Winfield's power device selection criteria, and his Figure Of Merit for ranking, are laid out in one of the attachments. I suspect these are part of his Secret Sauce that enables higher speed.
I am a bit busy this week (new product release) but I will look at this over the weekend.
No doubt, as Mark has alluded, the devil is in the detail, but it would be a very interesting challenge.
I have AoE (latest book). Seriously good introduction to the art.
It's this simple: the example amp has the -3 dB point at 20 MHz - see pics.
Attachments
The reason I came here was to find an amp that I can use to excite a psu regulator with DC + AC at the input, as well as DC + AC load current at the output to measure input ripple rejection, output impedance, that sort of thing.
In principle, an audio power amp fits the bill but I want to sweep the frequency to at least a few 100kHz, better to at least a Meg.
Jan
The circuit in post-45 has no device types listed. Doesn't the compensation depend on the use of specific types?
Probably not, because I also designed an amp with THD (100 W 8 ohm 20 KHz) below 30 ppb (no mistake). IMD is lowered similarly, likely even audibly so and the only extra part is a dual opamp, LM49860 plus a few RC's.
Understood.
That said, for what you are doing, would a current load (adjustable) in parallel with a pulse generator (also adjustable) do the trick?
If you are driving the PSU UT with a fast rise time pulse train, you will be exciting it to HF because of the harmonic content while the DC load will be shifting the operating point.
Last edited:
"sx-Amp" is ..
Phrase "sx-Amp" will show no results on forum search, but now we know a new name for CF amplifier.
Phrase "sx-Amp" will show no results on forum search, but now we know a new name for CF amplifier.
We are all talking ppb. Does it matter ?
Or over 200v/uS ? Audio only requires a paltry <20v/uS.
The only reason I go with a 200V/uS CFA is the "fat" gain margin
at 20K , so I can get my (realistic) <20ppm THD 20K.
Not for the slewrate.
Edit - I have ported "greenamp CFA " to a BJT 40W class A. 1-5ppm all the way.
But not too "green". It gets too hot here.
OS
Or over 200v/uS ? Audio only requires a paltry <20v/uS.
The only reason I go with a 200V/uS CFA is the "fat" gain margin
at 20K , so I can get my (realistic) <20ppm THD 20K.
Not for the slewrate.
Edit - I have ported "greenamp CFA " to a BJT 40W class A. 1-5ppm all the way.
But not too "green". It gets too hot here.
OS
Last edited:
Yes, understood. But not as nice and easy to interprete as, say, a frequency sweep.
I was looking at this as well:
https://www.batronix.com/files/ITECH/IT8800.pdf
Specifically the 8812B.
Jan
Last edited:
Whew. Nice, but it looks very expensive - I can see why you rather prefer to go with a fast amp.