MG-Cars.info

Welcome to our Site for MG, Triumph and Austin-Healey Car Information.

Parts

MG parts spares and accessories are available for MG T Series (TA, MG TB, MG TC, MG TD, MG TF), Magnette, MGA, Twin cam, MGB, MGBGT, MGC, MGC GT, MG Midget, Sprite and other MG models from British car spares company LBCarCo.

MG Midget and Sprite Technical - Midget 1500 Fuel Injection

For those of you that don't know, I am currently in a bit of a deep hole of a project, doing a DIY fuel injection conversion on my 1500.

If you are interested, here is a five minute video I made whilst doing my first systems check this morning. Cheers.

Malc.

http://www.youtube.com/watch?v=1JC1TdKYnlM&feature=youtu.be
Malcolm Le Chevalier

That was great but wow so much wires, boxes and plumbing.

You've a good presentation voice are you a tutor or instructor of any sort?

But - people do have to watch to discover if they like how you do things. :)

Lawrence would be very proud of that fuel rail, though it'd be far too neat for him.

I must admit I was rather hoping the laptop would fly out of the car when you switched the ignition on, just for the fun of it.

The K-Series Zetec Mafia Facebook Group would probably love that (or hate it?).

Great, really interesting, I look forward to seeing more, not that I understand any of it, can't wait for the actual road tests.

Well done.

Nigel Atkins

Thanks Nigel, you are very kind. No, I am not a tutor or anything, I work in an office. But I do work in a profession where you have to be able to explain technical things clearly and concisely and I do enjoy playing teacher when the need arises.

I am part of the K/Z group but not posted it there yet. It's not K, but it is probably more relevant to their interests than to the wider Spridgeting groups.
Malcolm Le Chevalier

Yes Malc I thought you'd gone to the darkside, do you have to have the curtains fully drawn to post in daylight and/or on here.

Your secret is safe with me and the CIA, NSA, Homeland Security, FBI, GCHQ, Google, Facebook, China, Russia and Mossad.
Nigel Atkins

Malc

I wondered what you'd been brewing as had been quiet lately! Looks fascinating, thanks for sharing. Makes me think of the various experiments with fuel injection on A series engines for competition use - rallycross minis I seem to remember. Wonder what the Triumph completions dept were up to in the 1960s and 1970s (there may have been a 4-wheel drive Triumph1300 rallycross saloon)

Cheers
Mike
M Wood

Looking good Malc.
Your excitement when the throttle sensor did it's sweep said it all
If I could make a little costructive suggestion-- When you run your ignition sensing wiring loom back to the ECU, Usually the ign. wiring needs an earthed shield if it gets tied in with other wiring in the loom. If it's not shielded it needs to be seperated from the other wiring going to the ECU to stop electrical noise playing with the system

willy
William Revit

Thanks Willy. I had not thought about that.

I was still pondering where to mount the little signal conditioning circtuit board, but based on that I will mount it under the dash with the ECU so there is minimal chance for the output signal to pick up noise.

Cheers,
Malc.
Malcolm Le Chevalier

Hi,
Nice video ! Definitely a lot of work been done there ! Impatient to hear the test with the injectors on :)
Just one small comment if I may: maybe it would be interesting at some point, to list/explain where you source the parts and why, as they are from a different vehicules... Just to understand if it is because they fit, or you had them on the shelf :)

Impatient to see the next one :) Good luck !
CH Hamon

eBay -> sort -> Price Low - High :-D
Malcolm Le Chevalier

Fascinating Malcolm. I hope it starts when the time comes! But what to do with the choke ;-)
Bill Bretherton

I haven't had a choke in years Bill!

Previous setup was a DCOE, which has more of a flood mechanism than a choke mechanism. So it was disconnected and binned a long time ago.

The choke hole in my dash now has the horn switch in it.

:-)
Malcolm Le Chevalier

The next riveting episode is now live!

http://www.youtube.com/watch?v=JVM5q1fF-6c
Malcolm Le Chevalier

This is great Malc. Watching with interest.
Greybeard

Lookin' good Malc.

