Resistor & grounding CDI ignition modules

[petertha]

I started Googling what the resistor was all about on typical RCEXL type CDI ignitions & it looks to be quite a controversial rabbit hole. Some say to prevent arcing issues if you cant rely on the ground return path through the crankcase assembly. Others say its for RF issues related to RC flying installations. Others say related to the plug itself. Others say don't need it if you have shielded cables (yet some vendors typically supply harness with resistor AND shielded cable EVEN WITH non-resistive plugs). And yet I see many examples here & even commercial model engines with slim-ish high strand (high kv?) wires, no resistor, no braided shielding, engine grounded distributor on the same 1/4-32 RCEXL type spark plug & RCEXL type CDI module. Some build I cant find anymore had/suggested a slim ground wire heat shrunk to the main harness & that solves some problems. My electrical knowledge is low, just wondering if there were any words of wisdom or experiences to share specific to this brand of module.

https://www.rcuniverse.com/forum/gas-engines-142/10898249-rc-xcell-ignition-resistance.html
https://ch-ignitions.com/product/repair-kit-1-4-32-90-degree-ngk-me8-rimfire-vr2-rcexl-spark-plugs/

 

[mayhugh1]

I started Googling what the resistor was all about on typical RCEXL type CDI ignitions & it looks to be quite a controversial rabbit hole. Some say to prevent arcing issues if you cant rely on the ground return path through the crankcase assembly. Others say its for RF issues related to RC flying installations. Others say related to the plug itself. Others say don't need it if you have shielded cables (yet some vendors typically supply harness with resistor AND shielded cable EVEN WITH non-resistive plugs). And yet I see many examples here & even commercial model engines with slim-ish high strand (high kv?) wires, no resistor, no braided shielding, engine grounded distributor on the same 1/4-32 RCEXL type spark plug & RCEXL type CDI module. Some build I cant find anymore had/suggested a slim ground wire heat shrunk to the main harness & that solves some problems. My electrical knowledge is low, just wondering if there were any words of wisdom or experiences to share specific to this brand of module.

https://www.rcuniverse.com/forum/gas-engines-142/10898249-rc-xcell-ignition-resistance.html
https://ch-ignitions.com/product/repair-kit-1-4-32-90-degree-ngk-me8-rimfire-vr2-rcexl-spark-plugs/

That resistor is to reduce the emi radiated by the plug wires. It's purpose is to reduce the Q (quality factor) of the inductance associated with the plug wires which can radiate interference to nearby electronics. It's more important in RC applications where the radios are susceptible. In our display models it's rarely a problem except for the one builder I followed on another forum who built a fuel injected model V8 and had a major difficulty with plug wire interference sending his Mega-Squirt controller off into the weeds. (The biggest problem was likely with the laptop he was using to monitor and control the Mega-Squirt. Running shielded wires (grounded at one end only) can be an even more effective solution especially when used with a resistor. Again, over-kill for a running display model.

Some of the plug wires in full-size engines I've owned used a carbon core instead of a wire plus a resistor in order to reduce AM radio interference. The resistor even at 10k doesn't rob the ignition of any significant power because there is only a negligible voltage drop associated with it. - Terry

 

[petertha]

Thanks Terry, that helps a lot. The interweb is useful once you know what you're searching for! There must be some reason to get the resistance nearest to spark source (in the plug) vs equivalent resistance in the plug wire itself. Just relating to with/without resistor I just noticed the CG specs says right there - no resistor. Maybe some of the larger format (model use / weed wacker plugs) do have resistors, but I was specifically interested in smaller 1/4-32. This might explain another source of variable results among modelers.

Small Spark Plug used on small gasoline engines up to 40 CC single cylinder.
Thread Size: 1/4-32 same thread as standard Glow Plugs.

