Projects - Trials - Experiments

Micro CHP 2

The objective of this exercise is (apart from keep me amused for yonks) to produce a gas powered engine and control system that powers a small generator. The waste heat produced (some 70-80% of the energy) will heat the workshop to start with and may be the house if I feel adventurous. The generator will produce electricity to charge the 2200Ah of storage I have attached to my PV System.

After some thought I decided I needed an engine that would run the old fashioned way i.e. distributor and coil rather than an ECU.So I could slowly add the required sensors and write the necessary software for the Engine Control Unit I wanted to design and build. Then be able to test it while the engine was actually running. This would allow me to debug both the hardware and software I was going to design bit at a time rather than have to have it perfect first time for the engine to run. Right first time was NEVER going to happen...

So I bought an old A series mini engine luckily mounted on a frame which saved me some considerable work fabricating one. It was spotted by a good friend on EBay, hell he even fetched it, Thanks Graham. It was an ex UK Military engine Refurbished, Preserved and packaged by REME in 1994. On arrival I filled it with oil, connected a coil and a battery, shoved some propane up the air intake and it started first time. The army certainly know how to look after things. It was £250 including the frame I thought it was a good price as I needed the frame. Anyway when I have the software and hardware done I will buy one that has the sensors and NEEDS an ECU and sell this one without the frame.

Just shoving the gas pipe up the air intake was ok for a quick test but as I wanted to run it for a short while regularly to check gas flow rates and RPM etc I did a safer job after removing the petrol needle assembly from the Carburetor stopping the pipe falling out while I was looking the other way.

This was a temporary solution but it actually fell together rather well as you can see.

It got to the point I needed to run it for more than 60 seconds now and again so I had to provide a coolant system Radiator and pipework etc. Found lying around at the scrappy a Peugeot 206 radiator with everything except the expansion bottle £20....

Cleaned it out and piped it up. Made the expansion bottle out of a pop bottle, they're good to at least 150 PSI. I've had one to 180 PSI playing bottle rocketsThe over pressure relief valve in the cap was fabricated too and blew off at 15PSI. ( I have a proper one now but it worked fine for a month or so till Graham located the correct thing. Thanks again Graham.

The Throttle needed to be under computer control so I fabricated a throttle control out of a stepper motor with gear assembly...

and a lever arm made from PCB and the inner out of chocolate block.

The gas flow also needed to be under computer control too this was not so easy, this is a needle valve re-engineered from a Brass water pipe T piece some other brass water pipe fittings and again a stepper motor this time from an old floppy disc drive. In the Tee section connecting to the base of the Carburetor there is a flashback arrestor (Stainless steel wool to cool the flame bit like a Davey lamp) Would you believe the Carburetor needle assembly was fitted with a BSP thread?

The gas flow control did not completely switch the gas flow OFF so a solenoid valve was added to ensure NO gas flowed in the OFF condition.

In order to know the position of the crank and detect Number 1 TDC a crank sensor is required. Yes this is a bit of a bodge but it works. Twelve equally separated magnets on a disc of laminated PCB and a 13th magnet just before #1 TDC and of course a hall effect sensor. Ideal nice big 5V signal with each magnet pass. This was tested on a small DC motor and was good well past 7000 RMP. (I'm not going there with an old mini engine scary)

The relays are for the starter motor (Solenoid), Coil cct (for the moment), Solenoid Gas Valve and a spare

The proto type ECU so far this only detects TDC and generates a pulse to trigger ignition sparks for cylinders 1&4 and 2&3. This currently only triggers a strobe. The firing angle advance and retard of this pulse is variable. The three attempt auto start sequence with lock out and stall fail safe (Isolate gas if engine not turning) stuff is all written too. The gas flow, throttle position, and firing angle, engine speed, and load table is started but still needs a lot of work. Load trials etc. i.e. I need to fit the alternator to the engine so I can load it and run some load trials (Whilst still using the distributor). I also need to fit the exhaust gas oxygen sensor (Lambda sensor) I guess a proper exhaust system would be handy at that point...

Incidentally there is a bunch of us that go out for breakfast regularly and one of these friends after seeing this lot in action gave me a CO monitor with audible alarm... Not sure if I should be flattered or insulted ...Thanks Jon...This is a picture of the Crank Sensor illuminated by the 10us strobe pulse... At TDC there is a photo luminescent sticker that can be seen in the photo.... The mess of light at the bottom of the picture is the strobed LED's. I have since improved this arrangement, I now use a multi LED 12v ceiling light.

2008/08/25-0511 The above was done in the latter half of 2007 and I have not been able to get back to do any more as I have had my hands full of house repairs and alterations. The next things that need doing are to port the software to a different PIC chip it currently runs on a PIC18F252 and I want it on the PIC18F452 which has more I/O and design a Prototype PCB to illuminate that mess of a prototyping board, this should improve reliability and noise immunity. Also the homemade gas control valve, while very pretty, intermittently sticks closed and Good old Graham has found me a Tartarini Control valve from a gas conversion kit and I want to adapt this.

There will be more as soon as I get the shit to do, done... I could use a spare lifetime...