Another little thing for you to think about--
I've found that with the pressure regulator mounted in the rear of the car, that the length of hose required for the vac. signal causes a bit of a delay in changes to the vac. reading as seen by the regulator, and will make it difficult to get the mixture correct for instant throttle response.This is caused by the hose holding vac for a split second which delays the required pressure increase that you have programmed in for that amount of throttle
You're better off not running a hose to the reg. and by doing that you have a constant known pressure which will make programming easier without the varying pressure altering your required mixture

Most modern cars like Ford Mondeo -Focus etc that run a single line non return fuel system just have the regulator as part of the pump assy. in the tank with no vac. referance at all-constant pressure

Can I ask what brand the ECU is--

Cheers
willy
William Revit

You reckon it really makes a noticable difference Willy? Interesting. I find it hard to believe that ~2m of 3mm bore hose can have any sort of perceptible lag. Especially when the speed of sound @ 350 m/s. Amazing if it does, I will see.

ECU is a Speeduino NO2C from diy-efi.co.uk. The Speeduino board is effectivly a switch board that piggybacks on to a cheap as chips open source Arduino microcontroller board that does all the heavy processing work. I soldered it all up myself.

The NO2C is their basic, two channel model. They do other boards too.

Malc.
Malcolm Le Chevalier

Really interesting, again :)

As far as I remember, it is also question of "fluids dynamic" and perturbations... Every time an injector open and close, it is creating a sort of "wave"/"perturbation" inside the high pressure system. Similar to opening and closing your tap: "coup de belier" in French, water hammer if my translator is correct...

Obviously, it depends on a lot of parameters: RPM, number of injectors, pressure, throughput, etc. Car manufacturers do use a lot of models and software to measure it perfectly. I do understand the scale of your installation, and the logic you went through to try a simpler method first: it would be interesting to see if it works like this :)

Good luck ! And....where is the next video ? :p
CH Hamon

Yeah, pondering this a bit more. It's not the vac' signal that is the problem is it. But the fuel pressure wave travelling in the other direction once the pressure regulator has altered it's setting. Interesting. *strokes chin*

I was aware that most cars these days have the PR in the tank, hence why I thought it would work OK this way.

Cheers guys,
Malcolm.

Malcolm Le Chevalier

I think the graph, coming form here (https://en.wikipedia.org/wiki/Water_hammer) explain visually what Willy said.

Then, the real question is "how big is the impact on your system": is it negligible or not...

Just for curiosity: which high pressure are you running ?

EDIT: posting while you were posting too :P

CH Hamon

Lost my edits cross posting with you CH. Oh well! :-)

The high pressure is 3 bar.

I understand water hammer, I work designing industrial pumping and piping systems (different scale when your pipe is 36" diameter and at 200 bar :-D) but hadn't thought about how it might affect this.

The proof of the pudding shall be in the eating.

Malc.
Malcolm Le Chevalier

Ground clearance might be an issue with 36" pipes... :p

I would say 3 bar is not so much, so I see where you come from to take your chance :)

Today, cars use mutli-injection + direct injection + > 1 500bar, so they do have a lot to do to keep it "sustainable".

Impatient to see the next video, good luck :)
CH Hamon

Why not mount the LP fuel pump and regulator up front and pipe the regulator overflow back into the fuel supply pipe?

Would save crawling under the car too!
Chris at Octarine Services

Malc,
on a very technical note - people less than twice your age struggle.

Even allowing for sizes of unions, was running rubber hose pipe instead of solid pipe just a matter of convenience?
Nigel Atkins

Malc.
Your setup is fine as is and worth a try with the vac hose connected-

I only mentioned it as I've been playing with an Omex ecu on a friend's Escort
Same setup as your car with the pump and reg. down the back,return back to the tank, and a single HP line to the front, but no O2 sensor and throttle body manifold, no vac sensing to ecu.
There was no vac fitting on the manifold so we actually drilled all 4 manifold runners to get a decent manifold vac. reading(at the manifold)with the result of3.0bar hose off 2.5 on- fuel pressure
So set the cruise mixture at 15:1(2.5b) and full throttle at 12:1(3.0b)
Cruising along ok at 80kph and give it some acceleration and the mixture would ping off to around 17:1 for a split second accompanied by a little hiccup then start into the 13-12's and away it went----Tried a cruising mixture as rich as14:1 but it'd still flick lean
Decided that although the full throttle mixture was set plenty rich enough, that setting is with full 3bar fuel pressure,and with the little delay in the regulator, when the throttle was first opened the delay in vac control was causing a low fuel pressure when it needed to be high
Disconnected the hose and it's all good
I feel it was just a matter of too much volume in the long hose causing the delay
Most cars with a single non return system run 3bar all the time--You'll find it a lot simpler to set up with a constant fuel pressure reading-

