Combat performance issue - RITA

[cliffa]

Continuing the process, I have encountered 3 challenges that make this upgrade not as easy as it could be.

The first 2 are issues because Rex's kit did not include a plastic connector terminal housing or rubber wire interface seal, both of which are integral to the wires soldered to the old circuit board, and will have to be safely removed in order to be re-used.
Removing the old male terminals from the housing was easy after I made an extractor out of a spare female terminal.
View attachment 101205

My plan for the wire seal (shown in the photo above where the wires reach the circuit board) is to unsolder the wires that attach to the board, remove the crimped on shrouds that terminate them and pull the wires thru the rubber seal. Hopefully it won't be too difficult to feed the new wires back thru the seal.

The 3rd challenge is how to eliminate the interference that will exist between the new circuit board and the terminal of the component that is mounted in the housing. The old board has a hole in it for the terminal and this hole allows the terminal to protrude above the circuit board without interference. The new board has no such hole and none can be drilled there without breaking printed circuit connections. My current idea for a solution to this is to remove the component from the housing by machining off the 2 rivets that attach the component's mount to the housing. Will have to make sure these holes remain sealed.
View attachment 101209
This one is only a challenge because one of my objectives is to be able to revert to an old style circuit board in the future, if desirable for some reason. Otherwise, I would just cut the terminal off.

If anyone has any ideas for more clever ways to deal with these issues, please let me know.

Thanks for the updates Nick.
Before doing anything, I would make sure that the seal is not bonded to the wires otherwise you will be wasting your time trying to get it off.
If you want to be able to revert to an old circuit, I would machine one of the rivets off then drill through the mounting plate and the case, countersink the outside, fit a countersunk screw and nut tight, then do the same to the other rivet. Once you have both drilled, remove the screws and component and fill the countersunk holes with silicone.

 

[NickZ]

Thanks for the updates Nick.
Before doing anything, I would make sure that the seal is not bonded to the wires otherwise you will be wasting your time trying to get it off.

Good point. They appear to be bonded because I haven't been able to make them slide in or out.
However, Rex's says on their fitting guide that it is supposed to be removed for re-use. I have emailed their tech support.

If you want to be able to revert to an old circuit, I would machine one of the rivets off then drill through the mounting plate and the case, countersink the outside, fit a countersunk screw and nut tight, then do the same to the other rivet. Once you have both drilled, remove the screws and component and fill the countersunk holes with silicone.

Not sure I am following you here. What is the reason for countersink?

 

[cliffa]

Good point. They appear to be bonded because I haven't been able to make them slide in or out.
However, Rex's says on their fitting guide that it is supposed to be removed for re-use. I have emailed their tech support.


Not sure I am following you here. What is the reason for countersink?

Countersunk screws would be flush with the case and would be easier to seal until required.

 

[tricatcent]

I have a Rita on my Norton 850 for the last 20,000 miles. It runs great. They do consume a lot of power though. Since I have a 3 phase alternator the higher power use doesn't matter. I set the full advance timing at 28 degrees. It is a touring bike so it runs most often between 3000 and 4000 RPMs. It pulls really well up to 6000 RPM in the lower gears. It passed rows of Harleys on the Bear tooth pass, with a passenger and luggage. I also ran a Rita on my Vincent for a while. That bike runs 34 degrees full advance. The box I use on that one has a different number on it that provides a lot more spark advance. Timing at kick over speeds is only about 4 degrees BTDC. If I used this box on the Norton it wouldn't work right.

 

[cliffa]

Rex’s claim that power consumption is greatly reduced with their replacement system.

 

[baz]

Rex’s claim that power consumption is greatly reduced with their replacement system.

Yes they claim it's lower
But they won't tell you what the consumption is,I have asked them in the past

 

[Thomasa]

This discussion makes me sad that I binned a Rita ignition years ago when it failed. One of the PNP driver transistors failed and on analyzing the specs on that transistor one could see why. The peak, inverse voltage rating was not up to snuff. There was an NPN transistor rated high enough but not a PNP which the circuit required. Apparently someone has worked out a fix. I thought and still think that the Rita is superior to the Boyer.

 

[SteveA]

This discussion makes me sad that I binned a Rita ignition years ago when it failed. One of the PNP driver transistors failed and on analyzing the specs on that transistor one could see why. The peak, inverse voltage rating was not up to snuff. There was an NPN transistor rated high enough but not a PNP which the circuit required. Apparently someone has worked out a fix. I thought and still think that the Rita is superior to the Boyer.