	Micro CHP 2	All Images

The objective of this exercise is (apart from keep me amused for yonks) to produce a gas powered engine and control system that powers a small generator. The waste heat produced (some 70-80% of the energy) will heat the workshop to start with and may be the house if I feel adventurous. The generator will produce electricity to charge the 2200Ah of storage I have attached to my PV System. After some thought I decided I needed an engine that would run the old fashioned way i.e. distributor and coil rather than an ECU.So I could slowly add the required sensors and write the necessary software for the Engine Control Unit I wanted to design and build. Then be able to test it while the engine was actually running. This would allow me to debug both the hardware and software I was going to design bit at a time rather than have to have it perfect first time for the engine to run. Right first time was NEVER going to happen...
	So I bought an old A series mini engine luckily mounted on a frame which saved me some considerable work fabricating one. It was spotted by a good friend on EBay, hell he even fetched it, Thanks Graham. It was an ex UK Military engine Refurbished, Preserved and packaged by REME in 1994. On arrival I filled it with oil, connected a coil and a battery, shoved some propane up the air intake and it started first time. The army certainly know how to look after things. It was £250 including the frame I thought it was a good price as I needed the frame. Anyway when I have the software and hardware done I will buy one that has the sensors and NEEDS an ECU and sell this one without the frame.

	Just shoving the gas pipe up the air intake was ok for a quick test but as I wanted to run it for a short while regularly to check gas flow rates and RPM etc I did a safer job after removing the petrol needle assembly from the Carburetor stopping the pipe falling out while I was looking the other way.

	This was a temporary solution but it actually fell together rather well as you can see.

	It got to the point I needed to run it for more than 60 seconds now and again so I had to provide a coolant system Radiator and pipework etc. Found lying around at the scrappy a Peugeot 206 radiator with everything except the expansion bottle £20....

	Cleaned it out and piped it up. Made the expansion bottle out of a pop bottle, they're good to at least 150 PSI. I've had one to 180 PSI playing bottle rocketsThe over pressure relief valve in the cap was fabricated too and blew off at 15PSI. ( I have a proper one now but it worked fine for a month or so till Graham located the correct thing. Thanks again Graham.

.	The Throttle needed to be under computer control so I fabricated a throttle control out of a stepper motor with gear assembly...

.	and a lever arm made from PCB and the inner out of chocolate block.

..	The gas flow also needed to be under computer control too this was not so easy, this is a needle valve re-engineered from a Brass water pipe T piece some other brass water pipe fittings and again a stepper motor this time from an old floppy disc drive. In the Tee section connecting to the base of the Carburetor there is a flashback arrestor (Stainless steel wool to cool the flame bit like a Davey lamp) Would you believe the Carburetor needle assembly was fitted with a BSP thread?

.	The gas flow control did not completely switch the gas flow OFF so a solenoid valve was added to ensure NO gas flowed in the OFF condition.

	In order to know the position of the crank and detect Number 1 TDC a crank sensor is required. Yes this is a bit of a bodge but it works. Twelve equally separated magnets on a disc of laminated PCB and a 13th magnet just before #1 TDC and of course a hall effect sensor. Ideal nice big 5V signal with each magnet pass. This was tested on a small DC motor and was good well past 7000 RMP. (I'm not going there with an old mini engine scary)

	The relays are for the starter motor (Solenoid), Coil cct (for the moment), Solenoid Gas Valve and a spare

	The proto type ECU so far this only detects TDC and generates a pulse to trigger ignition sparks for cylinders 1&4 and 2&3. This currently only triggers a strobe. The firing angle advance and retard of this pulse is variable. The three attempt auto start sequence with lock out and stall fail safe (Isolate gas if engine not turning) stuff is all written too. The gas flow, throttle position, and firing angle, engine speed, and load table is started but still needs a lot of work. Load trials etc. i.e. I need to fit the alternator to the engine so I can load it and run some load trials (Whilst still using the distributor). I also need to fit the exhaust gas oxygen sensor (Lambda sensor) I guess a proper exhaust system would be handy at that point...

	Incidentally there is a bunch of us that go out for breakfast regularly and one of these friends after seeing this lot in action gave me a CO monitor with audible alarm... Not sure if I should be flattered or insulted ...Thanks Jon...This is a picture of the Crank Sensor illuminated by the 10us strobe pulse... At TDC there is a photo luminescent sticker that can be seen in the photo.... The mess of light at the bottom of the picture is the strobed LED's. I have since improved this arrangement, I now use a multi LED 12v ceiling light.

2008/08/25-0511 The above was done in the latter half of 2007 and I have not been able to get back to do any more as I have had my hands full of house repairs and alterations. The next things that need doing are to port the software to a different PIC chip it currently runs on a PIC18F252 and I want it on the PIC18F452 which has more I/O and design a Prototype PCB to illuminate that mess of a prototyping board, this should improve reliability and noise immunity. Also the homemade gas control valve, while very pretty, intermittently sticks closed and Good old Graham has found me a Tartarini Control valve from a gas conversion kit and I want to adapt this. There will be more as soon as I get the shit to do, done... I could use a spare lifetime...