356 Spark plug gap and compression ratio

[Fred Winterburn]

Bored again and thinking of a possible change to the 6V CDI. Right now I'm keeping the maximum voltage quite low on the 6V models since some old 6V cars it might go onto are pre-war era. That accounts for maybe 5 or 6 of the 6V units I've built. Most have gone onto VW beetles and 356 cars. My question is: What is the recommended plug gap specified for the various models you have and what is the compression ratio? So far I haven't heard of any missing or hiccups but I was wondering if a bit more voltage margin is warranted. Right now with a typical 6V coil, the maximum voltage with 6V applied is approximately 22kV. Thanks, Fred

[Martin Benade]

Why do you keep the voltage low? Out of deference to an old coil or wires?

[Fred Winterburn]

Yes, and small diameter distributor caps, but mainly for antique coils and wire types. Fred

Why do you keep the voltage low? Out of deference to an old coil or wires?

[David Jones]

Fred, I have used as much as 43 thou gap with the Permatune CDI but realized that if I were to have a failure of the CDI while driving just changing back to the stock Kettering system would leave me with the potential of no spark because of the large plug gap so I now compromise with a gap of 30 thou with NGK iridium plugs. Big bore kit and less than 9 to1 CR. I run 87 octane gas.

[Al Zim]

The lower the voltage across the points the longer the points will survive. Carbureted 911's started on 12 volts then ran on 6 volts. Check your shop manual to see the way Porsche devised this system. It was very simple and allowed a hotter spark at starting. If the plugs are firing then there is no reason to alter the gap from the factory settings. CHECK THE RESISTANCE IN THE PLUG WIRES AND ROTOR

[Fred Winterburn]

Al, You're on a different subject. There is only 6V across the points with the 6V CDI and the current is limited by a resistor in the CDI trigger circuit. I was referring to coil secondary voltage. Actually a 12V ballasted system usually is more likely to be 8V applied to the coil once the ballast resistor is in the circuit after starting. I really don't think Porsche devised that system as it was very common on American cars, especially V8s. It's high current through the points that will turn them blue. Pitting on the other hand has to do with the energy available and the voltage subjected to the points as they open which can be a couple hundred volts or so and even higher on a misfire. Fred

The lower the voltage across the points the longer the points will survive. Carbureted 911's started on 12 volts then ran on 6 volts. Check your shop manual to see the way Porsche devised this system. It was very simple and allowed a hotter spark at starting. If the plugs are firing then there is no reason to alter the gap from the factory settings. CHECK THE RESISTANCE IN THE PLUG WIRES AND ROTOR

[Tom Wavrin]

Fred, at this point in time most of our cars have been modified in multiple ways along with changes in gasoline. The pre-A (through 1955) had compression ratios from 6.5 in the Normals to 8.2:1 in the Supers. In 1965 the cars CR ranged from 8.5 in the C to 9.5:1 in the SC. Points gap 0.4mm or 0.016 inch. Plug gap was 0.5 to 0.6mm or 0.020 to 0.024 inch. The SC, per my Porsche manual, shows a plug gap that is brand specific from 0.020 to 0.028 in (0.5 to 0.7mm). I'd guess a lot of us are using gaps nowadays more toward 0.028 inch. My 1600S with CR of 9.2:1, NGK iridium plugs gap set at 0.030 inch and points at 0.016 in, along with your 6V CDI, starts and runs very well - thank you.

[Fred Winterburn]

Tom, Maybe I should leave well enough alone. My idea was to increase the oscillator frequency slightly and the current draw by about 15%. With this power supply that would increase the voltage by about 2 kV and spread the useful operating battery voltage range possibly by another 2 volts for greater rpm potential at low battery voltage. IE better voltage regulation with rpm and battery voltage. It would make it nearly on par with the 12V unit. To do this, I need to make a quasi bridge rectifier out of P and N channel mosfets (6 in total, 2 N channel and 4 P channel) rather than the two large silicon bridge rectifiers that I use now. Those rectifiers make the power supply polarity insensitive and also make the supply more efficient as well as preventing the power supply from drawing too much current if the car's voltage regulator fails. The problem is that the silicon diodes in the bridge rectifiers drop roughly 1V, while the total with mosfets would be 0.1 volts. On a 6V system 0.9 V is a huge percentage. I do not want to change the transformer design as that has already been optimized for this voltage range. I could completely redesign everything, but I like the principle of using the least number of components to achieve the desired result. And then again, it has worked the way it is on every 6V car so far, so back to leaving well enough alone. Fred

Fred, at this point in time most of our cars have been modified in multiple ways along with changes in gasoline. The pre-A (through 1955) had compression ratios from 6.5 in the Normals to 8.2:1 in the Supers. In 1965 the cars CR ranged from 8.5 in the C to 9.5:1 in the SC. Points gap 0.4mm or 0.016 inch. Plug gap was 0.5 to 0.6mm or 0.020 to 0.024 inch. The SC, per my Porsche manual, shows a plug gap that is brand specific from 0.020 to 0.028 in (0.5 to 0.7mm). I'd guess a lot of us are using gaps nowadays more toward 0.028 inch. My 1600S with CR of 9.2:1, NGK iridium plugs gap set at 0.030 inch and points at 0.016 in, along with your 6V CDI, starts and runs very well - thank you.