• Thread Size: 1/4 Inch
• Thread Pitch: 32 inch
• Seat Type: Gasket
• Resistor: No
• Reach: 0.25″(6.25 mm)
• Hex Size: 5/16″(8 mm)
• GASOLINE Gap: 0.02″ (0.5 mm) CH Recommended
• Glow Gap: 0.012″ (0.3 mm) CH Recommended

 

[DKGrimm]

I started Googling what the resistor was all about on typical RCEXL type CDI ignitions & it looks to be quite a controversial rabbit hole. Some say to prevent arcing issues if you cant rely on the ground return path through the crankcase assembly. Others say its for RF issues related to RC flying installations. Others say related to the plug itself. Others say don't need it if you have shielded cables (yet some vendors typically supply harness with resistor AND shielded cable EVEN WITH non-resistive plugs). And yet I see many examples here & even commercial model engines with slim-ish high strand (high kv?) wires, no resistor, no braided shielding, engine grounded distributor on the same 1/4-32 RCEXL type spark plug & RCEXL type CDI module. Some build I cant find anymore had/suggested a slim ground wire heat shrunk to the main harness & that solves some problems. My electrical knowledge is low, just wondering if there were any words of wisdom or experiences to share specific to this brand of module.

https://www.rcuniverse.com/forum/gas-engines-142/10898249-rc-xcell-ignition-resistance.html
https://ch-ignitions.com/product/repair-kit-1-4-32-90-degree-ngk-me8-rimfire-vr2-rcexl-spark-plugs/

Petertha,

You raise a very interesting and important question. . . . but not an easy one. I bought a 1Ghz sampling scope and tried to do a little research. I found out early on that the noise from the spark current coupled into grounds and wires and even into the scope itself, and I really could not make decent measurements because the noise spikes would saturate the scope amplifiers. I found at using a resistor at the spark plug reduced that noise enough to get most of the measurements I was after.

I spent some time trying to figure out why this happened. Here are some pictures of measured voltage and current at the spark plug during the very beginning of the spark. This was done with and without resistor. Keep in mind that the time scale is very, very short, only a tiny fraction of the total amount of the whole spark.

My best guess as to what is happening is that:
a. The spark fires and conducts LOTS of current in an astonishingly short time, less than 10 nanoseconds.
b. This very quick current flow of current is very large, peaking at 2.5 amps.
d. The discharge pulse rings like crazy at an extreme frequency, around 80 MHz in the picture above.
e. The voltage swing of this pulse peaks at about 250 kvolt (almost incredible) for a few nanoseconds.
f. The effect of the resistor is to all but eliminate that initial spike of current.
g. Short spike current and voltage are reduced by over 10:1.
h. The amount of energy involved in this short time is only a tiny fraction of the total spark energy.

I believe the charge that supplies this initial current spike can only be in the stray capacitance of the HV plug lead. The inductance of the HV lead itself would block current with this risetime from any other source.

I know this is a lot more than you asked, but I thought it was interesting enough to share on the sting.

Don

 

[petertha]

Thanks for this Don. Electronics is not my strong suite for sure, but the charts & your description are quite revealing to this laymen.

- now would this issue 'scale' in that model engine A uses a commercial RCEXL CDI and engine B uses a car or motorcycle coil & so different voltage/current levels might be different? Would RCEXL type CDI be better or worse from energy standpoint or no different on this basis?

- when I see these tabletop Chinese engines it 'appears' they use 1/4-32 RCEXL plugs (no resistor), no shielding/ground on cable to distributer & no resistor between strand wire & plug. The no resistor is just an unsubstantiated guess based on occasional glimpses of the parts laid out, I just see a simple connection clip. So it obviously works, the engines run & there are no airborne/RC control/EMI issues to contend with. I thought I read somewhere in a homebrew resistor-less setup there was enough noise to spasm the CDI module itself & they had to revert to shielding/resistor (not Terry's example, I might be wrong but I think he was referring to YouTube Keith57000 which had Megasquirt related electronics, different again). When you do your scoping, is it in plug proximity or CDI module proximity or doesn't really matter?

 

[DKGrimm]

- now would this issue 'scale' in that model engine A uses a commercial RCEXL CDI and engine B uses a car or motorcycle coil & so different voltage/current levels might be different? Would RCEXL type CDI be better or worse from energy standpoint or no different on this basis?