Your pump has to stay at the back if you're not running a lift pump, the HP pumps don't suck very well they're pushers not pullers ,
Keep going as you are but if you find you have that little stumble on acceleration, give it a try with the hose off
Your ecu with vac. sensing may well self adjust-
Sorry if I've raised a ghost, probably shouldn't have said anything but there you have it
willy
William Revit

I appreciate your feedback Willy. It really is useful. As I say in my disclaimer, I don't know what I am doing, I just mess about until it works! :-D I was only going to hook up the vac' because there was a tail on the FPR to do so!

Thinking about it more... as I plan to use the TunerStudio auto tuning features, if I forget the vacuum line to the FPR it is basically one less variable. The software will do the rest based on throttle position, manifold vac', lambda and all the rest. Right?

KISS?

Malc.
Malcolm Le Chevalier

And Nigel, yes, flexi hose was easier and less connections. As the tank, filter, pump, regulator and fuel regulator all had tails for flexi hose, it seemed pointless then going from flexi, to hard line, to flexi etc. More joints, more places for leaks!
Malcolm Le Chevalier

Yep, KIS,(left one off just for you)- you're onto it, - have fun

willy
William Revit

I am much more clueless about all this than you are, but when I put a hot-wire injection system in my MGB V8 I used a low pressure pump feeding a 1 litre swirl pot, and then a high pressure pump to feed the engine, returning to the tank via the swirl pot. The reason I did all this complication is that someone told me that the single pump system works fine until the tank is getting low in fuel. Then as the fuel slops around with the movement of the car the fuel pump can draw air. In an engine with carbs this doesn't matter because the motor is running off the float chambers, but an injected engine will stumble and misfire as an air slug goes by the injectors. Modern cars have a pot within the tank I believe.

How true that is, I don't know. But my system has worked perfectly for over 10 years. Of course there is more room to accommodate all the gubbins in the MGB. I went down to a single battery and used the other bin for the fuel pumps and pot.

I'm 72 and still messing around under cars, although it takes longer to get back up again! There's hope for you yet. Fascinating work. Hope it turns out well.
Mike Howlett

It's a reasonable point you make Mike but i think it will be ok
With the non return system there's not as much fuel racing around as there is in a return type system but if it did suffer from surge, the T fitting on the outlet of the fuel pump could be replaced with a small 200-300cc tank with the inlet towards the top ,return out the top and outlet to the engine out the bottom
That would work like a mini swirlpot and allow any air to spit back through the reg. while maintaining fuel for the outlet
It's something that could be done later easily if there is an issue
Good point though
willy
William Revit

There you go Willy. Nice idea. I know my low pressure pump does draw air when the tank is low on fuel, particularly when braking. It is in the ex-battery bin right behind my seat. I can't normally hear the pumps once the engine is running, but if a slug of air goes through, the LP pump makes a louder noise which is audible.

Incidently Willy, don't you ever go to bed over there in Tasmania? No matter what time of day, you are always here with useful comments.
Mike Howlett

Episode three of my dulcet waffling about my 1500 fuel injection project. Perfect timing for sending you to sleep tonight! :-)

More stuff on the fuel system in this episode.

Hopefully, as there is a bank holiday weekend ahead, I will be getting close to firing it up... maybe...

http://www.youtube.com/watch?v=c5D7vQjur_s&feature=youtu.be&fbclid=IwAR1qziPUUZeTgCrkDHXAMQlsYarGiXrkFXMyilGIXTq5CSepmel21z9uDzc
Malcolm Le Chevalier

Malcolm, can you contact me re. hub gaskets please at:

brethertonwilliam(at)yahoo(dot)co(dot)uk

Bill
Bill Bretherton

We should open a community of people leak-testing "stuff" in their kitchen sink with foot pump :D
Nice common rail, I guess Bosch were doing pretty similar experiment for their first one :p
We all keep our fingers crossed ! Good luck !
CH Hamon

I had actually been suffering a bit of a mental block as to how I was going to test it. Then I saw your thread and that was the inspiration to get it done. So thanks CH! :-)
Malcolm Le Chevalier

You are welcome ! Finally feeling useful down here... :p
No patent issue either as you don't use any wine stopper in your leak-test... Definitely not made by a French ! ;)
CH Hamon

Well, it worked on the first try!

