syn08 said:
I'm glad to note your keen perception.
Indeed, you are right and righteous, the moderators are only conspiring against exposing the core truth of the matter.
Please stay alert and on guard.
G.Kleinschmidt said:
And that is without NFB.
(Yes, I know, it should be obvious – open loop – nevertheless).
ironhardware said:
Come on JC, I'd rather read "their" pointless banter.
Indeed, all here are friendly, loving and supportive. Only some fail to grasp it…
anatech said:
Indeed, this was your response to another question of mine.
Your attitude to my question is what it is, nevertheless.
stinius said:
Thank you very much for your friendly, supportive and loving advise.
Highly appreciated.
Joshua_G said:
Thank you Joshua, I knew that you would understand it, love it and appreciate it.
BTW: It was meant as a good advice.
Cheers
Stinius
Hi John, stinius, all,
I wasn't aware any posts were deleted. However, there are posted rules about what is acceptable and not. I suspect that those who did have posts deleted know why.
Now, there is another reason why a post may be deleted. Happens to me often enough. When one post references another that gets deleted, it is no longer relevant and may actually be confusing to other members. Those go along with the offending post. So you don't actually have to post something "bad" to have it removed along with the offending post.
Hi Grey,
Most amplifier designs that I have seen use bandwidth limiting at the input. So that even if that amplifier is -3 dB to 50 MHz, no worries. Anyway, that is common practice in the industry so you should have no fears building a very fast amplifier. Besides, by definition, all frequencies that are out of "the audio band" are just noise. The way you define "the audio band" is up to you. 50 KHz, 100 KHz, whatever. The output coil and zobel network helps with signals coming in the "back door". They would normally be attenuated by the feedback network (1/gain), but that little parallel capacitor may provide a clear road in. You could still use an RF filter there if it was a grave concern.
-Chris
Edit: The missing posts didn't really have any technical content. You are not missing anything special there.
I wasn't aware any posts were deleted. However, there are posted rules about what is acceptable and not. I suspect that those who did have posts deleted know why.
Now, there is another reason why a post may be deleted. Happens to me often enough. When one post references another that gets deleted, it is no longer relevant and may actually be confusing to other members. Those go along with the offending post. So you don't actually have to post something "bad" to have it removed along with the offending post.
Hi Grey,
Most amplifier designs that I have seen use bandwidth limiting at the input. So that even if that amplifier is -3 dB to 50 MHz, no worries. Anyway, that is common practice in the industry so you should have no fears building a very fast amplifier. Besides, by definition, all frequencies that are out of "the audio band" are just noise. The way you define "the audio band" is up to you. 50 KHz, 100 KHz, whatever. The output coil and zobel network helps with signals coming in the "back door". They would normally be attenuated by the feedback network (1/gain), but that little parallel capacitor may provide a clear road in. You could still use an RF filter there if it was a grave concern.
-Chris
Edit: The missing posts didn't really have any technical content. You are not missing anything special there.
syn08 said:
G.Kleinschmidt said:
Grey, syn08, and G.Kleinschmidt all on the same side of an argument? I can't be reading what I'm reading. No frickin' way.
Which of us is buying the beer?
anatech said:
You raise an interesting point, and one that John tends to like to talk about in terms of various distortion measurements. To my way of thinking, a 'fast' amp that's bandwidth limited at the input and/or output is (to borrow John's long-running car analogy) like taking a Ferrari that's been fine-tuned to within a gnat's hair of perfection...and putting a rev. limiter on it. It takes all the fun out of it.
Yes, you can argue that you'll stay out of slew-related trouble by designing a circuit this way, and you're definitely better off in terms of potential RF trouble, but when I say I'm running .5MHz, I mean I'm running .5M all the way though, stem to stern. (Ooops...syn08 & G.Kleinschmidt just left the table...beers unfinished. I guess we're back to being enemies again. Bummer.) No bandwidth limitation at the input and (at least as far as MOSFET output stages go) no Zobel at the output.
Incidentally, anyone who wishes to limit the bandwidth of the GR-25 or anything else I've done is welcome to do so. I won't be insulted. I've got a buddy here in town who gets near-perfect reception of a Country music station on his system about three days out of the week. He hates Country. He's clearly in a position where a little narrowing of the bandwidth is called for, but doesn't want to modify his equipment.
Charles Hansen uses bipolar outputs with no Zobels. From things he's said, I think I've got an idea or two about how to proceed, but I've not done so yet. To date, I'm still using MOSFET outputs. It's on my to-do list.
Thanks for the hints, Charles.
Grey
scott wurcer said:
Scott,
I've saved a place at the table for you. You can buy the next round.
Grey
G.Kleinschmidt said:
Wow! 50W? How about 400W at 8 ohms, 20KHz and 0.005% THD? Open loop. I guess if you haven't done it, it can't be done.
I can't do it, that's for sure.
Forgive me, I have been away eating out with Jack Bybee at a distant casino, while his wife hit the slot machines! It is always good to get out of the apartment, once in a while.
Was I right, or was I not? Joshua?
Cordell's comments are indeed, the hit of the week.
Forgive me, I have been away eating out with Jack Bybee at a distant casino, while his wife hit the slot machines! It is always good to get out of the apartment, once in a while.
Was I right, or was I not? Joshua?
Cordell's comments are indeed, the hit of the week.
Grey,
I took the time to read your GR-25 thread. Now that I know you and your writing style better, perhaps I did respond improperly.
