedback it really does help!

[LPGC]

It is A=1 B2 C3 D4.

[Dave Johns]

It is A=1 B2 C3 D4.

Thanks that makes sense to me!

[LPGC]

I haven't looked at the wiring diagram in a long time (I know what they're all for anyway) but it may be pictured in a way in which it could seem A=4 because back in the early days when a lot of installers did use plug-in petrol injector break looms the injector break looms had shortest connector at D and longest at A, that was because installers usually ran wiring from the back of a rwd orientation engine to keep wiring relatively neat/short and D would be the rearmost cylinder on a 4 cyl or V8 engine.

[Dave Johns]

I haven't looked at the wiring diagram in a long time (I know what they're all for anyway) but it may be pictured in a way in which it could seem A=4 because back in the early days when a lot of installers did use plug-in petrol injector break looms the injector break looms had shortest connector at D and longest at A, that was because installers usually ran wiring from the back of a rwd orientation engine to keep wiring relatively neat/short and D would be the rearmost cylinder on a 4 cyl or V8 engine.

Thanks, I'm not going mad then. I've checked the diagram a few times (research/homework!) and would still read it wrong! Must admit this is quite an involved project with very little gain for a mountain of work at the moment, lots of stages and different skill sets + knowledge required. Still trickle on and triple check it all as I go. At least the wiring can be broken down into small sub stages!

[Dave Johns]

Question
If I splice the gas injectors in does the ECU need to be live for them still to work? I'd guess not (bit like a closed relay with the circuit 'made' until...) and I hope not. If the ECU can remain disconnected until all is complete it would make things easier.

[LPGC]

No, the ECU doesn't need to be live, most do contain a set of double pole relays.

Edit - the above should read double throw. I always get mixed up between pole and throw but I know what I mean

Obviously the loom has to be plugged into the ECU though or the petrol injectors will be disconnected.

[Dave Johns]

No, the ECU doesn't need to be live, most do contain a set of double pole relays.

Obviously the loom has to be plugged into the ECU though or the petrol injectors will be disconnected.

Thanks, and yes obviously. Although I can see me splicing happily and forgetting that and panicking Oh yes.
Thanks again.

[Dave Johns]

Bolted the ecu in today, added the switch and the LED warning light to the dash. Usual headache of threading and access but its gone quite tidy. Pleased with the ECU position were it tucks nicely under the main fusebox. Switch and LED fit in with the other bank odf switches very well. Next up is the wiring to the tank and I'm waiting on some 5 core to be delivered. I shall wire in the ECU and leave the fuse out for now..
.https://drive.google.com/drive/folders/ ... sp=sharing
Also have to pull the intake (plastic section) back out and put the valve saver nozzle in as I lost the valve and had to get another one. Machine screw in there at the moment.

[Dave Johns]

Valve saver nozzle in, ecu wired in. Injector lines cut and wired .
Did not enjoy cutting those injector lines...
But all done and running, no fuse in the ecu at the moment, I'm waiting on a new sender and some cable and then wire that in....https://drive.google.com/drive/folders/ ... sp=sharing
Usual story, doing the job itself (which in this case was identifying the injector negative and cutting and soldering 8 wires) about an hour and a half. Getting ready dismantling and tidying the job up about 2 hours.

Read around about the Lambda hook up and it seems unnecessary and comments on here support that - sorry I'm a bugger for double checking. But, should I hook up to the obd2? It looks like a simple job, buy a male connector solder to the relevant ports and jobs a good 'un. Probably an hour or two. What do I gain?

Editted, previous comments are saying only if you need to. Sooo, unsure.

[Dave Johns]

Well that was a fun filled day . I've spent a few hours wiring everything in and today it was of we (didn't) go.

Have to say Mr Bird at Tinley was a Superstar!

Firstly the bloody thing wouldn't talk to the ECU. Code CO1 but then...we tried another socket on my swiss army knife lead and got c04 and then a c03 - what Joy! After Mr Bird sent me some updated drivers and software we got connected Bingo. And then, I couldn't get any revs so nowt moved, the setting 'weak' spark sorted that again (TT THANKS) Lambda was fine. OBD ON PIN 7 was an exclamation mark. butr everything else was working. Changed to gas, I had gas pressure and no smell of gas. Coupla stalls but hey ho. But oh yeah gotta be a but, those gas injectors were'ny registering which is a bugger. I had a full tank of gas and I could hear the rear solenoid ping. But still no injector movement. The switch was registering a full tank, and the gas light flashed, petrol stayed orange. Cuold it be that the front solenoid (integral to the reducer) isn't opening? This front one has a line feed from the ecu the rear also has a line feed from the ECU. Am I missing a software setting,? Put it down for a while, close but no cigar!

