Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Lazydocker]

ok, set to 1.1 and recalibrated, certainly massively improved. Tick over seems smoother, and on a full tank of gas the gear changes seem fine on a 5 minute blast. Sudden lift off from full throttle is far better but not quite as smooth as on petrol, acceptable though.

Will see how it goes as the tank empties.

Still switching back though, no change on that score.

Thanks LD

Under hard acceleration? If so it sounds like the limits of the multivalve have been found

You may only have a 6mm pick up pipe which can be changed to 8mm, which would raise the limits of flow a little but is very fiddly. Is it still only doing it once you have used some of the fuel from the tank and OK when first filled? If so it does all point towards the flow rate from the multivalve

sounds likely from what your saying, its far less likely to  switch back with a full tank, and gets peogressivly worse until empty...?

Yep... ISTRC that this was what you described recently :-/ :-/

[TheBoy]

Teilo has a supplier for a higher flow multivalve, but their is no stock anywhere.

It may be something I will consider in a few months if I find its constantly switching to petrol at high revs.

[Ken T]

Could I add a slight word of caution. The manpower required to design/test/ production engineer/write software/ source all parts for an ECU is quite large, you would need quite a large team to do it properly/well/reliably. And if you look at the price they are being sold for, ( a Stag front end kit for £140 ish ) there isn't much profit in there, not enough to fund all that R & D. It might be reliable, but if its fitted like the OMVL N stuff, it mimicks the Petrol ECU and controls the petrol injectors, so if it dies, so will the car. We're talking safety critical here.

The Stag stuff must be bad/ not enough profit  for him to change supplier.

Ken

I disagree.

Complex its not, and its only a basic microcontroller which many serious hobbiests can handle. Add some testing in etc and its the sort of thing a single compitent engineer could do in sub 12 months without to much of a challenge.

Then chuck it at an outsource company (such as Flextronics, Sanmena, Solectron, Jabil etc.....) and they will source all parts, build and test in return for some VA markup.

I would guess, in 1-10K quantities, that it would come in at around 70-80 quid.....and if you got into the millions at sub 30!

Yes, and design and produce a mould for a zinc diecasting. I'm not saying it can't be done, I'm just saying he will need to shift a lot of units to repay development.

Re the LPG ECU, well its in series with the petrol injector lead, and any element in series can fail. It would be interesting to see how they arrange to "fail safe ".  Subcontract manufacturing, well the place I am currently at, use a subcontractor to source parts and build stuff. Problems I had the other day were an IC the wrong way round (16Pin DIL), tombstone resistor and the soldering is way up the DIL ic legs. Oh and they are IPC certified. 

Interestingly enough, remember the Lexus that stuck on full throttle, and the Toyota sticking throttle recall  ?. Well a recent article on the EMC website 'suggests' that the Lexus problem might have been caused by the ECU deciding to go on full throttle, and there's basically nothing you can do about it. Car brakes can't absorb several hundred HP, and with electronic controls there is no way of cutting power to the ECU, so no way of stopping. And the Toyota pedal manufacture said that there was no evidence of their assemblies sticking. I wonder if Toyota ECU's occasionally stop monitoring the pedal. Not a nice thought.

And you would let some single engineer write code to control your engine ?. The amount of testing GM etc must do to ensure fail safe code must cost millions.

Ken

[TheBoy]

Thought in Toyotas case, it was pedal physically stick in the mat?  LOL, used to happen on your crappy vans at work

[Kevin Wood]

Interestingly enough, remember the Lexus that stuck on full throttle, and the Toyota sticking throttle recall  ?. Well a recent article on the EMC website 'suggests' that the Lexus problem might have been caused by the ECU deciding to go on full throttle, and there's basically nothing you can do about it. Car brakes can't absorb several hundred HP, and with electronic controls there is no way of cutting power to the ECU, so no way of stopping. And the Toyota pedal manufacture said that there was no evidence of their assemblies sticking. I wonder if Toyota ECU's occasionally stop monitoring the pedal. Not a nice thought.

If these rumours are true then the Toyota ECU is a truly sh1te design. There are plenty of watchdogs, in hardware and software then could monitor the operation of the system and prevent such a scenario. Safety critical firmware / hardware systems are everywhere. (airbag ECU, anyone?) How often does this type of mishap occur?

And you would let some single engineer write code to control your engine ?. The amount of testing GM etc must do to ensure fail safe code must cost millions.

But an LPG ECU doesn't control the throttle, and without opening the throttle there is nothing you can do to a petrol engine to make it fail unsafe.

I have been involved in an open source project to build an engine ECU for years  http://www.megasquirt.info/. The core code in that box is very small. About 2500 lines of 8 bit assembler. Most of that code is to allow the ECU map to be edited, and not used to actually run the engine.

