Measuring voltage (and frequency) of spark

[petert]

Hello,

I experimented with a piezo crystal from a lighter to generate a high voltage spark. A coherer was able to detect the generated EM wave (from the spark/spark gap) in close proximity.

Now I am trying to generate a spark with a coil that has the same effect. I charge a coil/inductor with a 4.5V battery and when I break the contact or make it, a spark is generated as expected.

How could I measure the spark voltage (and frequency) between the two leads? I am worried it might damage my scope.

[Rigolon]

It would probably damage the oscilloscope, a spark (if I remember correctly) of 1mm has 3kV (in air).

Here is a link that I found useful when learning about this effect, it's called Inductor Kickback by the way.


I don't know of a method to actually measured it, but you could at least do some math to calculate approximately.

[Zero999]

Hello,

I experimented with a piezo crystal from a lighter to generate a high voltage spark. A coherer was able to detect the generated EM wave (from the spark/spark gap) in close proximity.

Now I am trying to generate a spark with a coil that has the same effect. I charge a coil/inductor with a 4.5V battery and when I break the contact or make it, a spark is generated as expected.

How could I measure the spark voltage (and frequency) between the two leads? I am worried it might damage my scope.

The voltage depends on the current and the length of arc. Spark gaps have an inverse current/voltage relationship. A kV may be required to initiate the arc, but once it's struck, the current will increase and the electrode voltage will decrease. If nothing limits the current, then there will be magic smoke as the cable and power supply start to overheat.
https://en.wikipedia.org/wiki/Electric_arc#Overview

Sparks produce broadband electromagnetic radiation. The lowest frequency will be dependant on the length of the pulse, with harmonics extending beyond the microwave band towards infrared and visible.

[iMo]

The first transmitters/receivers were made of spark gaps and coherers.
The frequency is set by the standard LC tank, or an tuned/resonant antenna (what is the same basically).
So the spark creates a broadband emission, the tuned antenna or LC tank will select the proper frequency out of it and radiate towards the receiver. The receiver should get the same tuned antenna or LC, of course, and the coherer should detect something at that frequency then..

https://en.wikipedia.org/wiki/Spark-gap_transmitter
https://en.wikipedia.org/wiki/Coherer

[T3sl4co1l]

You could measure it with an oscilloscope and an inductive probe (say by clipping the probe ground to the tip and holding that in proximity to the coil), or a capacitive probe (holding the probe tip near the coil, without directly contacting it, and with the ground clipped to the battery).

You won't get a calibrated measurement this way -- the vertical scale will be arbitrary, and the inductive probe will sense change in current (which is to say, effectively the voltage) rather than current directly.  But it's linear, so the reading will be proportional to what's actually on the coil, and you'll be able to pick out the frequency from that.

Tim

[tautech]

1000:1 probes for scopes are not that expensive.
Look up Joe Q Smith's DMM test videos where he uses grill ignitors and builds a 1000:1 probe so to see and measure the waveforms.

[iMo]

..and you'll be able to pick out the frequency from that.

.. provided you have got an LC tank or a resonant antenna wired parallel to the spark gap..

From the wiki above:

In the 1950s a Japanese toy company, Matsudaya, produced a line of cheap remote control toy trucks, boats and robots called Radicon, which used a low-power spark transmitter in the controller as an inexpensive way to produce the radio control signals.[133][134] The signals were received in the toy by a coherer receiver.

[T3sl4co1l]

.. provided you have got an LC tank or a resonant antenna wired parallel to the spark gap..

The coil itself is an LC tank, no worry there.

That is another reason why I suggest not probing the coil directly -- the probe has a circuit equivalent to a lossy capacitor, so it will disturb both the frequency and the damping of the waveform.  For some coil values, it may even disturb the measurement enough that it no longer appears as an LC tank (it goes "thud" instead of "boing").

The coil has self-capacitance and loss, but we don't know how much until it's measured.  So the only thing we can do is observe it from a modest distance, and see how it behaves.

Tim

[den]

A common method used in cars is measuring the primary. The spike there (ignition coils have relatively high transformation ratio) reach 400 V, which is measurable. You can see a voltage spike which is forming a spark (breakdown voltage) and then (when the arc has formed) an arc holding voltage. You can recalculate to the secondary voltage knowing transformation ratio (probably not too precise, but a reasonable number).

Then you can measure what happens, when you connect the primary to the source and disconnect. You will be able to tell that those are very different processes and spark energy will differ significantly.
In the first ignition systems there was a cap connected in parallell to the primary - to reduce primary contact sparking when they are disconnected. This then creates a significant oscillation during the arcing process.

[iMo]

.. provided you have got an LC tank or a resonant antenna wired parallel to the spark gap..

The coil itself is an LC tank, no worry there.

I doubt so..
The HV coil generating the sparks and the LC tank (or a resonant antenna) are two different things..
You need a resonating "something" wired/coupled to the spark gap (see above the wiki) in order to produce your "frequency of interest"..

[joeqsmith]

Hello,

I experimented with a piezo crystal from a lighter to generate a high voltage spark. A coherer was able to detect the generated EM wave (from the spark/spark gap) in close proximity.

Now I am trying to generate a spark with a coil that has the same effect. I charge a coil/inductor with a 4.5V battery and when I break the contact or make it, a spark is generated as expected.

How could I measure the spark voltage (and frequency) between the two leads? I am worried it might damage my scope.

