Workhorse said:
Peak output current measurements are virtually meaningless unless they're accompanied by the associated measurements for linearity.
For instance, 126A peak is fairly useless if it's delivered at 99% THD+N.
Mike,
you are winding people up again, naughty
The very fact Kanwar's amp can do this is worthy of praise, not condemnation....... this headroom has to be good, not bad.....
Hugh
you are winding people up again, naughty
The very fact Kanwar's amp can do this is worthy of praise, not condemnation....... this headroom has to be good, not bad.....
Hugh
john curl said:
6uH...edge wound large diameter 2.25 inches...You know some things are needed...😉
mikeks said:
So you are going to Define SHORT CIRCUIT CURRENT LINEARITY in this forum....[Your Dself Bible must be teaching you this also..]
I think then we must define a new term for measurement of Short-Circuit Linearity. Iscpk+thd...hhhmmmm sounds good
Should we call it....126A Peak with 0.0001%THD.....
Kanwar, are you making sports cars or trucks? 6uH is REAL BIG. 2uH is normal 30 years ago. Today we don't use output coils.
john curl said:
We are making Powerfull Trucks out of N-Channel Vertical Mosfets, which can take any known type of speaker load upto 2 ohms, no matter how reactive its going to be when driven in to harsh steep hill climbing conditions....even in summers...😉
BTW, I have seen upto 15uH in regular pro-amps, meanwhile mine is very low as compared to them
Considering that you resolutely chose not to use SOA protection, I doubt very much that you actually measured this short-circuit without your amp. becoming an expensive fire-cracker.🙄
Well I said that I wasn’t going to post here anymore, but I’m idly waiting for a big application note to download via dial-up at the moment, so what the hell.
Along with a 12W class A amp discussed here previously, I’m currently building up this medium power amp:
http://users.picknowl.com.au/~glenk/160W.HTM
Its rated at 160W rms into 4 ohms with an output stage designed to give a lot of short duration output current, with eight 200W, 250V, 16A BJT’s in the output stage. At these rail voltages secondary breakdown isn’t an issue with these parts, so they’re essentially only thermally limiting.
Short circuit protection and peak current duration limiting for the amp will be implemented with a current sensing protection circuit that latches on and cuts the output stage supply rails (low RdsON MOSFET switched).
I’ll be putting up on my website some pretty sinewave-burst oscillograms into a low resistance load when it’s ready.
I might even do some destructive testing too, to see just how much oomph can be extracted from the BJT’s reliably before they blow. Would make an interesting comparison to Bob’s V-MOSFET results often cited here as a reference.
We being whom? Those who like building amplifiers prone to oscillation? 🙄
http://www.stereophile.com/solidpoweramps/774/index6.html
Cheers,
Glen
Along with a 12W class A amp discussed here previously, I’m currently building up this medium power amp:
http://users.picknowl.com.au/~glenk/160W.HTM
Its rated at 160W rms into 4 ohms with an output stage designed to give a lot of short duration output current, with eight 200W, 250V, 16A BJT’s in the output stage. At these rail voltages secondary breakdown isn’t an issue with these parts, so they’re essentially only thermally limiting.
Short circuit protection and peak current duration limiting for the amp will be implemented with a current sensing protection circuit that latches on and cuts the output stage supply rails (low RdsON MOSFET switched).
I’ll be putting up on my website some pretty sinewave-burst oscillograms into a low resistance load when it’s ready.
I might even do some destructive testing too, to see just how much oomph can be extracted from the BJT’s reliably before they blow. Would make an interesting comparison to Bob’s V-MOSFET results often cited here as a reference.
john curl said:
We being whom? Those who like building amplifiers prone to oscillation? 🙄
http://www.stereophile.com/solidpoweramps/774/index6.html
Cheers,
Glen
The discussion on output coils can be found in this new thread:
http://www.diyaudio.com/forums/showthread.php?threadid=101468
/Hugo
http://www.diyaudio.com/forums/showthread.php?threadid=101468
/Hugo
john curl said:
Hi John,
I agree that 6 uH is very big. However, I do still believe in using a small-value R-L combination to isolate the outside world. Typically I like 1 uH and 1 ohm. This causes very little ringing and virtually no frequency response issue, and pretty much guarantees VHF stability, even when using extremely fast output devices like Vertical MOSFETs. I guess I'm just being conservative here - better safe than sorry. In some cases, when possible, I will use a torroidal air core output inductor, but it is difficult to get these down to 1 uH with decent-sized wire.