As far as this very fast pulse when the spark first fires, I don't think the source of the spark would make any difference.

- when I see these tabletop Chinese engines it 'appears' they use 1/4-32 RCEXL plugs (no resistor), no shielding/ground on cable to distributer & no resistor between strand wire & plug. The no resistor is just an unsubstantiated guess based on occasional glimpses of the parts laid out, I just see a simple connection clip. So it obviously works, the engines run & there are no airborne/RC control/EMI issues to contend with. I thought I read somewhere in a homebrew resistor-less setup there was enough noise to spasm the CDI module itself & they had to revert to shielding/resistor (not Terry's example, I might be wrong but I think he was referring to YouTube Keith57000 which had Megasquirt related electronics, different again). When you do your scoping, is it in plug proximity or CDI module proximity or doesn't really matter?

I was working as a mechanic in my dad's garage when resistor plugs or resistor harnesses were introduced to the automotive world. Cars all had AM radios; FM had not been put in cars very much yet. Radio interference seemed to be the main issue. I don't think it matters to the spark plug, but these pulses certainly do couple into any kind of nearby electronics. For one example, lots of people have had Hall effect timing sensors mysteriously fail if they weren't isolated from metal by a layer of some kind of insulation. I'm betting that this big initial spike was the real cause of these failures. As you say, those spikes could easily disrupt CDI electronics, so I can see the need for resistor plugs.

Shielding is another way to corral the interference from this initial pulse and keep it out of other circuits, but it is really has to be coaxial shielding with full continuity of the entire shield to the metal. Pigtail leads aren't worth much for containing pulses this short.

Don

 

[peterl95124]

I know this is a lot more than you asked, but I thought it was interesting enough to share on the sting.

Don

Wow!, now if you could apply this same level of measurement detail to atmospheric lightning we might get somewhere on that topic !!!

 

[DKGrimm]

Wow!, now if you could apply this same level of measurement detail to atmospheric lightning we might get somewhere on that topic !!!

Perhaps it is not a coincidence that I took courses in lightning protection and had that as one of my jobs for a time. It's amazing what a 100,000 amp lightning pulse can do. If you ever want a job with a surprise around every corner, that would be a good choice.

Don

 

[peterl95124]

Perhaps it is not a coincidence that I took courses in lightning protection and had that as one of my jobs for a time. It's amazing what a 100,000 amp lightning pulse can do. If you ever want a job with a surprise around every corner, that would be a good choice.

Don

as a one time spark coiler and tesla coiler I've always said you can't have too many high voltage toys, you might be close to the limit though !!!

 

[peterl95124]

important question. . . . but not an easy one. I bought a 1Ghz sampling scope and tried to do a little research.

DK, perhaps, when you have a little time, research what happens to the supply voltage to the Hall sensor during a spark (with the Sage-Gedde driver circuit), because my hunch is we're burning out our Hall sensors with the kickback from the coil primary and I'd like to see someone with a better scope than mine confirm or refute
Peter L

 

[DKGrimm]

DK, perhaps, when you have a little time, research what happens to the supply voltage to the Hall sensor during a spark (with the Sage-Gedde driver circuit), because my hunch is we're burning out our Hall sensors with the kickback from the coil primary and I'd like to see someone with a better scope than mine confirm or refute
Peter L

Peter,

I've spent quite a bit of time with the Sage-Gedde driver; I even made my own circuit boards and bought parts for six of them. I used that driver for most of the coil testing I did. I see you were part of the discussions in the thread "Tests of CID Ignition Modules," and this question of blown Hall sensors did come up somewhere in the middle of all that thread somewhere.

