Vaccy,
I will presume you still have a working right hand! With a zig-zag layout the current of the resistors next to each other is by convention flowing in opposite directions so the magnetic fields are opposed to each other and mostly cancel. I will also presume you learned the right hand rule relating current flow to magnetic field so you can test the hypotheses with the assistance of a friend. 😉
I will presume you still have a working right hand! With a zig-zag layout the current of the resistors next to each other is by convention flowing in opposite directions so the magnetic fields are opposed to each other and mostly cancel. I will also presume you learned the right hand rule relating current flow to magnetic field so you can test the hypotheses with the assistance of a friend. 😉
If you search for the Putzeys designed 30W (balanced) class A amp, (Extreme A from memory?) , it had a pretty cool OP stage with local FB
loop current driving the emitters of a pair of 'backward' bjts.
I've used this arrangement with complimentary k170 / J47 jfets driven
(backwards) from the sources, (you can work the rest out) and it worked pretty well.
Have since given up all that guff, now playing with fully bridged CFB,
two pole compensation.
Killer!
T
Run some numbers. Pick a frequency omega ("w" on my keyboard) that you think is important, pick an inductance per resistor number that you think is reasonable, and then calculate 20*|j*w*L| . Compare that to the total resistance of the 20 resistors in series. Do the same thing for phase shift at omega. How significant is it?
I have used the suggested wirewound resistors. They are well made and designed. The thermo electric design is most impressive they are only 30 micro volts per degree C for the temperature differential across the body. (My measurement) They are reasonably low inductance but a bit higher capacitance than other types.
The issue is the welds between the end caps and resistance wire seems to be what makes some of them have unusual behavior on low duty cycle high power spikes.
Although I haven't used them in years as I have found better techniques using metal film resistors to get better results. Important note metal oxide film resistors are not the same as metal film!
The issue is the welds between the end caps and resistance wire seems to be what makes some of them have unusual behavior on low duty cycle high power spikes.
Although I haven't used them in years as I have found better techniques using metal film resistors to get better results. Important note metal oxide film resistors are not the same as metal film!
I recall it Halcro did enclose the OPS with fast (video?) opamps, many ways to linearize the OPS taking some of the burden off from GFB.
Mark,
If you look at the idea behind the inductance cancellation winding technique for wirewound resistors and the zig-zag of flat mounted resistors, I suspect you will have your early morning bit of humor.
ES
If you look at the idea behind the inductance cancellation winding technique for wirewound resistors and the zig-zag of flat mounted resistors, I suspect you will have your early morning bit of humor.
ES
That's why I've split my amp in two separate parts, a voltage amp and an OPS.
The OPS is error corrected by two nested error correction circuits, with very fast video opamps.
The filter between the two makes a load independent BW reduction.
Hans
Funny that you would post this, because just the other week I was doing some experiments with rather inductive 110 Ohm power resistors before soldering them into a ballast to test amps.
I thought along the same lines as you that theoretically it might happen possibly, but there is no effect. The magfields are just too far removed for the resistors to even see each other.
The inductance of one line is more or less canceled by the proximity of another where the current flow in the opposite direction. I have no idea if the result will be better than putting them in parallel.
The zigzag configuration was not suggested with the Syn08 suggestion. But, you are right, I made a mistake. Using .3nH, it is OK. So, may-be a good idea with thos in // at the end ?
I asked myself the same question. Without answer, of course.
A joke, of course.
[edit] Oh, its author already answered (and confirmed), sorry.
Last edited:
For up to 100W I parallel 5 0.5W resistors. For 240W I use a 1k 10W Mills (MRA IIRC) - very low inductance and good voltage co-eff. The lower leg of the feedback network is a Xircon MF.
For VFA, you can get away with much higher values (2k up to 10 or 15k) so 2 or 3 good quality 0.5W MF's in parallel will do the trick.
If you are designing a big amp that follows the 1~2V/us per peak output volt, I think the main feedback resistor inductance needs to be low otherwise you will have stability issues.
My 2 cents.
Peace 🙂
I did this with a video opamp as a LTspice sim a few years back - can't recall if I posted it up.
But, I do know Jan Didden came up with a similar approach and published it - if he's lurking maybe he could point us to it.
If you properly bias up a conventional class AB amp using decent OPS transistors, the OL OPS distortion is remarkably low. In sims, you can easily get 0.03% at 20 kHz with say 6 pairs at 200 Watts.
If the rest of the amp OL distortion is a 100 or so ppm worst case, 50 dB of feedback at 20 kHz theoretically gets you to single digit ppm.
The OPS bias and cross over distortion mechanisms are well understood. If I am perfectly honest, taking overall worst case distortion below say 50ppm will not deliver perceptible sonic benefits - although doing so is relatively easy and I've done that. Measurements showing a c. 5 fold improvement (30-35 ppm down to 7ppm which is the limit of my test gear) do not show a perceptible difference in sound quality using good speakers or a good set of headphones.
In general, more focus is needed on getting noise and hum lower where my perception is there is a big difference between say 80 or 90 dB vs say 110 dB. I don't know why, but it does sound cleaner.
A few of the "Oooh, I know electromagnetics!!" contributors here, also chimed in when I built an 8 ohm load resistor out of 28 pieces (56R @ 10W) wirewound resistors.
Building my own noninductive 8R 150W load using wire wound resistors
Building my own noninductive 8R 150W load using wire wound resistors
Waiting to see if that guy will troll you about level matched double blind testing...
- Status
- Not open for further replies.