[LPGC]

Could be that the front solenoid isn't opening. Have a squeeze of the pipe between reducer and injectors (does it feel like there's pressure in there now?), or read gas pressure in software, or disconnect said pipe to see if any gas comes out. You didn't turn off the gas at the tank using the manual/service valve thumbscrew did you (it needs to be open which is anticlockwise assuming it's a Tomasetto valve)?

Try holding the engine at around 3000rpm before pressing the button to switch to gas. On a new install all the gas lines are initially full of air so even if gas injectors pulse the engine might be getting air+air instead of air+fuel in which case the engine may stall before it gets fuel.

[Dave Johns]

Could be that the front solenoid isn't opening. Have a squeeze of the pipe between reducer and injectors (does it feel like there's pressure in there now?), or read gas pressure in software, or disconnect said pipe to see if any gas comes out. You didn't turn off the gas at the tank using the manual/service valve thumbscrew did you (it needs to be open which is anticlockwise assuming it's a Tomasetto valve)?

Try holding the engine at around 3000rpm before pressing the button to switch to gas. On a new install all the gas lines are initially full of air so even if gas injectors pulse the engine might be getting air+air instead of air+fuel in which case the engine may stall before it gets fuel.

Will give that all a go tomorrow. Really nice to have the support from people, helps one stay sane!
Thank you

[Dave Johns]

Another day down the treacle mines. Managed to get the maps acquired, huzzah! Checked the nozzles for the smell of gas and finally drilled them ..yep hadn't done that because I thought I could do it through gas pressure. Went for 2.5mm which is probably a tad small. After that it was all steady plodding and readING and checking. Acquired a basic map. Acquired a petrol map and then three maps for gas, but at what point do I stop aquiring? And switch off acquire and back to obd on...questions questions. No gas leaks! Checked with a gas sniffer so all good there.

[LPGC]

You can compensate a bit for slightly small / large nozzles with pressure, the extent of which depends on model of injectors and reducer that are fitted. But nozzle size is generally more critical than pressure because when injectors are fully open (and that's not clear cut because at low pulse duration LPG injectors spend relatively high proportion of active time opening and closing rather than at full extent of open... while pressure and temperature can effect both opening and closing speeds) flow rate of the injector (for a given pressure) is relative to the square of the nozzle size. While for any given nozzle size to compare flow rates for different pressures we use the square root of the pressures (not forgetting that pressure can affect opening and closing speeds).

So for example if we go from a 2mm nozzle (radius 1mm, square of 1 is 1) to a 2.5mm nozzle (radius 1.25mm, square of 1.25 is 1.562) then we've only increased nozzle size by 25%. Then if we do 1.562 / 1, we find the 2.5mm nozzle will flow 1.562 times as much gas as the 2mm nozzle, a 56% increase in flow for a 25% increase in nozzle size.
If we go from 1 bar pressure (square root of 1 is 1) to 1.5 bar pressure (square root of 1.5 is 1.225), an increase of 50% pressure, we find that the 1.5 bar will only flow 1.225 times as much gas as 1 bar (22.5% increase in flow).

The higher pressure can reduce opening speed (and can affect closing speed). The active component that closes most injectors is a little spring but when an injector's plunger is sealed against it's output port the difference in gas pressure and manifold pressure effectively adds to the spring pressure. During closing as the plunger nears the output port gas pressure and gas flowing between the plunger and output port can either work with or against spring pressure depending on the conditions and type of injectors.

So we can compensate nozzle size against pressure and vice versa to some degree but there are physical limits to how far we can take that. Not least, at high enough pressure the gas won't even be a vapour anymore, it will be a liquid. But usually way before we'd get to the problem of injectors being fed with liquid instead of vapour we'd get the problem of injectors not being able to open due to gas pressure adding to spring pressure and keeping the plunger closed... Big problem if the injectors don't even open. Or gas pressure affecting the injectors performance so much that what could be linear and accurate injectors at lower pressures become unlinear and inaccurate injectors at the higher pressure, especially at short pulse durations.