I did a video again, two in fact!

First start:
http://www.youtube.com/watch?v=zn-_pIL45z4

A little bit later once the adrenalin had subsided and I could actually check things over and fiddle a bit:

http://www.youtube.com/watch?v=g60JMX5hZ4A

I was able to drive it a little yesterday, first time since December. I now need to start the tuning process. Hopefully Tunerstudio will help me with that.

Malc.
Malcolm Le Chevalier

Magic stuff Malcolm--have fun fiddling
William Revit

Well done.

There was a fair amount of surprise that it works from someone off camera.

But carb fiddling with computers is definitely in fetish territory.
Nigel Atkins

I think there would be a lot of interest in what you've done here Malcolm if you were to display this at the NEC in the future.

Very impressive
Philip Sellen

I'm sure there would, and I would love to. Practical Classics show would be ideal.

But shows seem a distant memory at the moment :-(
Malcolm Le Chevalier

Restoration Show at NEC is planned for August.
Nigel Atkins

It may be planned but has zero chance of happening. I cant see any mass gatherings being allowed this year, even though I think they should start happening again soon.

Trev
T Mason

Wow, congrats ! The difference between the two videos is pretty impressive too ! So I imagine the difference you will get after fiddling for the video 3 !

Impatient to see the difference on a RR too ! But we will be patient and leave you some more time to play with it first :p

Cheers and congrats, hope it is only the beginning !
CH Hamon

A few issues to sort first, of course!

Been out twice and it has been great... Until about 30 - 40 mins it starts running like crap and I have limped home. Last night it completely conked out trying to get it in the garage and I had to push it!

I suspect I am cooking the fuel rail (with it being over the exhaust manifold). I have some insulation wrap to try out.
Malcolm Le Chevalier

A reflective heat shield may work better than insulation wrap. Heat penetration through insulation is a factor of time, and the insulation will prevent air circulation around the pipe. Reflective shielding still allows for air circulation - so long as that air isn't too hot !
GuyW

Thanks Guy. I agree.

For now, what I have kicking about is some thin, reflective foil backed insulation for insulating under laminate floors (?) and such like. I will see if this helps in the first instance.
Malcolm Le Chevalier

If the heat shield works--
You might have to end up fitting your regulator up the front and run the return right back to the tank
That way you'll have a nice fresh,cool supply from the tank circulating all the time---just a thought
William Revit

Trev,
you were right about the August show.-

"Practical Classics - Classic Car & Restoration Show Cancelled

Today 27-05-2020 we have been informed that The Classic Car & Restoration Show, NEC has been cancelled.
Members should follow this LINK to the Practical Classics website where more information is available.

The show has been rescheduled for 26-28 March 2021".
Nigel Atkins

Malc,

I have just watched all the videos - must have been great fun and very satisfying to get it running well?

After the first or second video, I did wonder why you have designed the common rail as a 'dead end', when it will be subjected to heat from the engine bay generally, and the exhaust manifold immediately below it even more.

Why would you not run the fuel return line from the opposite end of the common rail to the supply, back to the tank to ensure that you get a reliable flow of liquid fuel, and not a mixture of liquid and vapour?

Richard
Richard Wale

Well done Malc! I can't offer any advice but it's impressive.
Bill Bretherton

Malcolm,

Nice work and great it runs. I think though it would be a good idea to add a safety drain tube to the drain spigot on the fuel rail due to the proximity of the exhaust as I've had that style leak on my central heating system and wouldn't trust the rubber seal, if yours has one, to stand up to petrol long term. I think some early Sprites had something like that on the SUs to take any overflow safely away from the exhaust.
David Billington

Nigel, yes I saw that and also the Silverstone Classic a couple of days ago. Just leaves the Goodwood Revival and the NEC show of the big events now and I suspect they will follow in due course.

