How do Atom Computer Ignition modules work?

[legendre]

Couple of quick updates..

Thanks to Ken for his assistance in obtaining a couple of the BU931T devices he has on-hand.. presently awaiting their arrival.

Also, the second batch of 2SD1071 have arrived, and so far they do not appear to be fakes. Using a 0-15V @ 1.5A variable PSU, I have roughly measured three randomly-chosen devices from the batch of ten, and all indicate hfe in the range of several hundreds (minimum) to several thousands. As expected, gain varies considerably with temperature and collector current.

Interestingly, all of the devices in the batch indicate hfe of 7-22 on my swissknife component tester. This is similar to what I saw from the previous batch of (fake) devices.. I take this to mean one of two things – either the gain (hfe) is so high, that it overflows and confuses the component tester, or the tester is incapable of sourcing sufficient base voltage to turn-on both junctions in the 2SD1071 Darlington pair.

So for the moment, I cannot suggest that anyone use a similar component tester to evaluate a 2SD1071. Use a variable low-voltage PSU and appropriate resistors in the emitter (3-10R) and base (10-20K) circuits, then do the math. I’ve calculated hfe figures in the range of 400 to 9000 at various voltages and currents, within this single batch of ten.

[debe]

I didn’t test my faulty batch as ive no idea how to do it, but they are fine in another application. I am still waiting for the arrival of 10 x BU931T that ive ordered.

[legendre]

quote debe:

I didn’t test my faulty batch as ive no idea how to do it, but they are fine in another application. I am still waiting for the arrival of 10 x BU931T that ive ordered.

2SD1071 is very easy to ‘proof’ vs. a fake. You only need to do this:

Get a 0-15V @ 1A or better variable PSU. Connect the DUT as follows:

B – to B+ via 10K resistor
C – to B+
E – to GND via 5R 1W (or just straight to GND… but it will be hard to control)

Start to turn up the voltage while watching current though the B (drop across 10K) and E (drop across 5R). By the time B current is 1-2mA the E current will be almost 1A or possibly considerably higher.

Recall that the 2SD1071 is a Darlington NPN with a +minimum+ hfe of 300.. and an hfe of several thousand is more like it, under most conditions. So.. just 1-2mA through the B/E circuit should give you a couple of amps through the C/E circuit, if both are fed in parallel from the same PSU.

If the DUT is fake, you’ll see only 20-50mA though the C/E for a B/E of 1-2mA. Huge difference, you can’t mis-read this one..

[legendre]

Today I received a packet from Ken, with two BU931 parts – thanks, Ken!!

I built up a module with the following component substitutions: BU931 for 2SD1071, 1K3 for both 1K5 and 1uF/50V (electrolytic) for 0.47uF. This is just what I had on hand.. need to buy some 1K5 parts.

It seems to work just fine, nice bright spark – again, using the 1947/1948 TD20 magneto as the testbed. The (debe) suggested Transorb diode +is+ being used.

Unfortunately, this is that same "problem child" TD20 that I cannot make run correctly in any form. At this point, I’m chalking it up to low compression (60/65) – which seems like it should be semi-OK, but the dang motor just will not run correctly, or even at all, at lower throttle settings. Believe me, I have tried. So I don’t see any point in trying to actually start and run the engine on the prototype module.

Lacking a runnable OBM, I’m thinking hard about putting this on my lawnmower, lol.. but is it worth it??

[debe]

I use the lawnmower as a test as well. Another reason a 2 stroke may not idle properly can be a badly worn piston or cylinder. I had one that wouldn’t idle properly due to a badly worn piston. The give away in this case was exess fuel spraying out of the carby.

[legendre]

quote debe:

I use the lawnmower as a test as well. Another reason a 2 stroke may not idle properly can be a badly worn piston or cylinder. I had one that wouldn’t idle properly due to a badly worn piston. The give away in this case was exess fuel spraying out of the carby.

This particular motor is a real thorn in my side. I can count on half-a-hand the number of machines (of any sort) that have genuinely left me flat-footed – and this is one of them.

Pardon the vulgarity, but I’ve been inside the frickin’ thing so many times, I feel like I’m married to it. And it’s not like this is my first rodeo, you know?

But I think you’re on the right track, using a simple air-cooled engine to develop this stuff. I need to find a vertical mount Briggs or Tecumseh, that I can set up as a testbed for these purposes. My very good friend happens to have such a rig, already fixed to a slab of wood. Might be time to head over and borrow that item from him – at least then, I can look at dynamic timing curves & ignition power on an actual runner. If not, maybe another junker mower will show up in the back alley.

ETA: Did you collect circuit data from any of the ATOM modules, other than the ‘red’ unit we’ve been looking at? It would be great to build a table that links color / application and circuit values..

[debe]

Blue module R1-1k,R2-2k,R3-1k, C1-.15uf. Brown module R1-1k,R2-2.1k,R3-1k,C1-.47uf.

[debe]

My BU931T transistors arrived today from HongKong, about $12 Aust for 10 freight free. Testing it on my test unit they work fine upto about 2000 RPM then stop sparking until you drop the RPM. If you add the Mosorb in the original circuit they work fine at higher RPM, My unit only goes up to 3000RPM. So they should be fine so long as theres a protector used.

[legendre]

quote debe:

My BU931T transistors arrived today from HongKong, about $12 Aust for 10 freight free. Testing it on my test unit they work fine upto about 2000 RPM then stop sparking until you drop the RPM. If you add the Mosorb in the original circuit they work fine at higher RPM, My unit only goes up to 3000RPM. So they should be fine so long as theres a protector used.

Now that’s interesting..

Do you have any sense as to why the Mosorb raises the upper speed limit (essentially, gives you more HF bandwidth)? At the moment, I can’t think of any possible mechanism, other than some device-specific quirk.

Electronics can do the craziest things.

[debe]

The only thing I can think of is this particular Wipac Ignition system has particularly high voltage spikes that the transistors don’t like. If you want long term reliability I think the Mosorb is definitely worth fitting.

[Author]

Posts