Turbo GpZ550 Ignition misfires starting above 10 lbs boost

I'm headed out the door so I'll try to keepo this short... well, I guess not.

A current limiting circuit using a transistor basically works the same as a ballast resistor. But it's automatic and temporary, in that the resistance is removed after the spark happens. When it senses the current has reached the required limit, the transistor goes from being a very low resistance up to being the exact resistance needed to maintain the current at the limit value. That resistance jumps up suddenly. It has the effect, by virtue of being in series with the coil (on the negative end of the coil), of dropping the working voltage on the coil. It causes the coil's connection to the ignitor to have a sudden upward jump. So you have a constant 19v on the positive side of the coil, but now you will have say 6v on the negative side of the coil.

THe reason I mention this jump in such detail is that it can often wreak havoc with electronic tachometers and dwell meters etc. They see the current limit action as a spark event, then when the real spark event happens, (yet another big jump in voltage at the coil-ignitor connection) it registers another spark event. Theus the RPM appears to double when it hasn't.

In our case, a single current limiting device applied to the positive side of the coils would work because the dwell does not overlap. So the circuit would just be oscillating between 0 ohms and say 1 or 2 ohms. A transistor circuit could do that. So yes, that is an option.

I suggested the HEI option in hopes there is a chance that a modern manufactured 7-pin module would have a higher primary voltage blocking capability to give you slightly higher voltage capability. Also, the 7-pin current limit won't be reached so we no longer have to worry about current limiting at all even with the 19v. (Due to the short dwell time.) But in the event that the dwell unexpectedly jumps longer than necessary, the 7-pin will handle it with the internal current limit.

Without using the current limiter, the 7-pin modules will not need much of a heatsink at all. If they get unexpectedly hot in use, then it means there is time when the dwell must be long (although that did not show up on the scope) and current limiter is getting some use. In that case you might have to mount it to a piece of sheet metal somewhere.

I would not want to replace those Fuji transistors, especially if they are working. I doubt they affect the timing circuit, but there is a tiny chance they might.

EDIT: I may have to revise some of the above about the HEI current limit not being reached, but I really have to leave at the moment.

Thanks to Lou for his continued discussion about current limiting.

The topic continues and I have some information to consider.

I just ran a test on a spare GpZ coil (2.6 ohm cold) using the 19V, 15A power supply. I bolted the coil to a thick bar of steel to simulate mounting on the bike.
I verified the coil resistance at 2.6 ohm and the power supply output at 19V.
I connected the coil's primary directly to the 19V supply and measured a steady state current of just over 4.5 A.
I then connected the coil directly to a battery charger and read just over 3.5 A.
The coil was connected to voltage long enough to read my digital multi-meter. Probably 1000 mSec.

Now the battery charger read 17.5 Volts with no load. I didn't check it under load but it was probably somewhat less.
Edit. I just ran the test again with the battery charger. Under load the voltage at the coil was 9.5 V.
If the above was a valid test it sure eases my mind about using 19V to power the coils.

Ooh, I guess I should have mentioned you don't really want to leave a coil connected for more than a second without a lot of cooling time between. And never more than a couple seconds ever.

The predicted current using the cold resistance measurement of 2.6 ohms would be 19 / 2.6 = 7.3 amps.
But as you saw, the current usually measures lower than that. Part of it is that the supply may have some sag, or you may be getting a little loss in the connecting wires, but a large portion of that difference is because the wire in the coil is heating up very rapidly, and to a pretty high temperature. This greatly increases the resistance and thus acts to limit the current. (This is what a true ballast resistor does as well). The problem is the coil is not really meant to do that outside of a normal range. When left connected for a long time, coils have been known to get hot enough to melt the plastic casing leaving nothing but a burnt-melted blob.

The charger trial you did gets closer to the predicted value. But with such a lower voltage applied, you would expect much less heat generation.

One thing about most old-fashioned chargers and maybe even some modern ones is that their output is in pulses. Pulses do strange things to the voltage and current readings and so don't always hold up to the pure DC versions of ohm's law that most people use. Using ohm's law with AC becomes slightly more intricate than for pure DC cases. Some voltmeters don't give as accurate results for voltages that have pulses in them. That's a voltage with pure DC component and an AC component. The meter might not get the AC component accurately.

You can test this by simply measuring the AC voltage output on the charger when disconnected. If it is anything over 1 volt, there is significant pulsing in the voltage.

I have a whole paper on it for anyone who cares. There's a lot of math there for details, but the gist can be gathered without all the calculus. s3.amazonaws.com/gpzweb/ACvoltmeter.pdf

loudhvx wrote: Ooh, I guess I should have mentioned you don't really want to leave a coil connected for more than a second without a lot of cooling time between. And never more than a couple seconds ever.

The predicted current using the cold resistance measurement of 2.6 ohms would be 19 / 2.6 = 7.3 amps.
But as you saw, the current usually measures lower than that. Part of it is that the supply may have some sag, or you may be getting a little loss in the connecting wires, but a large portion of that difference is because the wire in the coil is heating up very rapidly, and to a pretty high temperature. This greatly increases the resistance and thus acts to limit the current. (This is what a true ballast resistor does as well). The problem is the coil is not really meant to do that outside of a normal range. When left connected for a long time, coils have been known to get hot enough to melt the plastic casing leaving nothing but a burnt-melted blob.