Did you made your coherer and can you post a picture of it?   

When I was looking at my piezo crystal lighter,  I was looking at the current rather than the voltage.   To look at the voltage on my ignitions, I normally will use a capacitive coupled divider.   For the lower frequency current, I have a wide band transformer.    I kept one of my first DSOs just for this sort of potentially destructive experimenting.     

[T3sl4co1l]

I doubt so..
The HV coil generating the sparks and the LC tank (or a resonant antenna) are two different things..
You need a resonating "something" wired/coupled to the spark gap (see above the wiki) in order to produce your "frequency of interest"..

Measure some coils, or try it yourself.  Or write it out from E&M theory.  I think you will be pleasantly surprised.

Tim

[petert]

1000:1 probes for scopes are not that expensive.
Look up Joe Q Smith's DMM test videos where he uses grill ignitors and builds a 1000:1 probe so to see and measure the waveforms.

I suppose you mean this video?


Other possibly relevant videos:




Interesting but not directly related:

[tautech]

1000:1 probes for scopes are not that expensive.
Look up Joe Q Smith's DMM test videos where he uses grill ignitors and builds a 1000:1 probe so to see and measure the waveforms.

I suppose you mean this video?

Yep, all those !

We can see from Joe's investigations that grill ignitor voltages aren't that high and a 1000:1 probe capable of handling 15kV should be ample to view their waveforms safely.
See if you can find any Pintek HV probes sold locally:
http://www.pintek.com.tw/files/pintek/HVP-provided-Fig.pdf

[petert]

Did you made your coherer and can you post a picture of it?   

It's not the clearest, because I used it several times already and the flexible tube has become a bit tarnished.
I also adjusted it several times (moving the screws to adjust the distance) until I found a working "configuration", tapped it to decohere it, so that some of the filings are "spread out" on the screw threads.
The filings are of Cupronickel, a mix of copper and nickel (75 % Cu, 25 % Ni).

Full resolution pic (click to zoom):

[petert]

See if you can find any Pintek HV probes sold locally:
http://www.pintek.com.tw/files/pintek/HVP-provided-Fig.pdf

Thanks. They look interesting, however what I found is around 270€, and my scope is just a Rigol 1054Z, so it's pretty close in price haha.
(for future reference: https://www.reichelt.de/hochspannungs-tastkopf-40-mhz-passiv-bis-20-kv-testec-hvp-15-b-p267712.html)

[tautech]

See if you can find any Pintek HV probes sold locally:
http://www.pintek.com.tw/files/pintek/HVP-provided-Fig.pdf

Thanks. They look interesting, however what I found is around 300€, and my scope is just a Rigol 1054Z, so it's almost the same price haha.

Then keep an eye on the 2nd hand market. 

If you think HV probes are expensive then start looking at decent current probes !
I recently got a 100 MHz Tek P6022 for just $145 US to go with my 60 MHz P6021 and new they're $3k ! ! !
Want a good AC/DC current probe expect to pay even more, lots more.

The cost of a scope itself is just the start and to be well equipped to measure all manner of things you will always pay more for good accessories than the scope itself.

[joeqsmith]

1000:1 probes for scopes are not that expensive.
Look up Joe Q Smith's DMM test videos where he uses grill ignitors and builds a 1000:1 probe so to see and measure the waveforms.

I suppose you mean this video?

A few comments.    The homemade differential probes I made all have a fairly low bandwidth and voltage rating.    The box used to measure the AC power uses some odd ball transformer that again would have a fairly low bandwidth.   

If you really wanted to look at a spark with a scope you would need something fast.  The large HV probe I made is a 10,000:1, not 1000:1.  I built it to monitor the transients I apply when bench marking DMMs.  While these transients are far beyond what a typical cheap HV probe could measure, they are about 1000X slower than a spark.   While I did show this probe looking directly across a spark gap when I was looking for where the breakdown problems would be, its really not fast enough to show the waveform.   You can see that testing here (shortcuts are all close to the relative parts, you may need to skip around a bit).
https://youtu.be/OUfWYwPB9uI?t=305

Me reading a few paragraphs from an old RCA applications book on car ignitions.   Again, just a little background on what to expect.
https://youtu.be/bgz-pqg0rKo?t=1177

When I looked at the grill starter, I show the IEC standard and how I measured it.  You can find that in the UT181A PartB, here:
https://youtu.be/1kYcY2ogmqo?t=125

I ended up making a better spark gap and here you can see some of the data I collected for it. Again though, I am looking at current.  Note the rise and delay times. 
https://youtu.be/Qimtx8z6FUQ?t=132

Attached pictures looking at the output from one of my ignitions with a homemade capacitive probe and my current transformer.   This is my first real DSO which has a BW of 4GHz.  I am really only interested in making relative measurements and would need to improve the setup to see what the waveforms actually look like.   Nothing $$$ wouldn't solve.

[joeqsmith]

Did you made your coherer and can you post a picture of it?   

It's not the clearest, because I used it several times already and the flexible tube has become a bit tarnished.
I also adjusted it several times (moving the screws to adjust the distance) until I found a working "configuration", tapped it to decohere it, so that some of the filings are "spread out" on the screw threads.
The filings are of Cupronickel, a mix of copper and nickel (75 % Cu, 25 % Ni).
(Attachment Link)
Full resolution pic (click to zoom):
(Attachment Link)

  I would vote this post of the week if I could.   It's always good to see people experimenting.