For others looking at the DSO screen shots. The jagged lines on this particular DSO is due to the limited resolution of the internal DAC. It will be more smooth on higher resolution DAC's like I have with my RIGOL DS2xxx series. The CT is fully analog so is not the reason.
It's NOT the triangle waveform, as I explained above, this is a well know effect of the DUT heating/cooling called blooming.
Sure. I am not implying anything else.
I was simply wondering if I understand correctly how the system works. As far as I can tell, even if the triangle wave was not perfectly triangular, the scope would measure the true voltage of the signal and put the measurement value at the correct/measured x-axis position. Correct?
Jeez. I apologize for asking. I was aiming to further my understanding. I didn't think self-heating would cause that much variation. I have never designed a curve tracer so what's well-known and obvious to you may not be intuitively obvious to me even though I'm well-versed in semiconductor device physics. I also have the occasional brain fart. I'm human. Sorry...
Tom
Hi Tom, I was not trying to lecture you. Sometimes people don't read earlier posts.Oh, DUH! Yeah. See my comment about being human and having brain farts above.
Tom
If you look at the display, you can see that with a small Base current (the lower steps), there is no blooming. That starts to happen with higher Base currents, as you can see in the top traces. A higher Base current results in a higher Collector current, and that heats up the device and changes the gain a little. The heating on one way of the curve, and the cooling on the traverse is the cause for the blooming effect. A non-perfect triangle would show up on all I/V lines, not just the top ones. On my blog: look here, search for "dealing with thermal issues". I show a picture that was made with a Tektronix CT, and it shows the exact same effects.
https://www.paulvdiyblogs.net/2022/06/the-vba-curve-tracer.html
The building is somewhat challenging but manageable for someone experienced in SMD soldering. It works great and with a decent oscilloscope provides decent results. A digital oscilloscope with a analogue input is preferred.
All kinds of transistors supported incl. IGBTs within the available power ranges.
There are three main ranges that can be selected.
The building is somewhat challenging but manageable for someone experienced in SMD soldering. It works great and with a decent oscilloscope provides decent results. A digital oscilloscope with a analogue input is preferred.
All kinds of transistors supported incl. IGBTs within the available power ranges.
There are three main ranges that can be selected.
- 0-35V @ 0-2A
- 0-70V @ 0-1A
- 0-200V @ 0-100mA
Yes, D.S. oscilloscope's are usually 8 bit, with some input stage non-linearity as well (you can't use heavy feedback on a 200MHz amplifier, that would become unstable). The human eye looking at a triangle curve might spot 1 or 2% deviation, but the underlying data points could be 0.5% off easily anyway and you probably won't see that.
My vote for the discrepancy is thermal cycling though, the way to test this is increase the sweep rate - thermal distortion decreases with frequency.
My vote for the discrepancy is thermal cycling though, the way to test this is increase the sweep rate - thermal distortion decreases with frequency.