I was powering my 19v Boost power supply with the battery charger.
I repeated the test and read 13.5V under load at the coil. That explains the 4.5A current
The charger was probably dropping down to 9.5 V while it supplied the 19V converter.

Back to testing but this time I will power the 19V converter with a fully charged 12V battery.
Voltage and current at the coil should be much higher.

I will be back to update this post with the new data.

I see, so the boost supply has significant sag. I was going to ask more about the details of this boost supply. With that much sag I wonder if it will even be of much help.

The 13.5V / 4.5A = 3 ohms is a more reasonable match.

I think you may want to go with slightly faster coils and just do away with the 19v boost thing. You are going to want new coils anyway as old stock coils might not be too enthusiastic against the higher cylinder pressure. With two HEI modules and say a pair of 1.5 to 2 ohm coils, you will just start to tap into the current limit of the HEI. The lower inductance of the faster coils will be a better match for CDI as well.

I've been looking at Herlux Herko B116 coils. They are just some generic Hyundai car coil replacements. But they are nice and small and run about 1 ohm. They would need a ballast or decent limiter, but they are cheap and small. I have one for testing, but I don't know what kind of voltage they can put out yet. I haven't done much other than map out the charge rate (it's on the coil testing link).

I was only considering them for norm. aspirated, 10:1 comp.
Preliminary tests were very good. Nice long sparks.

Look at the price!
www.amazon.com/Herko-B116-Ignition-Hyund...0-2002/dp/B01MPVPQT0

loudhvx wrote: I've been looking at Herlux Herko B116 coils. They are just some generic Hyundai car coil replacements. But they are nice and small and run about 1 ohm. They would need a ballast or decent limiter, but they are cheap and small. I have one for testing, but I don't know what kind of voltage they can put out yet. I haven't done much other than map out the charge rate (it's on the coil testing link).

I was only considering them for norm. aspirated, 10:1 comp.
Preliminary tests were very good. Nice long sparks.
www.amazon.com/Herko-B116-Ignition-Hyund...0-2002/dp/B01MPVPQT0

WOW that is affordable.
I had been looking at a VW coil pack that has internal coil drivers. It has 4 separate coils so they would have to be triggered in pairs for wasted spark. $29.
Not sure on the coill resistance so also not sure if they could handle the zx ignitor's 5 msec dwell at idle.

If you look at the chart you see the graph has a slope inversion at about 2 msec. That is the range at which the magnetic saturation is starting to happen so there is a decreasing return in continuing charge after that time. The slope of current/time goes from being gradually decreasing to gradually increasing. That happens to coincide with about 4 to 5 amps. So that is basically perfect for high-RPM operation.

s3.amazonaws.com/gpzweb/Ignition/Ignitio...s/IgnitionCoils.html

But yes, that is not so great for low RPM. A lot of current limit time will be used.

Maybe just hedge it with a .5 ohm ballast, and let the HEI modules do the rest of the limiting, and mount them on a metal plate.

I don't think I would use them direct to the Zx igniter unless you run a 1 ohm ballast. But then at high RPM you cut the current in half. I'll have to think about it a bit more.

At any rate... more pressing... there are only two left in stock. If you don't want them, I'll probably grab them. At that price, one of us should grab them. Then I can blow up the one I already have to see if it can give me a 1-inch spark.

No eBay APP ID and/or Cert ID defined in Kunena configurationHere's another link to those B116 coils

I just ran the coil/current test again using the 19V boost converter supplied by my 90% charged MGB 12 V battery.

This time I read 6.2 amps through the 2.6 ohm (cold) Gpz ignition coil.
That would be at 100% duty cycle.
Not sure how responsive my meter is but a very momentary circuit connection, by hand, didn't reach much over 5 amps.

6 msec at 1000 rpm would be about a 10% duty cycle?
And 2.6 msec at 10000 rpm would be a little over 40% duty cycle but at a lower current.

Maybe the ignitor would be OK?

He also only shows 2 left. I wonder if it's the same seller.

So the boost converter seems to have a lot of voltage sag under load. That's not really a good reliable factor to try to deal with if it's unpredictable.

Generally speaking you want to try to stay under the limits of any published specs for a component. I usually like to have a 100% margin or another way to put it is only use about 50% of the rated spec. That "absolute maximum" Ic current rating is staring us in the face pretty hard. I don't think I would want to push it to 6.2A. Now, I can't really read the Japanese in that pdf, so is that rating average or instantaneous? I usually assume it's instantaneous which means you would be over the limit regardless of duty . Earlier I threw out the possibility that it could be an average rating, but without knowing for sure, you have to assume instantaneous.

I did find an English PDF of the datasheet that refers to the max collector current as "continuous"
It doesn't provide a separate pulsed current limit but I have seen spec sheets for other Darlington Pairs that do.