I’m restoring a Triumph Herald 13/60 built in 1970 and the Stromberg carburetor is a mess. These carburetors were clever devices for their time but I’m wondering if I could build a fuel injection system using just analog electronics which has an improved performance with the possibility to fine tune.

I’ve considered a processor driven approach but old cars are an electrically noisy place to be so I’d like to try analog. I’d use a throttle body from maybe a VW Golf with throttle sensor in it, a pressure sensor for manifold pressure and a fuel injector firing from outside the throttle body, timing from sniffing the sparks and add a temperature sensor. A facia mounted pot to lean or enrich the mixture could be useful. Not as good as a regular system with lambda sensors, etc, but simple!

I anticipate responses which say this is crazy but I’m wondering if anyone has actually tried this? I’ve seen a schematic for a system for an aero engine which uses a 555 monostable to control the on time of injectors!

 

I’ve considered a processor driven approach but old cars are an electrically noisy place to be so I’d like to try analog.

Surely electrical noise would be more likely to mess up analog signals than digital ones?

 

I would think that building you own would be problematic in getting the proper fuel mixture under various driving conditions.
Have you looked at aftermarket systems?

 

I just happen to have a Stromberg carburettor on a Sunbeam Alpine Fastback engine in my shed. . .
The carburettor doesn't know the position of the throttle, but it does need to know the air flow and inlet vacuum, and it does know the position of the choke control (which moves the orifice downwards to engage with a narrower portion of the needle to allow more fuel through)
The profile of the needle acts as a look-up table to give the amount of fuel from the mass flow and inlet vacuum. I'm not sure how that could be implement in analogue.
As the Stromberg exists to circumvent Skinner's patent on the SU carburettor, perhaps a good place to start would be an explanation of how the SU works.
http://bobmccluskey.com/SU Carburettors.shtml

 

Thanks for all your thoughts. Whilst I have considered aftermarket systems like megasquirt and speeduino I thought it might be fun to build a simple system, albeit without the sophistication of these systems but potentially at least as good as a carburetor. The thinking behind analog rather than digital is that if a processor loses its way then the failure is absolute. I’m not expecting it to mimic the performance of a Stromberg or an SU but to base injector on time to deliver a stoichiometric mix based mostly on the manifold pressure. The difference between optimal, lean and rich mix is around 20% which allows carburetors to get away with relatively poor performance compared to modern fuel injection systems

I’m still hoping I might find someone on this site who has actually had a go at this!

 

I entertained the possibility of making my own Analog-Fuel-Injection,
and after spending a stupid amount of time and effort on it
and I determined that it was just a plain dumb, overly-complex and expensive-idea.

I went with a DIY-Parts-Kit for one of the MegaSquirt designs and never regretted it.
It performed flawlessly for over ~13-years, and was sold, still running, with almost 300,000 miles on the car.

MegaSquirt is adaptable to virtually any application.

Don't waste your time, go to DIY-Auto-Tune, and pick-out one of their packages.
They also have Sensors.
.
.
.

 

An analog injection control may indeed be simple, depending on the performance you require, and the emissions standards you need to satisfy. I was a part of the analog fuel control development project , department 7430, I think, at a Highland Park segment of a large auto company, back in the 1974 to1976 era. That scheme did not use injectors, which is why the concept failed miserably..Injector IS the only way to go. Pat of it must be digital because injectors are either on or off. And your system will need to measure manifold pressure/vacuum, air mass flow and temperature.

BUT you can drive a car with a 400 CID engine using a manually controlled fuel delivery and a gas pedal to work the throttle opening, and nothing except a single knob to adjust fuel delivery. The economy with that $15 control will be poor, the emissions will be illegal in most of the world, but it can be done. that simply.
Take advice from LQC and at least investigate the kit. It does take more than a 555 and two transistors and a simple pot to do it right.

 

An analog injection control may indeed be simple, depending on the performance you require, and the emissions standards you need to satisfy. I was a part of the analog fuel control development project , department 7430, I think, at a Highland Park segment of a large auto company, back in the 1974 to1976 era. That scheme did not use injectors, which is why the concept failed miserably..Injector IS the only way to go. Pat of it must be digital because injectors are either on or off. And your system will need to measure manifold pressure/vacuum, air mass flow and temperature.

BUT you can drive a car with a 400 CID engine using a manually controlled fuel delivery and a gas pedal to work the throttle opening, and nothing except a single knob to adjust fuel delivery. The economy with that $15 control will be poor, the emissions will be illegal in most of the world, but it can be done. that simply.
Take advice from LQC and at least investigate the kit. It does take more than a 555 and two transistors and a simple pot to do it right.

Thanks for this, it reassures me that I’m not completely crazy, just mildly crazy. In the UK the permitted emission levels are related to the age of the engine so the bar is set quite low (or do I mean high?). My view is that the approximate on time of the injector is mostly determined by the manifold pressure and you can get away without a MAF sensor. And the exact timing of the injector pulse is not critical for injection into the single manifold of a four cylinder engine, so long as it fires in sync with the engine speed.

The idea of a $15 system is intriguing, but maybe you guys are right, I should go for an aftermarket system, or simply rebuild the carburetor…

 

That $15 system was for emergency use when the developmental system failed on the road. It provided fuel delivery by the driver twisting a knob, and could easily run rich enough to make black smoke at 50 MPH in low gear, close to engine redline RPM.