I think the worst "bug" I have experienced in running round on beta firmware was a problem where the engine would only start when cold. It did cause a bit of embarrassment when I stalled it but no safety issue at all.

Sadly, it doesn't yet incorporate a stability control system compatible with driving on icy puddles. Recent experience shows where the real point of failure is inmost car designs - the wheel-to-seat interface!

I, for one, would not hesitate in trying Teilo's ECU. In fact if, in the light of the above, he thinks my input would be useful I'd be happy to do so.

Kevin

[Ken T]

http://www.compliance-club.com/

Ken

[Kevin Wood]

http://www.compliance-club.com/

Ken

Ahh, yes. Good guys. Have come across them as they share my loathing of power line networking adaptors. I've got that journal on my desk at work. Will have a good read of that.

Kevin

[feeutfo]

been to work and back up the m4, still happy with the gear changes, full throtle lift off is not quite as improved as first appeared this morning, weird, still needs work but better, defo.
 

[Marks DTM Calib]

Yes, and design and produce a mould for a zinc diecasting. I'm not saying it can't be done, I'm just saying he will need to shift a lot of units to repay development.

Or use an off the shelf item....there are plenty about!

Re the LPG ECU, well its in series with the petrol injector lead, and any element in series can fail. It would be interesting to see how they arrange to "fail safe ".

Easy enough, relays, OC outputs, watchdogs etc will see this sorted easily. 

Subcontract manufacturing, well the place I am currently at, use a subcontractor to source parts and build stuff. Problems I had the other day were an IC the wrong way round (16Pin DIL), tombstone resistor and the soldering is way up the DIL ic legs. Oh and they are IPC certified. 

The build is only as good as the design....and there is more to the design than just the circuit and pcb.....

I have been using outsourced manufacture for uber complex buils for 8 years plus now.....and no major problems at all.

And you would let some single engineer write code to control your engine ?. The amount of testing GM etc must do to ensure fail safe code must cost millions.

Ken

Yes, a compitent engineer yes....it happens all the time already today......

[Lazydocker]

I had a good play today and it's still not perfect on mine either, but much better.

I can sense another small pressure adjustment and re-map coming on

[feeutfo]

I had a good play today and it's still not perfect on mine either, but much better.

I can sense another small pressure adjustment and re-map coming on

was wondering in the car on the way home tonight, is it worth going to 1.00 bar?

What are we compromising between with this setting? if you see what i mean? we know what happens when going too high, lag/flat spots. What happens with too little pressure? Not sure i understand it fully, i'm thinking it does the same as the fuel pressure regulater on petrol i suppose?
 :-/

[Kevin Wood]

was wondering in the car on the way home tonight, is it worth going to 1.00 bar?

What are we compromising between with this setting? if you see what i mean? we know what happens when going too high, lag/flat spots. What happens with too little pressure? Not sure i understand it fully, i'm thinking it does the same as the fuel pressure regulater on petrol i suppose?
 :-/

Insufficient fuel at wide open throttle and running lean. Always check that, on a full throttle acceleration run, the Lambda sensor outputs stay rich right up to the red line. Do this whenever you've adjusted anything.

Kevin

[feeutfo]

was wondering in the car on the way home tonight, is it worth going to 1.00 bar?

What are we compromising between with this setting? if you see what i mean? we know what happens when going too high, lag/flat spots. What happens with too little pressure? Not sure i understand it fully, i'm thinking it does the same as the fuel pressure regulater on petrol i suppose?
 :-/

Insufficient fuel at wide open throttle and running lean. Always check that, on a full throttle acceleration run, the Lambda sensor outputs stay rich right up to the red line. Do this whenever you've adjusted anything.

Kevin

oh yeah, that. Will do.

[TheBoy]

was wondering in the car on the way home tonight, is it worth going to 1.00 bar?

What are we compromising between with this setting? if you see what i mean? we know what happens when going too high, lag/flat spots. What happens with too little pressure? Not sure i understand it fully, i'm thinking it does the same as the fuel pressure regulater on petrol i suppose?
 :-/

Insufficient fuel at wide open throttle and running lean. Always check that, on a full throttle acceleration run, the Lambda sensor outputs stay rich right up to the red line. Do this whenever you've adjusted anything.

Kevin

Won't it detect lean running, an pop back to petrol?

Drill the injectors out more if it does?

[Kevin Wood]

Won't it detect lean running, an pop back to petrol?

No - it only monitors the vapour pressure so if that pressure is not able to deliver enough fuel through the injectors...

Drill the injectors out more if it does?

I think we're at the max. at 3mm. Then again, according to the documentation that's supposed to be good for more than 200 BHP @ 1 BAR.  :-/

Kevin