TDS744a (scope) technique: how to capture widely different yet related signals

[k1mgy]

The following measurement problem is associated with a high voltage (very low current) power supply.  HV is around 2.5KV.  The supply is charged through a 1:100 transformer by quickly completing a current path through the primary.  This is a very fast operation with a very slow repetition rate.

I'm attempting to measure the two signals for this circuit - the pulse which recharges the HV, and the ripple component of the HV itself.

The pulse is a negative-going signal of about 20uS that (allegedly) occurs at a 9Hz rate.  The HV ripple is a result of very slow bleed-off of the peak charge value.  We're talking nanoamps here.

I'd like to get at least two of the charge pulses in one sample so I may view the dynamics associated with the HV output.  I should see, for example, the HV output reaching a low-level threshold where it triggers a re-charge pulse, and then see the HV output rise until it dips once more and gets another boost.

Using the TDS744a, I'm struggling with how to set it up such that I may capture the needed data.

If I bring the sweep speed up sufficiently to see the pulse in a single X-axis division, the HV signal is all over the place - aliasing I guess.  If I trigger on the HV signal at around 9Hz, I see a stable semi-sine-wave but cannot see the pulse.

I suspect this problem has to do both with my setup and possibly the amount of memory and sample rate. 

Any suggestions out there?

[PA0PBZ]

One thing that comes to mind is loading the HV side so that the charge pulse will occur more often, or are you trying to capture the behavior with the load as it is now?
Is it necessary to see the charge pulse or would it be enough to trigger on it?

[k1mgy]

I really would like to capture present behaviour.  The pulse repetition rate changes very little under load.. there's simply no load.

[MarkL]

...
If I bring the sweep speed up sufficiently to see the pulse in a single X-axis division, the HV signal is all over the place - aliasing I guess.  If I trigger on the HV signal at around 9Hz, I see a stable semi-sine-wave but cannot see the pulse.

Are you doing this with a single shot acquisition?  Maybe the reason you're seeing the "HV signal all over the place" is because the scope is falling into auto-trigger mode due to the slow repetition rate.  You should be able to do a single shot to capture the full 9Hz cycle and then zoom in on the pulse (at least to an extent; 20us wouldn't be represented by many points if you only have the base 50kpts of memory).

Or set "Normal" trigger mode and it will at least not fall into auto-trigger mode and then you can work with the delay and sweep settings to move around in the waveform.

If that's not the issue, perhaps some screen shots and/or the scope settings would help us make some suggestions.

[k1mgy]

I recall faintly that the acquisition memory can be expanded in this scope?  Would this help?

Meanwhile I will get some screen shots and post settings to further any ideas.

[danadak]

As stated earlier your memory depth for a single shot event along with
sample rate will determine if you can capture the event with reasonable
fidelity. Lower sample rate to 3X highest freq of interest to see if you
can do the capture effectively. Newer scopes have a lot more triggering
capability and memory.


Regards, Dana.

[MarkL]

I recall faintly that the acquisition memory can be expanded in this scope?  Would this help?

Meanwhile I will get some screen shots and post settings to further any ideas.

More memory is (almost) always better.  There is an option 1M for this scope which increases the memory to 250kpts for 2 channels on or 500kpts for 1 channel.

There is a hack for it if you're interested:

  https://www.eevblog.com/forum/testgear/enabling-option-1m-extended-acquisition-memory-tektronix-tds754a/msg565086/#msg565086

[nctnico]

AFAIK there are two ways: use segmented capturing or use persistence. Using long memory for capturing very slow events is not going to work for making comparisons.