You know, I have been designing solid state power amplifiers for almost 50 years, and I find many here with a lot less experience than me, and therefore prone to make the same errors that I had made up to about 45 years ago. It's time to catch up with what modern designers know. We make amps with 50V/us or more, if at all possible. Rise time is usually 5us or less. Heck, I made a famous power amp design more than 45 years ago with 100V/us slew rate with 4MHz f(t) transistors. I have never looked back to the early days of 10-20V/us for the last 45 years since either, because my faster designs, as well as those from Matti Otala, just sounded better than anything slower, that was available at the time and probably even today.
You just can't sit in an armchair and decide that slew rate is unimportant, or that a rise time of 10us isn't possible in the real world. I think we call that sort of thing 'sophomoric'.
You just can't sit in an armchair and decide that slew rate is unimportant, or that a rise time of 10us isn't possible in the real world. I think we call that sort of thing 'sophomoric'.
Rise time values higher than 10 µS exist in the real world,
but are certainly much more rare at the output of most microphones.
Just to clarify. You're not saying that you've been recycling the same design for 45 years?
Thanks to bring it to my attention.
I really dislike this kind of error,
particularly when I am its author.
Question for Mr. Curl...
John, I deeply respect the innovations and designs you've added to the opus of quality, fresh innovation in audio engineering in the last 40+ years. I have three questions which have been bugging me for decades.
ONE - in relation to a comment earlier in this thread: “I had no amplifiers below 100 watts [per channel]” (paraphrased). … I'm almost certain you imply “for 8 Ω load”, which in turn requires at least √( 100 W × 8 Ω ÷ 2 RMS factor ) or 20 VRMS which only for sines would be 28.3 peak volts. The question is, what level of overvoltage headroom do you design in?
I know this is a question fraught with side analysis. Nature and content of the music. Degree of 'headroom' built into the recordings. And I'm almost sure that you've addressed this elsewhere, but it is hard to find. Keeping it simple, do you design for 50% peak-before-clipping voltage headroom? 100%? 150%?, since all amplifiers must have some upper limit before clipping.
TWO - Is it the case that a “100W with 250W dynamic output capacity” amplifier also having (say) 50 V/μs performance, has better performance at more modest (and yes, let's call it “musical”) levels? The question reflects on my personal mmm… theory or analysis or something between: played at more modest levels, all amplifiers (short the miracle few) perform more accurately, having to fight a whole lot less speaker-motor reverse EMF, than at the highest levels.
THREE - Is not one of the most important design criteria (which is admittedly very hard to numerically resolve), the matching of the damping factor (servo-like output voltage tracking viz a vis input signal) of an otherwise competent amplifier to the intended set of speakers? Again, perhaps a personal theory or predilection, but it seems to me (from physics of second and third order derivatives and oscillatory/energy-storing systems) that there must be a critically important tracking-factor that both “leads” the speaker cone excursions, and that also “binds” the speakers to the evolving signal.
Humbly, I look forward to your response.
GoatGuy
John, I deeply respect the innovations and designs you've added to the opus of quality, fresh innovation in audio engineering in the last 40+ years. I have three questions which have been bugging me for decades.
ONE - in relation to a comment earlier in this thread: “I had no amplifiers below 100 watts [per channel]” (paraphrased). … I'm almost certain you imply “for 8 Ω load”, which in turn requires at least √( 100 W × 8 Ω ÷ 2 RMS factor ) or 20 VRMS which only for sines would be 28.3 peak volts. The question is, what level of overvoltage headroom do you design in?
I know this is a question fraught with side analysis. Nature and content of the music. Degree of 'headroom' built into the recordings. And I'm almost sure that you've addressed this elsewhere, but it is hard to find. Keeping it simple, do you design for 50% peak-before-clipping voltage headroom? 100%? 150%?, since all amplifiers must have some upper limit before clipping.
TWO - Is it the case that a “100W with 250W dynamic output capacity” amplifier also having (say) 50 V/μs performance, has better performance at more modest (and yes, let's call it “musical”) levels? The question reflects on my personal mmm… theory or analysis or something between: played at more modest levels, all amplifiers (short the miracle few) perform more accurately, having to fight a whole lot less speaker-motor reverse EMF, than at the highest levels.
