Reactive power - Reactive Generator research from GotoLuc - discussion thread

[TinselKoala]

Mine does, and I think most variacs do too. Just saying that they can, but yours apparently doesn't.

I suggest checking with an ohmmeter just to be sure that the two sides are actually completely isolated and that the neutral isn't common to both sides of the isolation transformer. I've seen ones (even Solas) that had the neutral common, so they didn't achieve full isolation after all.

Every Variac I've seen has one terminal in common to both input and output sides, so they don't really isolate.

[TinselKoala]

Here is a zip of the original file.

BTW, the load is 15 Ohm 1% 50w resistors connected to 3 Series 12v batteries at 37 volts DC, so no error on power to the load.

It's 1:25am and been a long day, so I'm closing up for the night. I'll pickup more of my 12v batteries from storage tomorrow so we can see how far we can take this.

Stay tuned and please try to find the error as this is too simple!

Luc

OK, the negative math mean is only about one-twentieth of one vertical division (math trace is 200 V-V per div). This could easily be a sampling or calibration error. The "mean" of an approximately sinusoidal waveform like the math result should be approximately zero. In order for human eyeballs to check the scope's math it would be nice to have only three or four full cycles displayed, so that areas above and below the zero reference could be compared by external means.
Since the negative value is so small compared to the overall signal (nearly 800 V-V p-p) I am skeptical that it is real.

[d3x0r]

I don't know where the error is, but I see now why the scopeshots are so pixellated, the scope is set to save them at 320x240 resolution. And from looking at manuals it seems that might be as good as you can get. Maybe Poynt99 can give some advice as to how to get better resolution.

Sorry for the distraction, it's not really important, just hard for my old eyes to read your images especially with a lot of cycles and traces displayed.

And personally; my scope (rigol ds10something), the math is done with the value shown on the screen and not what is in the sample buffer... so it's pathetic low resolution like that even though it's 2Msa stored....like the values change if I change the time...

[TinselKoala]

And personally; my scope (rigol ds10something), the math is done with the value shown on the screen and not what is in the sample buffer... so it's pathetic low resolution like that even though it's 2Msa stored....like the values change if I change the time...

I think that's why .99 asked for many cycles on-screen. Maybe Luc's Tek also only does math on the screen display too. It's always a tradeoff, I guess. The old Link pc-DSO that was donated to me can do its math and measurements on the entire sample buffer, or just what's displayed on the screen, as the user chooses, but its math is limited and it can't do trace multiplication. Some other scopes I've used, like the early LeCroy WaveRunner 104x,  seem to incorporate the trace offset into the math, so one needs to be very careful about where the baselines are set. Luc is setting them all at the screen center, I think, so that's not likely to be a problem here.

[poynt99]

Probably because I open them in windows image viewer and use screenshot pilot and make a png file so I can upload then. This site is not accepting bitmap (Tek regular file format) for some reason.

Luc

According to the manual, you have the following image file format options to save to:

bmp, pcx, tiff, rle, epsimage, jpeg

I would suggest trying tiff and jpeg. I don't see where one can change the resolution though.

btw, nice scope Luc. Where did you buy it?