If we choose 2 different combinations of nozzle size and pressure, both which should flow the same, the resultant gas map will be different between setups. The one with the smallest nozzle and highest pressure will usually need the highest multiplier at low petrol injector duration but they should both have the same multiplier at the longest petrol injector pulse durations, which also implies the one with the biggest nozzle should have the flattest map - but not necessarily depending on the type of injectors and the engine they are fitted on. We should always aim for gas injector pulse length to be longer than petrol injector pulse length at short petrol injector pulse lengths (or we will almost certainly loose some fuelling accuracy) but gas injector pulse length cannot much exceed petrol injector pulse length at full engine load or we might risk 'running out of window' which is where physical (not electronic) overall gas injector pulse length (accounting for the injector's opening time and closing time) exceeds the length of time it takes for the engine to complete a 4 stroke cycle. On some engines window can be shorter than the 4 stroke cycle if good results are to be had (e.g. pseudo Atkinson cycle engines.... we don't want gas to be injected a bit late at a time when the Atkinson cycle means cylinder charge is being blown back out of a cylinder into the intake manifold).

If we've fitted injectors that are linear from short pulse durations but are a bit slower at higher pressures we might think it better to have the big nozzles and lower pressure, or if we fitted injectors that are not so linear from short pulse durations but are less effected by higher pressures we might think it better to have the smaller nozzles and higher pressure. But there are other things to consider too... For starters we could have fitted faster more linear injectors in the first place (and some injectors are fast/linear and accurate). After considering this it gets a bit complicated but suffice to say that if you have a simple/oldskool engine then you'll get away with just about any injectors capable of flowing enough gas at peak load for the engine (if you want it to be able to run at peak load on gas, some installers don't even care if the car can run on LPG at full tilt or at the other extreme if it can idle at bottom tilt i.e. minimum engine load). But with some more modern/advanced engines, turbo'd/supercharged, VVT, Atkinson cycle, no petrol return, etc, for best results you need to know not only whether that high pressure smaller nozzle or lower pressure bigger nozzle would be best but also which make/model injectors (which in turn feeds back to pressure and nozzle size), where to fit spuds, how long to have pipe-lengths, how to set up compensation for things like pressure, temperatures, petrol pressure, which ECU(s) would work best, etc. Get it running on gas anyway with almost any components but the finer aspects at least add the polish to the job and at worst make the difference between an install which might seem to give decent cruise ability seat of the pants but falls down during changing driving conditions and/or at various engine loads and/or has to run on petrol for a longer warm-up period before switching to gas and/or doesn't feel quite as crisp to drive seat of the pants and/or is incapable of seeing fuel trims running on gas as consistent as they are running on petrol (which could also mean varying mixture under open loop conditions etc) and/or gives poor mixture (too rich or lean) at higher engine loads.

You've got an LPGTech ECU - Did you set it up in Tech mode or Standard mode? The 2 modes are very different, actually you get a lot more control in standard mode than in tech mode (and in my book tech mode should be called standard or basic mode while standard mode should be called tech or advanced mode) but if you ever switch between tech mode and standard mode you need to recalibrate from scratch. There was a time not so long ago when a few firms liked the (what they seemed to assume) 'fit and forget' tech mode... done in a day installers. I sorted hundreds of LPGTech equipped vehicles such installers converted, they were nearly all set up in tech mode, I could see why that had caused problems and I changed them to stanadard mode and recalibrated from scratch.... Sometimes also changing components such as injectors, spud positions, changing pressure, etc.

I never connect an LPG ECU to OBD... If the system comprises correct parts, fitted in correct positions, setup and calibrated properly there is no need and results can be better than if connected to OBD. On some vehicles (most notably Mercs) connecting to OBD can mess up all sorts of things... Can even result in an engine that won't even crank if you turn the key, and/or harsh automatic gearbox changes, almost at the very least it can prevent connection using an OBD scan tool in future. But if you don't have an OBD scan tool that will show live data you might want to connect the LPG ECU to OBD so you can read fuel trim data... Just don't activate any LPG ECU functionality that will see it respond to readings.

