Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

O-scope to measure resonance of automotive coil

Status
Not open for further replies.
I have a chance to borrow a Tek TDS1000B o-scope and it comes with two 10x probes. I have a microprocessor that is driving a mosfet which is running an automotive coil. I found this online, but wasn't sure if I could hook up the o-scope with a 10x probe to the high voltage side of the coil? If not, how do I measure the resonance of the coil? I want to maximize the voltage output with the least amount of energy.

This is the tutorial I found online (looking at page 6 in particular):

https://www.electro-tech-online.com/custompdfs/2010/01/E1b_3b.pdf

I know the tutorial says you can connect a 10x probe...but will I be able to with my o-scope and with the coil running off 12v and putting out thousands of volts on the 2ndary side? I just don't want to do any damage to the o-scope or the probes. Any input you can give would be greatly appreciated!
 
Whomever loaned the scope will never again loan you anything if you connect it to the coil.
 
Do NOT connect the scope probe to the high voltage terminal on the coil, or you will be paying for an expensive repair to the borrowed scope...

Here is how I have done it: Connect the HV terminal to a spark plug. Ground the threaded portion of the spark plug to the coil case (common connection between the primary and secondary windings) through a 1/2W 100 Ohm resistor. Connect the scope probe across the 100 Ohm resistor, with the ground clip to the coil case. The scope reading will show the current through the 100 Ohm resistor as the plug fires. I = E/R...
 
In Figure 4 on page 6, the primary is being driven by a Function Generator which puts out a low voltage (perhaps a volt), thus the coil secondary voltage (typically 100 times higher than the primary) is still low and you can measure it with the oscilloscope. Just because it says you can use an oscilloscope on the coil secondary in that particular configuration, doesn't mean you can in a different setup.

Your concern is valid if the circuit is operating with 12V on the primary in the normal Kettering circuit, such as Figure 1, where the output is typically 20,000 to 30,000 volts. That will almost certainly zap the oscilloscope and/or probes if you hook it to the secondary. The oscilloscope has a maximum stated input voltage it can handle and you must never exceed that.
 
I've often though about trying an experiment to get a relative measure of voltage and current in an ignition wire. I would wind a few truns of wire on a torroid, and slip it over an ignition wire. A differential measurement of the wire should output a current, and single-ended should output a voltage. Of course, the instrument would need to be protected from over voltage. Maybe something like that would work too.
 
Do NOT connect the scope probe to the high voltage terminal on the coil, or you will be paying for an expensive repair to the borrowed scope...

Here is how I have done it: Connect the HV terminal to a spark plug. Ground the threaded portion of the spark plug to the coil case (common connection between the primary and secondary windings) through a 1/2W 100 Ohm resistor. Connect the scope probe across the 100 Ohm resistor, with the ground clip to the coil case. The scope reading will show the current through the 100 Ohm resistor as the plug fires. I = E/R...

Gotcha. So you just adjust the freqency going into the coil and watch the scope till you get the maximum current from the output of the coil. That frequency is the resonance frequency of the coil. Correct?

Yes I figured connecting the scope to the 2ndary to be a bit nuts...I did recall seeing something about 300v max on the 10x probe, so I wondered how they could do it in that example. Makes sense that the function generator would only put out a small voltage and the output will be under 300v which would allow for the use of an o-scope to connect directly to the output. Thank you for the heads up!
 
The set-up depends on what you are trying to measure. Are you using the coil to fire a spark plug, or are you using it for something else? If you are tying to maximize the "spark", then monitoring the current as I outlined worked for me. If you are using the coil for some other purpose, then introducing the load represented by the ionized air in the spark gap will modify what the coil puts outl...
 
The set-up depends on what you are trying to measure. Are you using the coil to fire a spark plug, or are you using it for something else? If you are tying to maximize the "spark", then monitoring the current as I outlined worked for me. If you are using the coil for some other purpose, then introducing the load represented by the ionized air in the spark gap will modify what the coil puts outl...

I'm building an electric fence module so a spark plug wouldn't be the exact same representation... Maybe I can have the Mosfet just run the coil primary at a 1-2 volts and then measure off the 2ndary with the scope like the example? Does the resonance of a coil change with voltage input? I would assume so?
 
How will knowing the resonance frequency help you? The usual way to make a fence charger is to apply rated current to the primary of the coil for a few milliseconds, once a second. The fence chargers I've seen in the past used a mechanical gizmo to pulse the spark coil once a second. You don't need to know the resonance frequency; it all takes care of itself.

But, just for grins, I measured the resonance frequency of 3 automotive spark coils I have lying around. I connected a 200 megohm resistor to the secondary, in series with a 2k resistor, then to the minus terminal of the coil. I passed about 1 amp through the coil primary (with a .22 uF cap across the primary) and then interrupted the primary current with a mechanical switch. I monitored the voltage across the 2k resistor (which was about 1/100000th of the secondary voltage) with a scope.