I had the big power transistor fail a couple of times. This happened if the spark plug was not connected well enough to properly load the circuit.

We figured a TV line switching transistor would fix it, and it did, but it was no more robust than the original item, hence reoccurrence.

The old tech transistors drew a lot of current, particularly that final switching transistor, which is why the case was a substantial heat sink, earlier ones were even finned to help disperse heat.

So there will be little doubt that more or less the same circuit implemented with more modern transistors will draw less current. As can be seen, there is no need to bond the new board to the case for cooling purposes.

 

[maylar]

So there will be little doubt that more or less the same circuit implemented with more modern transistors will draw less current. As can be seen, there is no need to bond the new board to the case for cooling purposes.

A minor nit-pic (sorry, it's the EE in me): The current is the same (ignition coil load) but the "on" resistance and therefor power dissipation of modern transistors is much lower, generating less heat.

 

[Thomasa]

A minor nit-pic (sorry, it's the EE in me): The current is the same (ignition coil load) but the "on" resistance and therefor power dissipation of modern transistors is much lower, generating less heat.

If I understand what you're saying, V(sat) is lower. That is , the voltage drop across the switching transistor at saturation. This has also been an issue with "rebuilt" alternators with internal regulators (looking at you GM). Cheap ones have a higher Vsat and of course the heat shortens their lives when loaded. Most rebuilders use the cheapest regulators they can find.

That apparently wasn't the issue with Rita. There simply wasn't a suitabley high Peak-Inverse-Voltage, PNP transistor available at the time. And also as I understand, it the failure mode was junction punch-through because of flyback voltage from the coils rather than heat. Would not a protection diode have solved the problem. One sees those all over automotive circuitry these days, or last I looked anyway. Seems like that would have been an obvious fix once they understood the problem. Is that the fix that Rex's has implemented? One supposedly Rita circuit diagram I found shows just such a diode but all the transistors are NPN and the call outs are in French. The input is labeled 12V not B+ or +12V. It's not a complex circuit, really.

Are other electronic ignitions using variations on the Rita theme or are they using a capacitive discharge design?

I know they made these units for Jaguar; did they fail in the same way?

 

[maylar]

If I understand what you're saying, V(sat) is lower. That is , the voltage drop across the switching transistor at saturation.

Bipolar power switching transistors have largely been replaced with power MOSFETs. There is no V(sat) since a FET is a resistive channel. The on voltage is a function of RDS(on) and some are under .01 ohms at 10 amps.

There simply wasn't a suitabley high Peak-Inverse-Voltage, PNP transistor available at the time. And also as I understand, it the failure mode was junction punch-through because of flyback voltage from the coils rather than heat.

The failure mode was referred to as secondary breakdown, and it's a consequence of switching an inductive load. Transistor manufacturers did mega R&D in the 70's to address that as solid state off-line converters, TV horizontal amplifiers and electronic ignitions were becoming common.

Would not a protection diode have solved the problem.

A reverse diode clamps the "inductive kickback" but doesn't solve secondary breakdown. Fortunately MOSFETs are not plagued by that.

Sorry for the derail, but I designed power supplies for a living in the 70's and was deeply involved in this stuff.

 

[Thomasa]

By "secondary breakdown" do you mean the effect of the "ring" as the voltage reverses? I'm not familiar with that term. Clearly I have not kept up with circuit design in the past couple of decades.

 

[NickZ]

Bipolar power switching transistors have largely been replaced with power MOSFETs. There is no V(sat) since a FET is a resistive channel. The on voltage is a function of RDS(on) and some are under .01 ohms at 10 amps.


The failure mode was referred to as secondary breakdown, and it's a consequence of switching an inductive load. Transistor manufacturers did mega R&D in the 70's to address that as solid state off-line converters, TV horizontal amplifiers and electronic ignitions were becoming common.


A reverse diode clamps the "inductive kickback" but doesn't solve secondary breakdown. Fortunately MOSFETs are not plagued by that.

Sorry for the derail, but I designed power supplies for a living in the 70's and was deeply involved in this stuff.

No need for derail apology. To me , this is interesting and relevant stuff. Probably helps that I also have EE background, but have forgotten most of it.

 

[maylar]

By "secondary breakdown" do you mean the effect of the "ring" as the voltage reverses? I'm not familiar with that term. Clearly I have not kept up with circuit design in the past couple of decades.