I can give you my opinion, even though I can't necessarily prove it. I doubt the driver circuit is the source of damage; instead, I believe the source of energy that damages the hall device is the extremely rapid discharge of the electric charge stored in stray capacitance of the secondary lead to the plug. The "generator" of that energy is the extremely short buildup of current in the arc when the plug fires. You can see such a pulse in the waveforms above. If a numerical example helps, assume the stray capacitance is 20 pf (micro microFarads) and that the spark breakdown voltage is 15 kv. Total stored energy is

That's a bunch! If that all discharges in about 150 nsec (as seen in the pictures above) then the average power delivered during that short pulse is almost an almost unimaginable 15 megawatts. The time is so short that lumped constant models of the circuit can't really describe what happens. One could try to model what happens using the geometry of the wires and parts and applying field theory, but that's beyond my pay grade. A large part of that energy goes into creating e-fields and h-fields, a lot of which is radiated as radio interference, and both fields -- especially the h-field -- couple into any other wiring that happens to be in the vicinity.

Bottom line: this burst of energy is absolutely dominant as a source of interference with other circuits. Semiconductors that are vulnerable to damage from static electricity (like Hall devices) don't stand a chance unless they are somehow protected from this pulse.

 

[peterl95124]

Peter,

I've spent quite a bit of time with the Sage-Gedde driver; I even made my own circuit boards and bought parts for six of them. I used that driver for most of the coil testing I did. I see you were part of the discussions in the thread "Tests of CID Ignition Modules," and this question of blown Hall sensors did come up somewhere in the middle of all that thread somewhere.

[ ... ]

Bottom line: this burst of energy is absolutely dominant as a source of interference with other circuits. Semiconductors that are vulnerable to damage from static electricity (like Hall devices) don't stand a chance unless they are somehow protected from this pulse.

DK,
I'm not buying it, because I've seen my supply voltage swing 90 V opposite, from the primary kickback <--- that's what I'd like to see if anyone else can replicate, if a device is designed for +5V but you send it -90V I'm guessing that's a problem.
Pete.

 

[DKGrimm]

DK,
I'm not buying it, because I've seen my supply voltage swing 90 V opposite, from the primary kickback <--- that's what I'd like to see if anyone else can replicate, if a device is designed for +5V but you send it -90V I'm guessing that's a problem.
Pete.

Hmmmm. sounds to me like something that could be induced by coupling from the spark wiring. I've even seen this magnitude of pulse coupled into other circuits, even the ground. Does a spark plug resistor of 10 kohms or so make it go away?

 

[DKGrimm]

Just happened to remember, the S/G driver has a blocking diode on the power input to block negative voltage swings. Also, try clipping the scope input lead to the S/G driver GROUND and test it again. Does the pulse go away? The pulse could easily be coupling into the scope lead or the scope ground, or both.

 

[peterl95124]

Hmmmm. sounds to me like something that could be induced by coupling from the spark wiring. I've even seen this magnitude of pulse coupled into other circuits, even the ground. Does a spark plug resistor of 10 kohms or so make it go away?

DK,
many thanks for the replies, currently up to my ears in alligators with other projects, but will get back to this, hopefully in the not too distant future
Pete.

 

[petertha]

Coincidentally, this video just got posted showing partial disassembly of a Toyan L-400G distributer, one of the Chinese tabletop engines I was mentioning. I was going to ask some questions but I suspect more to come whether he orders replacement parts or carries on with disassembly. Not super informative at this point but anyways, shows a rather basic rotor contact assembly. I'm not clear if the contacts when new should be rubbing or should be gapped.

Just eyeballing the wires, they seem like regular stranded copper conductor wires? I suspect he is running RCEXL type 1/4-32 spark plugs but from other pictures / similar engines I don't get the impression there is an inline resistor near the plug unless maybe its integrated axially into the wire & hidden within the heat shrink boot? Its certainly not the external braided grounded type of wire on RC systems.

I'm told by a friend that the early Bill Carpenter CH ignition modules used braided carbon ignition wires, likely off an existing ignition system but stripped of the thicker sheathing so it just had the thinner silicone type layer. He crimped his spark plug cable connector onto the carbon. But I'm not sure if the new CDI systems have this same wire or went to different conductor with external ground mode.

generic kit for RCEXL type metal/grounded boot showing inline resistor