Trev
T Mason

Silverstone relies quite a bit more on the roundy-round elite celebrity circus.
Nigel Atkins

Richard,

Yeah, at the time I was thinking simplicity and ease of plumbing (lazyness!) and that loads of moderns run dead end systems just fine. I might reconfigure things next winter. For now, I am happy to be driving for the first time in six months.

David,

Yes, that is on the list of things to do. Although I yet to decide what to do long term. Convert it to a return line or re-solder it with an elbow. I am not convinved I really need a bleed.


Took the car to work today, a 20 minute blast of city, country and motorway.

It is very lively in the mid-range and pulled strongly in 4th beyond 5000. Too many slow BMWs in the way on the m-way though! ;-)
Malcolm Le Chevalier

Nigel, I know you are not a fan of racing but the Silverstone Classic is as much a huge car show as it is about racing.

Trev
T Mason

Malcolm
Are you sure it's temperature /boiling vapour locking, I think you said somewhere it's a 3 bar system, my copy of Perry is at work so can't find the vapour pressure of Petrol but at 3 bar I'd be surprised if you were getting fuel hot enough , and if you were wouldn't it flash as it went through the injector (rather than spray) so you'd get all sorts of different issues (quicker bang if nothing else)

Could you be picking inerts up with your spirited driving - with a carb they separate in the float chamber and vent but in your system there's nowhere for them to go except form a bubble at the end

Just a thought

timmyk

Sorry just thinking on, serious point now
If it really is that hot, how good is your soldering? If you've got Petrol hotter than its atmospheric boiling point and you get a leak ( or even worse mechanical damage) and it flashes off under the bonnet things are going to go pretty nasty pretty quickly.
Ok the inventory is small in the header rail, but there's a lot of stored energy in Petrol if it all goes off at once
timmyk

Very interesting,
I recently made my own fuel injection system for my Suzuki Katana.

I configured my fuel system so it went from the tank to a filter to the HP pump to the fuel rail to the reg to the top of the tank.

This way the fuel is always circulating and not getting hot.

I used alpha N rather than vac and its works well on a NA engine.

Skelly.
Jon Skelly

I had a 1974 Volvo with a Bosch K-Jetronic system (continuous injection with relays and no CPU) with a single, high pressure fuel pump. These cars all had vapor locking issues. Volvo added a second, low pressure generic AC in-tank pump in 1975 and the problem was solved. The K-Jetronic system returned excess fuel to the tank. The vapor locking in the 1974 occurred in the high pressure pump itself (not in the engine bay) when it reached a certain temperature. The immediate fix was a squirt bottle in the boot so the the pump could be sprayed to liquify the vaporized fuel when the car stopped.
Glenn Mallory

That is interesting Glenn, thanks for sharing. We have had similar with ignition modules breaking down with heat.....CO2 extinguisher squirt soon showed what the problem was!
Peter Burgess Tuning

Glenn
Interesting about the K-Jetronic. I had two Volvos with it, a '79 245GLE and an '89 745GLE (I think the USA exports had electrical piezo injectors for the 740 range). Mine had the in-tank pumps and I remember repairing the 745's one, broken plastic blade I think. But it would run without the tank pump as you allude to. Apart from air filters, I didn't touch the injection system once (maybe tweaked the mixture for the MOT as there was no lambda sensor or ecu) and the 745 covered over 100,000 miles in the 20 years I had it, including several trips to the Alps and one to Vienna and Prague. Those redblock engines were superb. It was the 745 that towed the Frogeye 200 miles when I collected it for renovation, a few months before I sold it (the Volvo). Damn, I'm going all misty eyed now....

Sorry for the thread drift Malcolm.
Bill Bretherton

Glenn
Interesting about the K-Jetronic. I had two Volvos with it, a '79 245GLE and an '89 745GLE (I think the USA exports had electrical piezo injectors for the 740 range). Mine had the in-tank pumps and I remember repairing the 745's one, broken plastic blade I think. But it would run without the tank pump as you allude to. Apart from air filters, I didn't touch the injection system once (maybe tweaked the mixture for the MOT as there was no lambda sensor or ecu) and the 745 covered over 100,000 miles in the 20 years I had it, including several trips to the Alps and one to Vienna and Prague. It was the 745 that towed the Frogeye 200 miles when I collected it for renovation, a few months before I sold it (the Volvo). Damn, I'm going all misty eyed now....
Bill Bretherton