When the primary current was interrupted, there was a voltage with a decaying sine wave characteristic at the secondary. The peak voltage was about 10,000 volts, and the frequency of the ring (the coil resonance) was about 2.85 kHz for one of the coils. The other two coils also had a resonance frequency of around 3 kHz.

I measured the resonance frequency without the .22uF cap across the primary. The resonance frequency increased to about 3.1 kHz. The primary cap has very little effect on the coil resonance; it's almost completely dominated by the self resonance of the secondary. But, the .22uF primary cap is needed for proper operation. It absorbs most of the spike that would otherwise appear at the primary due to the leakage inductance.

You should also know that automotive spark coils aren't designed to withstand the voltage that appears at the secondary when they are operated with rated current, and with the secondary open-circuited for extended periods of time. When they are connected to a spark plug, the breakdown of the plug gap limits the secondary voltage. You'll say, everybody who has ever worked on a car has disconnected a plug to see if there's a spark, so you might think that the coil won't be hurt by doing that. But running open-circuited for extended periods of time can be harmful to the coil.

If you're going to use an automotive spark coil to charge a fence, there will not be a spark most of the time; you'll be running open-circuited. Maybe if a cow or some other conductive object is touching the fence, the voltage would be limited; but most of the time it will be open-circuited. So, you should run the coil at a reduced primary current, or I suppose you could connect a spark plug across the secondary to act as a voltage limiter. With a spark plug in place, you will still get a fairly high voltage out, and you'll protect the coil from internal arc over.
 
How will knowing the resonance frequency help you? The usual way to make a fence charger is to apply rated current to the primary of the coil for a few milliseconds, once a second. The fence chargers I've seen in the past used a mechanical gizmo to pulse the spark coil once a second. You don't need to know the resonance frequency; it all takes care of itself.

But, just for grins, I measured the resonance frequency of 3 automotive spark coils I have lying around. I connected a 200 megohm resistor to the secondary, in series with a 2k resistor, then to the minus terminal of the coil. I passed about 1 amp through the coil primary (with a .22 uF cap across the primary) and then interrupted the primary current with a mechanical switch. I monitored the voltage across the 2k resistor (which was about 1/100000th of the secondary voltage) with a scope.

When the primary current was interrupted, there was a voltage with a decaying sine wave characteristic at the secondary. The peak voltage was about 10,000 volts, and the frequency of the ring (the coil resonance) was about 2.85 kHz for one of the coils. The other two coils also had a resonance frequency of around 3 kHz.

I measured the resonance frequency without the .22uF cap across the primary. The resonance frequency increased to about 3.1 kHz. The primary cap has very little effect on the coil resonance; it's almost completely dominated by the self resonance of the secondary. But, the .22uF primary cap is needed for proper operation. It absorbs most of the spike that would otherwise appear at the primary due to the leakage inductance.

You should also know that automotive spark coils aren't designed to withstand the voltage that appears at the secondary when they are operated with rated current, and with the secondary open-circuited for extended periods of time. When they are connected to a spark plug, the breakdown of the plug gap limits the secondary voltage. You'll say, everybody who has ever worked on a car has disconnected a plug to see if there's a spark, so you might think that the coil won't be hurt by doing that. But running open-circuited for extended periods of time can be harmful to the coil.

If you're going to use an automotive spark coil to charge a fence, there will not be a spark most of the time; you'll be running open-circuited. Maybe if a cow or some other conductive object is touching the fence, the voltage would be limited; but most of the time it will be open-circuited. So, you should run the coil at a reduced primary current, or I suppose you could connect a spark plug across the secondary to act as a voltage limiter. With a spark plug in place, you will still get a fairly high voltage out, and you'll protect the coil from internal arc over.

Wow thank you for the awesome response. The setup I'm using is not like a typical electric fence, so I don't have much to worry about with open-circuited issues. Thank you for the heads up on it, as I wasn't aware of that. Just curious what coils don't mind having an open-circuit that would output similar voltage as an automotive coil?

As for why I want to know the resonance frequency, I want to set up the microcontroller to operate the coil at resonance frequency for a fraction of a second so the coil is operating as efficient as possible. I want the highest voltage output for the least current input, which I'm under the assumption only occurs at the resonance frequency of the coil. I may take this into a solar/small battery solution, so the most efficient way to electrify a fence is the desired outcome.
 
Last edited:
Wow thank you for the awesome response. The setup I'm using is not like a typical electric fence, so I don't have much to worry about with open-circuited issues. Thank you for the heads up on it, as I wasn't aware of that. Just curious what coils don't mind having an open-circuit that would output similar voltage as an automotive coil?

I don't know specifically what kind of coil fence chargers use, but my guess would be that they just use a standard coil, but with reduced drive. An automotive coil is capable of 30,000 to 40,000 volts output, and you don't need that much for an electric fence.

