Oh, it does!
Gives more phase margin, or one can reduce compensation capacitance, so increase slew rate.
If you have no output coil, you need more phase margin.
Gives more phase margin, or one can reduce compensation capacitance, so increase slew rate.
If you have no output coil, you need more phase margin.
darkfenriz said:
Hi Adam,
Alright, although in an indirect way, it might improve SR. Now I see his point.
Cheers,
john curl said:
I got even more crazy and also stopped using the output RC to
ground around '91.
Wakawakawakawaka!
😎
You still use power supply though, don't you?
😀 😀
Neslon, what do you think of capacitive loading? Is it real threat with speaker cables or do speaker cables have enough serial inductance to still be not so hard load?
regards
Adam
😀 😀
Neslon, what do you think of capacitive loading? Is it real threat with speaker cables or do speaker cables have enough serial inductance to still be not so hard load?
regards
Adam
pooge said:
G'day.
I hope you don't mind, but I've posted a reply in the new "audible coil thread":
http://www.diyaudio.com/forums/showthread.php?postid=1203809#post1203809
darkfenriz said:
Other forms of compensation than Miller tend to break the restricting relationship between slew rate and gain crossover frequency. Miller compensation is pretty crude in this regard. Look at the compensation I used in my MOSFET amplifier, where I easily achieved 300 V/us in a 50 watt amplifier.
Bob
pooge said:
Thanks! I was unaware of this paper, and it looks like a good read. This is the sort of information I was hoping to see. I wonder if John does something like this.
Cheers,
Bob
Nelson Pass said:
Hi Nelson,
Cool! Did you just find out that the Zobel was unnecessary in the topolgy you were using, or did you make some kind of a circuit change to allow getting rid of it. It is certainly true that we sometimes do things almost by wrote, for historical reasons, because at sometime in the past it was really needed or solved a problem under circumstances that may no longer be relevant. I plead guilty to doing this sometimes, perhaps just out of conservatism. Any insight you can provide would be very helpful.
Also, what are the kinds of conditions under which you found the presence of the Zobel network to be audible in a negative way?
Is there any measurement that correlates the presence of the Zobel to the audible difference? Or must we just chalk it up to the X Factor?
Thanks,
Bob
darkfenriz said:
There was a time when cable capacitance/inductance was a
serious problem (1970's), but everyone seems to have adjusted.
The issue was not capacitance per se, but the resonance of the
unterminated cable.
I wrote an article on cables which addressed this:
www.passlabs.com
😎
john curl said:
Hi John, I completely understand and respect your desire to not spill all of the beans, and I wasn't expecting you to share a schematic. I was hoping that you would share in generalities, however.
For example, your conservatism in design is legendary in many areas, like ability to supply very high currents and deal with ridiculously difficult speaker loads, and this is laudable. So the question naturally arizes if you are applying the same conservatism to possibly de-stabilizing loads. Obviously, the coil was needed in the past. How do you get away with not using it now? For example, an answer might be that you chose to use low amounts of feedback. Or, that by using a huge number of fast output devices, you have shown it to be unnecessary. Or that you have used a proprietary form of feedback compensation. That's what I was getting at.
For example, have you tested your amplifier into a 2 uF ceramic capacitor load placed right at the output terminals, where the "capacitor" may consist of numerous 2 uF ceramic caps in parallel to achieve extremely low ESR and ESL? Did your amp pass this test?
Have you ever SPICE'd your amplifier into such a load?
We will naturally always try to pick your brain, but you always have the right to say no.
Thanks,
Bob
Well, if using only .5uH or less as output coil hopefully everyone is using mono amps and also placing the power supply in direct proximity to the output stage and is also stacking the NPN/PNP or NMOS/PMOS output trannies in alternating order, do not using any wiring to the output binding posts and is also using multilayer PCBs. Otherwise it's getting difficult to cut down the output parasitic impedance to .5uH or less.
But I also agree that it is a good idea to remove as much as possible components in the signal path as they are not absolutely necessary. Unfortunately this takes a lot of time when not degrading other parameters.
But I also agree that it is a good idea to remove as much as possible components in the signal path as they are not absolutely necessary. Unfortunately this takes a lot of time when not degrading other parameters.
Bob Cordell said:
Some of the beans then ?
(the thing that strikes me most of the JC-1 is the steep bias level of the Toshiba driver duo)
Bob Cordell said:
At some point I found myself with low enough feedback that it
didn't look necessary. So I deleted it.
The Zobel didn't create any problems, it was just extra parts.
😎
Bob, I would not deliberately subject my power amp to paralleled 2uf ceramic caps across the output terminals, but it would be a tough test. I hope that no audiophiles out there will do it either.
Bob Cordell said:
On my amps: yes, thought I did use a good non-inductive film cap, and yes, they did pass 🙂 . One of my usual tests on stability was to observe an overshoot on 1 kHz squarewave into 2.2uF load directly on the output terminals, at different signal levels including clipping.
Alex
john curl said:
Just implying that i steal your schematics, but no intention of starving you.
Stay healthy and continue the car talk, Mr Curl.
The 2uF test seems to have become a tradition with mags to tailor the fuses to size.