THREE - Is not one of the most important design criteria (which is admittedly very hard to numerically resolve), the matching of the damping factor (servo-like output voltage tracking viz a vis input signal) of an otherwise competent amplifier to the intended set of speakers? Again, perhaps a personal theory or predilection, but it seems to me (from physics of second and third order derivatives and oscillatory/energy-storing systems) that there must be a critically important tracking-factor that both “leads” the speaker cone excursions, and that also “binds” the speakers to the evolving signal.
Humbly, I look forward to your response.
GoatGuy
Last edited:
Forget "Damping Factor"
Look instead at the total impedance feeding the speaker and determine whether that speaker requires less, or the same, or more impedance to operate as you need.
In my biased view of advertising script writers, I think they invented that phrase to impress the gullible buyer.
eg.
"Our 1000 Damping Factor amplifier is far better than all our competing Manufacturers' Damping Factor of only 500"
Completely useless nonsense.
Look instead at the total impedance feeding the speaker and determine whether that speaker requires less, or the same, or more impedance to operate as you need.
In my biased view of advertising script writers, I think they invented that phrase to impress the gullible buyer.
eg.
"Our 1000 Damping Factor amplifier is far better than all our competing Manufacturers' Damping Factor of only 500"
Completely useless nonsense.
AVA amplifiers with infinite slew
Last year I was about to buy a Frank Van Alstine hybrid amplifier with Exicon Lateral MOSFET: Fet Valve 400R
avahifi.com Power Amplifier Buying Guide
Fet Valve 200R, Why not, Frank? by maty
-> Fet Valve 200R, Why not, Frank?
Last year I was about to buy a Frank Van Alstine hybrid amplifier with Exicon Lateral MOSFET: Fet Valve 400R
avahifi.com Power Amplifier Buying Guide
By the way, the amplifier bandwith of Fet 400R/600R is about 350 kHz.
Fet Valve 200R, Why not, Frank? by maty
-> Fet Valve 200R, Why not, Frank?
If an amp can have an "infinite slew factor" then I can only assume one of two explanations:
1. alien technology,
2. "slew factor" is something different from, and less demanding than, 'slew rate limit'.
Of course, there is a third explanation for this remarkable claim but it would be unkind to suggest it.
1. alien technology,
2. "slew factor" is something different from, and less demanding than, 'slew rate limit'.
Of course, there is a third explanation for this remarkable claim but it would be unkind to suggest it.
I suppose (and everyone) "infinite" = very high
Very fast -> slew rate very high and rise time very low
Job, Goldmund and others use Exicon lateral MOSFET too. Better with double die => not problem with any impedance (8, 4, 2).
JOB Products
The JOb 225: (maty: better with > 4 Ohms because is not double die)
SPEED
SPEED
Very fast -> slew rate very high and rise time very low
Job, Goldmund and others use Exicon lateral MOSFET too. Better with double die => not problem with any impedance (8, 4, 2).
JOB Products
The JOb 225: (maty: better with > 4 Ohms because is not double die)
SPEED
- Slew rate : > 60 V/us.
- Rise time : < 200 ns.
SPEED
- Slew rate unloaded > 80 V/us.
Last edited:
Only in Alice's world. On my planet infinite means infinitely more than very high.maty tinman said:
I will be so unkind. Perhaps the amp rolls off before it can ever hit a slew limitation. If it never slew-limits, one could claim (in a puffery sense) "infinite slew factor."
Infinite slew rate would require infinite bandwidth and infinite power. I suggest we divide and conquer. You build the amp. I go work on the nuclear reactor in the backyard... 🙂
Tom
Tom
Back to reality . . . we're all looking for a teenage girl to test her hearing. And her "slew rate".
Well, yes. Some are. But many are not. This seems to be a common refrain that leaves out so much of the rest of the signal.
Hi-End Class A-AB Hybrid Amplifier - the reference with "incredible slew-rate"
Allow me a small digression:
Hybrid Amplifier by Andrea Ciuffoli
Allow me a small digression:
Hybrid Amplifier by Andrea Ciuffoli
Slew limiting is non linear, it must different from bandwidth limiting (ex. using Resistor and Capacitor).
But I did not find any double blind test some amplifier with different slew rate (huge different), only bandwidth limiting at input amplifier.
Yeah, so true.
I was introduced to this forum by a coworker who restarted my interest for hi-end audio, but I mostly get some good laughs when I visit this forum.
Good laughs indeed 😀
- Status
- Not open for further replies.