What specific measures do you and others ("we") take in modern power amplifiers that make it possible to dispense with the output inductor?
Thanks,
Bob
G.Kleinschmidt said:
Hi Glen,
Good to see you back. The destructive BJT testing you mention would be very valuable. I would recommend you use the ring emitter devices like those that I think John Curl uses (don't remember the numbers offhand). That would give us a good comparison. My measurements showed that the HEXFETs could handle 2-3 times the published SOA-rated pulse power when tested at 100V and 10 ms, if I recall correctly. An interesting question is whether the BJTs have as much margin against published SOA.
Cheers,
Bob
Hi Bob C,Bob Cordell said:
are you saying 2 to 3 times the 10mS SOA figures?
eg, irf240 is a 150W device that can take about 3Apk @ 100Vds for a 10mS pulse. Multiply that up by a further 2 to 3times takes us to 6Apk to 9Apk @ 100Vds for the same 10mS pulse.
Those failure currents are of the same order as published for a 1mS pulse.
Hi Bob, been there, done that. Better to use no coil if practical. But then, I make amps for the commercial marketplace. ;-)
1uH-1ohm? Did it decades ago.
Of course, I thought that Nelson Pass was being a little crazy in 1980 or so when he mentioned not using an output coil and the potential tradeoff with slew rate. Remember that conversation, Nelson? I was a BIG slew rate fan, myself, in 1980 and was pushing over 500V/us. Why slow down, when just a simple output coil could make stability easier and what harm could it possibly do?
1uH-1ohm? Did it decades ago.
Of course, I thought that Nelson Pass was being a little crazy in 1980 or so when he mentioned not using an output coil and the potential tradeoff with slew rate. Remember that conversation, Nelson? I was a BIG slew rate fan, myself, in 1980 and was pushing over 500V/us. Why slow down, when just a simple output coil could make stability easier and what harm could it possibly do?
john curl said:
Hi John,
I don't think you answered my question. If an output coil was needed in the past for a given degree of stability into a given assumed bad load, what change in circuit design did you make that permitted you to dispense with the coil and still have comparable stability margin?
Or, did you just conclude that you would throw out the coil and live with the resulting reduced stability margin into bad loads? (This could be a reasonable approach - no criticism is meant here).
This is a good discussion, and answers to questions like these are what really help educate those reading the thread.
In fairness to you, I do not take the categoric position that output coils cannot be heard. The way that I look at it is that it may fall into the category that I call the X-factor, where we must accept the possibility of, and allow for, audible differences that we can't (yet) measure.
The way I look at it, if you believe that high-end audio cables sound different from #12 Home Depot cable, then you must equally allow for even a 1 uH inductor sounding different. Apart from those few high-end cables that deliberately do some funny business, it is actually not easy in many cases to come up with measured differences between audiophile speaker cable and Home Depot #12 that engineers would normally find to be relevant to a sound difference.
Please don't misunderstand what I am saying here - I am NOT saying that people should use Home Depot speaker wire; I'm just saying that there is an analogy here (possibly) between audibility of speaker cables and audibility of an output inductor.
Cheers,
Bob
Bob, did you use an output coil in your 1984 mosfet power amp? If no, then just like you did!
john curl said:
Hi John,
Yes I did. It was 0.5 uH and 0.5 ohms ("decades ago"). That reduced the damping factor at 20 kHz to about 125.
Without the coil, the damping factor was still about 3000 at 50 kHz.
This information is in my amplifier paper at www.cordellaudio.com.
So, John, what was the answer to my question?
Cheers,
Bob
I would stick to .5 uH, if I were you. My schematics are not published, and I'm not going to give them out here, either.
john curl said:
Huh? An output inductor affecting slew rate? Agreed, it does affects rise and fall times, but it has nothing to do with SR.