As for why I want to know the resonance frequency, I want to set up the microcontroller to operate the coil at resonance frequency for a fraction of a second so the coil is operating as efficient as possible. I want the highest voltage output for the least current input, which I'm under the assumption only occurs at the resonance frequency of the coil. I may take this into a solar/small battery solution, so the most efficient way to electrify a fence is the desired outcome.

When you say that you want to operate the coil at its resonance frequency, that sounds like you intend to drive its primary with a sine wave. The standard way to do it would be to apply a square pulse of voltage for a few milliseconds to the primary (probably through a current limiting resistor is your drive voltage is 12 volts as in a car). After that time, the flux in the core will have built up and you can terminate the current.

At this point, the primary voltage will rise, and your circuitry must allow it to rise to a few hundred volts, which it will do on its own if there isn't some aspect to your drive circuitry that would limit it (such as breakdown of the device which is switching the current, perhaps a FET or IGBT). In other words, the switch device which turns the primary current on and off must be able to withstand several hundred volts when the current is switched off.

If you were to try to drive the primary with a sine wave, you would have to provide several hundred volts of sine wave drive if you expect to get secondary voltages of 10,000 or 20,000. This would not be a good way to go.

The way you get low energy consumption in an electric fence charger is to energize the fence with a high voltage pulse from the coil only once a second or so. You don't need to provide a high voltage on the fence continuously. At least, not if it's farm animals (or something similar) you're trying to discourage.
 
I finally got around to reading the .pdf file you linked to in your first post, and I now see why you're focusing on resonance.

It mentions choosing a capacitor to resonate the inductance of the primary with the resonance frequency of the secondary. This is very non-critical, and the value of capacitance typically used in car ignition systems is .22uF. In fact, I just measured a distributor capacitor I have, and, sure enough, it's .22uF. If you're using a mechanical switch, just use a .22uF cap, and don't worry about whether it's exactly tuning the primary resonance to match the secondary resonance.

All you need to do is connect a capacitor as shown in Fig. 1 of the .pdf file (the capacitor needs to be rated for 400 volts working voltage). Then the whole primary/secondary system will be resonant.

When you use a FET to switch the primary current, you don't need a capacitor across the primary; the secondary self-resonance will do the job.

Instead of using a mechanical switch (such as a relay), you could use the Fig. 7 circuit, driven by your microprocessor, turning the FET on for a few milliseconds, once a second, and you will get the kind of high voltage pulses a standard fence charger provides. Notice in Fig. 7, he mentions that the IRF640 FET is rated at 200 volts. Limiting the voltage across the coil's primary to 200 volts, rather than allowing it to rise to, say, 400 volts would be a good way to make sure that the secondary voltage isn't too high (thereby avoiding internal arc-over).

You could change the on time, making it shorter and shorter while monitoring your output voltage (see how long the spark is), and make the on time just long enough to give the desired high voltage out. The shorter the on time, the less overall energy the system takes.
 
You can also just use a timer and a big old power transistor. Just be careful to keep HV spikes out of the power rails. They tend to destroy everything connected to the power, including any components and power suppplies being used. I learned that the hard way.
 
The Electrician, thank you once again for your input!

The microcontroller will be putting out a square wave pulse, running through a mosfet to power the coil. I've had some 3/4" arcs, so running off a benchtop power supply without a concern for efficiency is fine...but with running the coil at say 3kHz (like you mentioned with your experiment) with the coil in resonance would be more efficient. Then run the coil at 3kHz for 20ms every second to electrify the fence (I will determine the desired on/off time cycle, no worries there).

The microcontroller and primary side of the coil will be running off a 12v source. The HV 2ndary side will have an ignition wire lead to clip onto the wire fence. To decrease the voltage I can either use a non-resonant frequency or just use a shorter on time?

So with an electric fence setup using an auto coil, what would be the best way to measure the resonance of the coil? The method mentioned by MikeMl?

Here is how I have done it: Connect the HV terminal to a spark plug. Ground the threaded portion of the spark plug to the coil case (common connection between the primary and secondary windings) through a 1/2W 100 Ohm resistor. Connect the scope probe across the 100 Ohm resistor, with the ground clip to the coil case. The scope reading will show the current through the 100 Ohm resistor as the plug fires. I = E/R...
 
Go back and re-read post #14.

You really don't need to worry about resonance. The coil will automatically generate a damped sine wave at the self resonance of the secondary if you apply a few millisecond pulse (narrow square wave) to the gate of the FET and then turn it off.

You can safely look at the voltage on the primary of the coil with a 10x probe and scope, and you shouldn't need to look at the secondary voltage because it's pretty much a replica of the primary voltage if the secondary is open circuited.

You can calculate the resonance frequency from the time between peaks of the damped sine wave; this time is the period (1/frequency) of the sine wave.
 
Status
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips

Back
Top