Electronics > Beginners

Triggering scope upon removal of periodic AC signal

<< < (4/5) > >>

Circlotron:

--- Quote from: iroc86 on December 03, 2019, 02:35:39 am ---Just for context, I'm tuning an RC snubber on a small shaded-pole AC motor, so I'm using the scope to measure the back-EMF at shutoff. That's why I'm looking for a nice, defined event upon which to trigger.

--- End quote ---
Maybe measure the peak AC current. Then put this much DC (only) through the winding, switch off and trigger on the spike and see what your snubber does.

MarkL:
It's not a trigger per-se, but if you have the mask testing option you could create a tight mask on the AC waveform, set no persistence, and set the scope to stop if the mask test fails.  When the AC is removed, the mask will be violated and the scope will stop, but how much of the "off" event you will capture is still left to chance.  With a few tries you should be able to get one that starts near the left edge.

On some DSOs, mask test failures can also generate a trigger output.  This could be used to trigger a second scope to capture a longer record of the event.  I don't know if the 54645A has this.

You could also set infinite persistence.  All the traces will be layered, but you can probably pick out any interesting deviations from the normal sine wave.

If you wanted a nice, clean, repeatable trace, I would probably use a relay to cut AC power and use the relay control voltage as the trigger, as others have already pointed out.

macboy:

--- Quote from: vk6zgo on December 02, 2019, 03:34:43 am ---
--- Quote from: gcewing on December 02, 2019, 12:37:42 am ---Maybe use a small plugpack power supply if you have one lying around. Might need a lamp or something as a load so the output falls off quickly.

--- End quote ---

That's the hard bit!
The loss of AC has to happen before the trigger can be generated, so you are really battling.

Perhaps the OP could reverse things so that the 'scope triggers, then the AC disappears a fraction of a second later.
That is, of course, no help if they are looking for a way to see something which happens in an unplanned manner.

Switch off times with mechanical switches are surprisingly fast, & obviously the timing is quite random.

Another thought is, if you are only looking at one AC source & you have another one which doesn't break with it, would be to compare both sources, (first ensuring they are in phase) & if one disappears, generate a  trigger pulse.

If you can't have another source which doesn't break, perhaps sync an oscillator to the incoming AC frequency & compare that.
It will drift over time if it loses its reference, but hardly at all over the first cycle after that event.

--- End quote ---

This would be a big problem 30 years ago with an analog scope, but all digital scopes have the capability to capture the waveform leading up to the trigger event, rather than exclusively after the trigger, like with an analog scope. In fact, most digital scopes place the trigger event at the center (horizontally) of the screen by default, rather than at the left edge as was normal in analog days. You can further manipulate the controls to capture exclusively prior to the trigger, or after the trigger or anywhere in between. Of course you will need to set sample rate and sample depth appropriately in order to ensure that you can capture both enough time (pre-trigger) and enough HF detail.

The glitch trigger may be the best bet here. Usually this will halt capture when a periodic trigger comes early or late. Try to set up a one-shot glitch trigger so that the absence of a trigger transition after 1/50 or 1/60 second (+10% or so) will cause the capture to halt. You might be able to set up the scope to save the waveform and start again, or do something similar from a connected PC. Or you might need to just monitor the scope and deal with each captured waveform individually.

Some scopes have an envelope trigger (or something similar under another name) which lets you specify a min/max that the waveform must entirely fit within. If any sample point falls outside the envelope, capture is halted. It seems you probably don't have this, but the glitch should work for you just as well or better.

David Hess:

--- Quote from: iroc86 on December 03, 2019, 02:35:39 am ---Just for context, I'm tuning an RC snubber on a small shaded-pole AC motor, so I'm using the scope to measure the back-EMF at shutoff. That's why I'm looking for a nice, defined event upon which to trigger. When the motor is running, triggering off the AC source is just fine to see the whole 60 Hz waveform (10 ms/div), but I'm looking to capture the spike over a very short period of time (10 µs/div or faster).
--- End quote ---

Depending on how quickly shut-off occurs, setting the trigger to reject low frequencies will selectively trigger on the back EMF spike.  I do this all the time to trigger on just reverse recovery of a diode in a low frequency circuit.

iroc86:
Those are some cool ideas! There's always more than one way to skin a cat, and I appreciate learning some other techniques from you all.


--- Quote from: Circlotron on December 03, 2019, 04:38:58 am ---Maybe measure the peak AC current. Then put this much DC (only) through the winding, switch off and trigger on the spike and see what your snubber does.

--- End quote ---

That's an interesting alternative approach to measuring the back-EMF. Could you elaborate a bit more on this, and how the test setup might be configured in practice? The motor is drawing about 1 A peak-to-peak at 120 VAC according to my Fluke clamp-on current probe. Since the motor is designed to operate at mains voltage, wouldn't I need a fairly high-voltage DC source to generate that sort of current?


--- Quote from: MarkL on December 03, 2019, 02:54:09 pm ---It's not a trigger per-se, but if you have the mask testing option you could create a tight mask on the AC waveform, set no persistence, and set the scope to stop if the mask test fails.  When the AC is removed, the mask will be violated and the scope will stop, but how much of the "off" event you will capture is still left to chance.  With a few tries you should be able to get one that starts near the left edge.

--- End quote ---

I do have the HP measurement module with mask testing on this scope, so this is definitely worth trying!


--- Quote from: macboy on December 03, 2019, 04:01:12 pm ---The glitch trigger may be the best bet here. Usually this will halt capture when a periodic trigger comes early or late. Try to set up a one-shot glitch trigger so that the absence of a trigger transition after 1/50 or 1/60 second (+10% or so) will cause the capture to halt. You might be able to set up the scope to save the waveform and start again, or do something similar from a connected PC. Or you might need to just monitor the scope and deal with each captured waveform individually.

--- End quote ---

That's also a slick approach. I'll admit that I'm only familiar with basic triggering techniques, but the criteria for the glitch trigger as you described makes a lot of sense in this situation.


--- Quote from: David Hess on December 04, 2019, 06:09:53 pm ---Depending on how quickly shut-off occurs, setting the trigger to reject low frequencies will selectively trigger on the back EMF spike.  I do this all the time to trigger on just reverse recovery of a diode in a low frequency circuit.

--- End quote ---

Yeah, I think triggering on the event itself would probably best reflect what I'm actually looking to capture. I can see how rejecting LF would allow for triggering on the spike (relative to the 60 Hz signal) even if the overall magnitude of the spike is being limited by the snubber.

--

Okay, so now you've all got me thinking about a variety of ways to accomplish this task without resorting to external hardware. I should have some time over the next few days to investigate each of these ideas and report back. :)

Navigation

[0] Message Index

[#] Next page

[*] Previous page

There was an error while thanking
Thanking...
Go to full version
Powered by SMFPacks Advanced Attachments Uploader Mod