You said:
Scenario two, the approach favored by those who want to get their THD as low as possible, is to design for high open loop gain, so as to be able to apply plenty of negative feedback. The unavoidable consequence of this, unfortunately, is reduced bandwidth and higher distortion. How reduced? That depends on the design, obviously, but it's not unusual to see an amplifier with 1kHz open loop bandwidth. That means that if you want to get the bandwidth out to 20kHz, you're going to have to apply 26dB of feedback. Want 100kHz bandwidth? You're looking at 40dB of negative feedback. That's before you even begin to worry about distortion figures. Now, assuming that you've designed your circuit such that you have sufficient open loop gain to use that much feedback and still have your target gain for your circuit, you're in pretty good shape. If your open loop bandwidth is lower than 1kHz or if your distortion is higher than whatever you can handle with the available amount of feedback, you need to build in more gain, which in turn leaves you with even more distortion and even narrower bandwidth, so it's a delicate balancing act. If you play your cards right, you can achieve THD down in the .001% range, perhaps lower.
This, to me, implies you feel many designers consider a 1KHz open loop bandwidth acceptable. I do not. I manage both bandwidth and low distortion without feedback.
You said:
In solid state gear, I usually look for about 200-250kHz bandwidth, minimum. I usually come in higher than that--400-500kHz. I can get away with fairly wide bandwidth because there's little RF where I live. I've not done a formal study of what's on the market, but I've got a sense that the 'average' bandwidth these days is around 100kHz.
OK. That is not what you said in the previous post. I was working with what you had said.
You said:
From what you're saying, I'm assuming that you live in an area where RF isn't a problem. That's cool. But criticizing people for not meeting your bandwidth ideas is not necessarily a particularly good idea, as it could lead to fried circuits and/or tweeters, not just lower sound quality. Most people seem to think that I'm lunatic fringe for running .5MHz bandwidth. If you think that's too low, and if your living conditions permit, then by all means go for it. I'll support you when everyone else starts yammering about how crazy you are, because I happen to agree with you.
RF is not a problem where I live now. I can only see one cell tower out one window. A pair of my amps ran just fine less than a block from a TV broadcast tower. One did overheat and shut down when the owner took the top off to show the insides. After cool off and replacement of the cover, the amp and speakers were OK. You said:
I'm not quite sure how you got the impression that I believe low THD to be detrimental to sound quality, but...my philosophy is that if I want ice and water freezes at 32 degrees, then 25 degrees is sufficient. If you feel that 10 degrees produces better quality ice, then by all means make it as cold as you like.
It seemed (once again to me) that you felt 0.1% THD was good enough. I can do better, without feedback, so I do. Perhaps I misunderstood your position on that.
I took the time to read your GR-25 thread. Now that I know you and your writing style better, perhaps I did respond improperly.
You said:
Scenario two, the approach favored by those who want to get their THD as low as possible, is to design for high open loop gain, so as to be able to apply plenty of negative feedback. The unavoidable consequence of this, unfortunately, is reduced bandwidth and higher distortion. How reduced? That depends on the design, obviously, but it's not unusual to see an amplifier with 1kHz open loop bandwidth. That means that if you want to get the bandwidth out to 20kHz, you're going to have to apply 26dB of feedback. Want 100kHz bandwidth? You're looking at 40dB of negative feedback. That's before you even begin to worry about distortion figures. Now, assuming that you've designed your circuit such that you have sufficient open loop gain to use that much feedback and still have your target gain for your circuit, you're in pretty good shape. If your open loop bandwidth is lower than 1kHz or if your distortion is higher than whatever you can handle with the available amount of feedback, you need to build in more gain, which in turn leaves you with even more distortion and even narrower bandwidth, so it's a delicate balancing act. If you play your cards right, you can achieve THD down in the .001% range, perhaps lower.
This, to me, implies you feel many designers consider a 1KHz open loop bandwidth acceptable. I do not. I manage both bandwidth and low distortion without feedback.
You said:
In solid state gear, I usually look for about 200-250kHz bandwidth, minimum. I usually come in higher than that--400-500kHz. I can get away with fairly wide bandwidth because there's little RF where I live. I've not done a formal study of what's on the market, but I've got a sense that the 'average' bandwidth these days is around 100kHz.
OK. That is not what you said in the previous post. I was working with what you had said.
You said:
From what you're saying, I'm assuming that you live in an area where RF isn't a problem. That's cool. But criticizing people for not meeting your bandwidth ideas is not necessarily a particularly good idea, as it could lead to fried circuits and/or tweeters, not just lower sound quality. Most people seem to think that I'm lunatic fringe for running .5MHz bandwidth. If you think that's too low, and if your living conditions permit, then by all means go for it. I'll support you when everyone else starts yammering about how crazy you are, because I happen to agree with you.
RF is not a problem where I live now. I can only see one cell tower out one window. A pair of my amps ran just fine less than a block from a TV broadcast tower. One did overheat and shut down when the owner took the top off to show the insides. After cool off and replacement of the cover, the amp and speakers were OK. You said:
I'm not quite sure how you got the impression that I believe low THD to be detrimental to sound quality, but...my philosophy is that if I want ice and water freezes at 32 degrees, then 25 degrees is sufficient. If you feel that 10 degrees produces better quality ice, then by all means make it as cold as you like.
It seemed (once again to me) that you felt 0.1% THD was good enough. I can do better, without feedback, so I do. Perhaps I misunderstood your position on that.
- Status
- Not open for further replies.