Interesting couple of comments there.
Peter, thanks for reminding me about the hot electronics problem. I'd forgotten about a similar episode with the CDI module on a Kawasaki bike which was tracked down by someone cleverer than me using a freeze spray.
Dad had a V12 Jaguar briefly which had an earlier version of the Bosch system. I think it was called just Jetronic without the K. We initially thought it was a Lucas system because it had Lucas labels on it but it turned out that Lucas had something to do with the design, but it was a Bosch system. It gave endless trouble and he got rid of it.
If I remember rightly the K system relied on an airflow sensor. Dad later had a BMW with it and it would occasionally falter in heavy city traffic on a hot day.
I once rented a BMW bike (K750) in California for two weeks which also cut out on me twice in traffic with vapour lock but I don't know what system it used.

Edit: I looked it up. Bosch L Jetronic, but gravity fed of course owing to the tank being above the engine.

The CAT diesels I work on have the CPUs in a box cooled by the passage of fuel through it, pulled through by the lift pump. The best part of it is Caterpillars name for it. They call it the Personality Module.

More thread drift...
Greybeard

Hello again folks.

Lengthy post alert!

It has been eight weeks since I first started the car on the injection system, time for an update and a few conclusions before this is finally laid to rest in the archives!

I have since been working through a bunch of stupid niggely problems and tidy ups, such as:

- Knackering, trying to rebuild and eventually replacing the starter motor.
- Almost boiling it dry as I didn't notice and deal with a small coolant leak (doh! noob error!)
- A whole load of messing with maps and the software.
- Heat wrapping the exhaust manifold, down pipe and fuel rail.
- Trying out different injectors.

I got a bit stuck and confused with the tuning process as whatever I did it wasn't quite right and it certainly wouldn't accept full power. Lambda readings were always tending toward lean on the power however much I messed and tried to enrich things.

I convinced myself, with the aid of lots of datalogging, that the injectors were maxing out and swapped them out for something a bit bigger, but still no joy.

Going back to basics I checked if the pump was pumping correctly and was a bit surprised to find it barely dribbled out of the end of hose! It wasn't the injectors but a distinct lack of fuel supply. Always check the simple things first. Easy to see in hindsight.

I removed the filter and it was so clogged the fuel didn't even come out when I disconnected the hose! I cleaned the filter and it ran brilliantly for about 30 minutes, then the issues started to slowly come back - the filter was impeding the flow again.

With the pump rated at 135 L/hour and anything above what is required (maybe 20L/hour?) being recirculated to the tank that is 115L/hour being recirculated. Or about a tank full every 15 - 20 minutes (very rough maths). It was doing a great job of washing out the tank and the filter - designed for carbs really - stood no chance!

Anyway, a much bigger and purpose designed filter swapped in and the car has been brilliant ever since. The maps are still self-learning, but are pretty much there.

It idles, cruises and accepts the power beautifully. Much better than before. I don't think there is much difference in the top end, but it has been a long time since I drove it on the carb setup. A rolling road sesh (Peter?) will be the proof.

It's been a big project (looking at receipts, I started buying things in July 2019!) but a great way to learn and had the desired effect of improving the general driveability and creating something quite unique and interesting.

So, what next? Turbo?! ha ha :-D

Well done if you have read this far! :-)

Cheers,
Malc.
Malcolm Le Chevalier

That's a pretty impressive piece of work Malcolm. Thanks for all the progress reports. You've ventured into realms that are well beyond my understanding!

Re. the filter problem - I wonder if a period of use with the new high capacity pump will effectively flush out the tank and remove all the crud that's built up over the years so that ultimately you have a completely clean tank?

If, however, you think it's going to be a recurring issue, would some sort of cyclone filter do the job? This should collect the sediment so you can remove it periodically without clogging a filter element further down stream.

I'll follow the turbo installation story with interest!

Colin
C Mee

Really interesting, and thanks for posting.
The rolling road numbers compared to your carb set up will be interesting too.
Chris Madge

This thread was discussed between 19/04/2020 and 20/07/2020

MG Midget and Sprite Technical index

This thread is from the archives. Join the live MG Midget and Sprite Technical BBS now