Secondary breakdown is a failure mode in bipolar transistors in which negative resistance (current concentration) occurs under high-voltage and high-current conditions. Current concentration causes local heating, resulting in a small hotspot. Thermal runaway quickly ensues.

When I was designing power supplies we had a rule for switching regulators: nothing over 24 volts and nothing faster than 25 khz. Transistor technology hadn't advanced yet to a point where they would survive the stress.

 

[Thomasa]

25 Kilohertz! Modern Switching power supplies raised all kinds of harmonics problems with the utilities, till they figured out the cause. Blown up 3-phase transformers at Microsoft for one example.

And to think today we have phones in our pockets with solid-state circuits operating in the Ghz range. On my first tech job, at Ft. Monmouth NJ, I taught radar theory and repair. The only solid-state components in our mortar-locating radar was a stack of "door knob" diodes that were part of a voltage septupler in the thyratron driver circuit that triggered the magnetron. The mag operated at 16.5 Ghz IIRC. You could see rain coming on the CRT scope. You could also see .50cal machine-gun rounds but you couldn't track them. Fortunately, in combat they were mostly going away from you; not so, the mortars and rockets. A modern version of this radar is presently being used in Ukraine to find Russian artillary. A very handy device. But I digress...

 

[NickZ]

Finally getting back to RITA rebuild with Rexs circuit board.

I had to fabricate a new grommet that the wires go into the unit through because I could not get the old wires to move out of the old grommet. Rex's documentation says the old grommet is to be re-used, and when I contacted their tech support they stuck with that position, but no way I could get that to work.

I milled off the 2 rivets used to mount the discreet component to the housing so it could be removed and not interfere with the new pcb. #8 screws to close the holes.

Crimped on new connectors and did a bench test. It sparks! I elected to not use the terminals supplied with Rexs kit to replace the obsolete 560 type connector used in the original installation and ran the wires directly to the coils & pickup. This eliminates a set of connections that aren't really necessary.

Sealing everything with silicone and refit to bike. Unfortunately, I won't be able to get back to this for another 10 days to set the timing with the strobe. Also, I will need to find some assistance to do that properly at 6000-6500rpm.

 

[NickZ]

P.S.
I now believe the component removed from the housing was not a zener, but instead a Metal Oxide Varistor, or MOV, which would provide over-voltage protection for the output transistor that switches the coils. It is connected between the collector and emitter of the big transistor which is connected to the coils.

 

[motorson]

I had a RITA on my bike for thousands of miles and loved it. My understanding is that they consume more power than all the others because of the fact that they saturate the coils for a lot longer before firing the spark. That makes for a really hot spark though. When I was stopped in traffic, I could feel that the coils were getting really hot.
I fried that RITA on a really long ride in the rain from Kentucky to Ashville NC. My voltmeter kept telling me I was overcharging, and the RITA fried.
Now I have a Pazon on the bike. I like it but I don't think it has as aggressive of an advance curve as the RITA. I called those guys a Rex's to ask about the advance curve of their replacement circuit board. They would only say that it was a direct replacement.

 

[baz]

I had a RITA on my bike for thousands of miles and loved it. My understanding is that they consume more power than all the others because of the fact that they saturate the coils for a lot longer before firing the spark. That makes for a really hot spark though. When I was stopped in traffic, I could feel that the coils were getting really hot.
I fried that RITA on a really long ride in the rain from Kentucky to Ashville NC. My voltmeter kept telling me I was overcharging, and the RITA fried.
Now I have a Pazon on the bike. I like it but I don't think it has as aggressive of an advance curve as the RITA. I called those guys a Rex's to ask about the advance curve of their replacement circuit board. They would only say that it was a direct replacement.

So they won't tell you what the advance curve is
And they couldn't tell me what the power consumption is !
The thing is
If I knew for sure it'd be as reliable as my 40+ year old one and I knew it had a definite lower power consumption
And it was the same advance curve I'd probably buy one

 

[ILLF8ED]

So they won't tell you what the advance curve is
And they couldn't tell me what the power consumption is !
The thing is
If I knew for sure it'd be as reliable as my 40+ year old one and I knew it had a definite lower power consumption
And it was the same advance curve I'd probably buy one

I guess to be fair Lucas didn’t let anyone know the advance curve of the original RITA. We got that information from Dave Comeau aka DynoDave on his Atlantagreen website. Modern MoSFETs will have lower power consumption than the original transistors.