[Dave Johns]

Now that's a lot to take in at one swallow and shows an expert point of view. I'm using an AEB ecu which seems to have adapted quite well. Acquiring the maps seemed to be acheived reasonably easily the problem was my unfamiliarity with the sofware and a lack of faith that maps have been acquired and everything is running as it should. I can understand your point of view re the use of the OBD my memory is getting on a bit but I seem to recall that I set up Megasquirt on modified vehicles using a Lambda feedback, the first time around was a supercharged Punto and that was very reliable (and bloody fast!). No OBD connection needed. However, if OBD can provide a further feedback loop is it not useful? I'm talking about a tow car that will be encounter unusual loads but then maybe not. Currently my slow scan K is giving me a 6 percent + correction. Enjoyed reading through the technical side of injector choice balanced with gas pressure. Quite interesting to read that and at the end of the day we are just using a drill bit based on a guesstimate index!

The whole of my system seems to be working appropriately at the moment, I have to change the gas line set up underneath at some point but that's OK as I'm working with a system that now works and I'm familiar with. Makes a difference.
I'm suprised at the work involved and glad I completed it as it keeps my grey cells up to speed. As for fitting that lot in a day or even the day and half I was quoted - rather them than me! I do think it is a challenging project first time round. Would I do it again, Oh yes but the criteria for this are quite narrow. I have a real life 60mpg deisel MPV (useless tow car) which is a 'better' choice I'd guess in the real world. I'd like to build a super economy car at some point. LPG Citroen C1? But then I could buy a Volkswagen e-up for an economy car.
Anyway, thanks for everyone's comments and feedback it really did help . Cheers.

[LPGC]

Enjoying your posts and insights too.

Would you mind linking to the AEB ECU and software that (jointly) 'learn maps' for petrol and LPG? I had assumed you'd opted for LPG Tech because it was mentioned and because it does in a way 'learn maps' by plotting injector duration (on Y axis) against manifold pressure (on X axis) on a graph, a graph which will have plotted points at different heights on the X axis because engine breathing for any given manifold pressure is also dependant on volumetric efficiency which itself changes with rpm and depends on the cam / valve arrangement.

There's a big difference between LPG injectors (which inject a vapour / gas) and petrol injectors (which inject liquid). Petrol injector response is relatively immediate and the density and pressure of fuel doesn't usually need to be a consideration factor, while LPG injectors injecting vapour need to open much further/wider than petrol injectors to inject the equivalent mass of fuel (liquid denser than gas and liquid is at higher pressure than gas) and the extra mechanical distance they have to open/move means a far greater opening/closing time on top of the far wider density fluctuations due to pressure and temperature changes.

I mentioned a few potential problems with OBD connection above. Even if none of those potential problems is realised connecting to OBD still wouldn't be a great idea.. Watch the fuel trim ranges change on an OBD live data reading tool as you're driving around and it becomes apparent that the LPG ECU doesn't know when conditions have changed and a different set of LTFT's becomes active (or what the conditions were that caused the switch to a different LTFT). So say you're cruising at 50mph in top gear at 2000rpm with LTFT at +5% and STFT fluctuating around zero%... Well you could enable OBD adaptivity in the LPG ECU to steer that LTFT down to zero%, or you may even be able to set it up so that at 2000rpm at (say) 6ms or (say) 0.6 bar map to steer LTFT to the same +5% that it's usually at when running on petrol. But if you move the throttle or change gear, something that's almost bound to cause the petrol ECU to switch to another underlying LTFT, the LPG ECU doesn't change the adaption at the same time. If you had set the system to steer trims to 0% then when running with +5% LTFT the LPG ECU might add -5% to fuelling to compensate but then when you move the throttle the new LTFT might not be +5% while the LPG adaption is still -5%, so during this crossover period the fuelling on LPG can be further away from correct than would be if the OBD adaption were not enabled. And then we can think that if only we had set the LPG system up properly in the first place the fuelling would be closer to exactly correct more of the time than it would be when we had the OBD connected,

[Dave Johns]

I can certtainly pop up some screen shots of the adaptivity screen , I took a few pics when static in order to be able check things at home if needed - and when I was less stressed at working on a new system with 48 ltrs of gas in the boot! Even though I'd used a sniffer I was still edgy!

I'm listening to what you say about the obd and can see situations where it may change things and not for the better. Although I have to say I'm stretching my thought patterns and knowledge! Back to school for me .

Your second paragraph is a given, from my days of study I'd have thought vapourised LPG would be more manageble to inject consistently? I think that's what you are saying. However it will bring problems of its own. Mind you liquid (especially refined petrol is a given nowadays.

When I've finished sitting grandchildren over the next two days I'll hook up the computer and link more screenshots. If I can video the acquistion phase I will.

https://drive.google.com/drive/folders/ ... sp=sharing