[David Jones]

Fred, could you use the 12 volt unit but use a 6 to 12 volt converter to feed the CDI? The output would be a constant 12 volts which may be an advantage for voltage stability. That way you need do no more R&D.

[Fred Winterburn]

Could, but not without unintended consequences. The converter would become part of the circuit for some phases of the spark profile which might not play well with whatever electronics it uses. Also, I doubt that very many 6v t0 12v converters work well below 5V. They don't need to get the car started. The 6V CDI in its present form will do an easy 18kV at 3.2V ( I advertise 3.5V). The starter motor won't work at that battery voltage but it will bump start easily. The trouble is that the inverter frequency is low at 3.2V and although it will start the car, the battery voltage needs to be increased for increasing rpm. That's the nature of a free running oscillator; the higher the voltage, the higher the frequency. The voltage is more easily controlled than the frequency which is charging system voltage dependent. And it's the frequency that determines the rpm capability. It's still a better method (IMO) than the one shot charging method of the MSD or later permatunes that charge the capacitor in one half cycle rather than several cycles. Those CDIs give the same spark energy and voltage regardless of rpm but the voltage drops off nearly linearly with battery voltage and they tend to have very short duration sparks. The oscillator design if done right, runs at a high frequency so the rpm voltage drop off is low with decreasing battery voltage and increasing rpm, yet can maintain spark voltage even at low charging system voltages or low battery voltage for starting purposes. That's the basic difference between the two types other than spark duration, spark polarity (alternating or not) etc. Fred

Fred, could you use the 12 volt unit but use a 6 to 12 volt converter to feed the CDI? The output would be a constant 12 volts which may be an advantage for voltage stability. That way you need do no more R&D.

[Fred Winterburn]

Here I was touting the advantages of an oscillator type CDI. Well, here's an example of one that was poorly executed. It's called the Hot Box. It was built in the Seattle area in the late sixties by the look of the components and was sold out of a Porsche owner's collection. 12V negative ground. This came with a rev limiter that doesn't work and probably never did very well. The transformer is an off the shelf tube filament transformer and the oscillator frequency is approximately 50Hz at 12V. At low rpm it produces 70 thousand volts and then runs out of steam at about 12000 volts at 333 sparks per second. The frequency should be 300 times that to be any good. It's a cool looking unit though and I think only a handful would have been built. See the pic.

Could, but not without unintended consequences. The converter would become part of the circuit for some phases of the spark profile which might not play well with whatever electronics it uses. Also, I doubt that very many 6v t0 12v converters work well below 5V. They don't need to get the car started. The 6V CDI in its present form will do an easy 18kV at 3.2V ( I advertise 3.5V). The starter motor won't work at that battery voltage but it will bump start easily. The trouble is that the inverter frequency is low at 3.2V and although it will start the car, the battery voltage needs to be increased for increasing rpm. That's the nature of a free running oscillator; the higher the voltage, the higher the frequency. The voltage is more easily controlled than the frequency which is charging system voltage dependent. And it's the frequency that determines the rpm capability. It's still a better method (IMO) than the one shot charging method of the MSD or later permatunes that charge the capacitor in one half cycle rather than several cycles. Those CDIs give the same spark energy and voltage regardless of rpm but the voltage drops off nearly linearly with battery voltage and they tend to have very short duration sparks. The oscillator design if done right, runs at a high frequency so the rpm voltage drop off is low with decreasing battery voltage and increasing rpm, yet can maintain spark voltage even at low charging system voltages or low battery voltage for starting purposes. That's the basic difference between the two types other than spark duration, spark polarity (alternating or not) etc. Fred

Fred, could you use the 12 volt unit but use a 6 to 12 volt converter to feed the CDI? The output would be a constant 12 volts which may be an advantage for voltage stability. That way you need do no more R&D.

[Wes Bender]

Folks back then were experimenting a lot. Most were not successful. Although they were pretty crude, I built several Delta units from kits. All were still operating when I sold the vehicles they were in.

[Fred Winterburn]

Deltas were good, although they still ran the voltage higher than they should have to give the units enough steam at high rpm. The Mark 10 infringed on my father's watered down patent. The 10B did not. The most failures I've seen with Deltas is with the transformer failing, second the SCR. Most were long lived. Sydmurs were an almost direct copy and did infringe without any doubt, but the couple I have taken apart failed due to the high voltage affecting the transformer and the output diodes. Fred

Folks back then were experimenting a lot. Most were not successful. Although they were pretty crude, I built several Delta units from kits. All were still operating when I sold the vehicles they were in.

[Neil Bardsley]

A big thank you to Fred for the CDI unit. Installed and running. Seems to run with a bit more zip but hard to quantify. Time to do a tune up

[Mike Wilson]

I, too, installed Fred's CDI unit. The engine does seem to run better, starts easy and I need to quantify it but, seems to be more fuel efficient. I'm using the iridium plugs gapped at .035.

My installation.

Mike