Slew rate quality point
- Thread starter mfaz28
- Start date
HTWillie
Super Member
I don't recall off the top o' me head. I do remember reading some time ago on Nelson Pass' First Watt site one of his musings on amp design. He was talking about how slew rate for a while was a one-upmanship spec, and he decided to conduct some tests.
The test signal he think he used was a pop or click from a vinyl record as it's a handy example of a signal with very fast rise time and very high amplitude. Damned if I can even remember his conclusion apart from the actual numbers.
The test signal he think he used was a pop or click from a vinyl record as it's a handy example of a signal with very fast rise time and very high amplitude. Damned if I can even remember his conclusion apart from the actual numbers.
soundmotor
super modified
When referring to opamps & sound quality a general rule of thumb is that more is better and more starts around ~9 (5532).
mfaz: I'd guess one should be able to determine the minimum requirement by calculation, shouldn't one?
Example: Let's assume we'd like the amp to deliver 100 W into 8 Ohm, and it should play up to 50 kHz. The former means we'd need not quite 28.3 V effective, so that would be pretty exactly 40 V peak - and the latter means that a complete sine wave would take 20 µs, so 5 µs for a quarter wave. And 40 V in 5 µs would equal 8 V/µs. I'd guess for a "ballpark approximation" that calculation would already do. For a more exact calculation we'd have to additionally consider that with sine-shaped signals we wouldn't have a linear rise or fall per quarter wave. I.e., if we look at the first quarter of a sine wave, the initial slew rate obviously needs to be a bit higher compared to a linear rise...
I'd like to leave a refined calculation to others, though - 'cause I'm somewhat tired today and have already forgotten too much of my high school maths anyway. And it's Sunday. And I need a coffee. And... and... und überhaupts! (~ "and anyway" = Bavarian general point, one would finally make, as soon as one has run out of others, to indicate that one could easily add some more, even if one actually couldn't...
).
Anyway, as the example shows, there would be no general answer to your question, 'cause the required slew rate would depend on the desired output voltage level and upper frequency limit.
Geetings from Munich!
Manfred / lini
Example: Let's assume we'd like the amp to deliver 100 W into 8 Ohm, and it should play up to 50 kHz. The former means we'd need not quite 28.3 V effective, so that would be pretty exactly 40 V peak - and the latter means that a complete sine wave would take 20 µs, so 5 µs for a quarter wave. And 40 V in 5 µs would equal 8 V/µs. I'd guess for a "ballpark approximation" that calculation would already do. For a more exact calculation we'd have to additionally consider that with sine-shaped signals we wouldn't have a linear rise or fall per quarter wave. I.e., if we look at the first quarter of a sine wave, the initial slew rate obviously needs to be a bit higher compared to a linear rise...
I'd like to leave a refined calculation to others, though - 'cause I'm somewhat tired today and have already forgotten too much of my high school maths anyway. And it's Sunday. And I need a coffee. And... and... und überhaupts! (~ "and anyway" = Bavarian general point, one would finally make, as soon as one has run out of others, to indicate that one could easily add some more, even if one actually couldn't...
).Anyway, as the example shows, there would be no general answer to your question, 'cause the required slew rate would depend on the desired output voltage level and upper frequency limit.
Geetings from Munich!
Manfred / lini
mfaz: Not sure, whether I should interpret your reply straight or rather as ironical. If the latter were the case, you'll need to accept that the variables due to the dependency on output voltage and upper frequency limit simply are too large for a generalised statement like x V/µs are an indicator for a good amp design. I.e., one user will be happy with 20 kHz, the other will demand 100 kHz -> factor 5; one user will be happy with 10 W into 8 Ohm or roundabout 12.65 Volt peak, the other will demand 1000 W into 8 Ohm or roundabout or roundabout 126.5 Volt peak -> factor 10. So combined that's factor 50 - simply too much that a particular absolute value could be an adequate answer.
Greetings from Munich!
Manfred / lini
Greetings from Munich!
Manfred / lini
There is no single answer to your question: How slow is too slow ?
I maintain that there is no statistically sound, empirical data when it comes to comparing audio amplifiers.
A lot of people accept the premise that all amplifiers sound the same.
You'll discover that this topic is quite contentious here.
Last edited:
Perhaps this interview in Stereophile is what you recall where the answer was 50 uV/sec.
E-Stat: Me thinks you'd need to move that µ a bit...
And maybe one should add, that the quoted value was mentioned in the context of a 200 Watt amp and very short impluses of high intensity. So, as already predicted, an answer in form of a particular value will only make sense in context with a certain output voltage level (or certain power output into a certain load) and a certain upper frequency limit.
dlucy: Yup, 120 V/µs for the hk825 and 160 V/µs for the matching hk870, iirc - with the bandwidth reaching up to 180 kHz (-3 dB) for each. I didn't really buy my own 825/870 combo for that, though - but rather because it seemed nice and attractively priced, as it was on sale-out at that time. The only drawback was that they only had the 825 in champagne and the 870 in black - but that didn't disturb me. In fact it turned out that I even like it, that the pre-amp sticks out a bit from my otherwise predominantly black gear.
Just a pity I haven't used the combo for too long after a move - so that it now would most probably be in need of a thorough restoration. And as I couldn't do that myself (already for insufficient workspace, but also for it not being unlikely to exceed my skills...), a professional restoration would probably cost me about as much as the combo cost me new back then... Hence I'm still debating with myself, what I should do. Because on one hand I don't really have a need for it, especially as I listen via headphones most of the time anyway, but on the other hand I already own it, and it's still in pretty immaculate cosmetic condition - and it's not like I had absolutely no use for it... So I'm rather torn in that matter.
Greetings from Munich!
Manfred / lini
And maybe one should add, that the quoted value was mentioned in the context of a 200 Watt amp and very short impluses of high intensity. So, as already predicted, an answer in form of a particular value will only make sense in context with a certain output voltage level (or certain power output into a certain load) and a certain upper frequency limit.
dlucy: Yup, 120 V/µs for the hk825 and 160 V/µs for the matching hk870, iirc - with the bandwidth reaching up to 180 kHz (-3 dB) for each. I didn't really buy my own 825/870 combo for that, though - but rather because it seemed nice and attractively priced, as it was on sale-out at that time. The only drawback was that they only had the 825 in champagne and the 870 in black - but that didn't disturb me. In fact it turned out that I even like it, that the pre-amp sticks out a bit from my otherwise predominantly black gear.
Just a pity I haven't used the combo for too long after a move - so that it now would most probably be in need of a thorough restoration. And as I couldn't do that myself (already for insufficient workspace, but also for it not being unlikely to exceed my skills...), a professional restoration would probably cost me about as much as the combo cost me new back then... Hence I'm still debating with myself, what I should do. Because on one hand I don't really have a need for it, especially as I listen via headphones most of the time anyway, but on the other hand I already own it, and it's still in pretty immaculate cosmetic condition - and it's not like I had absolutely no use for it... So I'm rather torn in that matter.
Greetings from Munich!
Manfred / lini
whoaru99
Epic Member
Indeed, but only because what was really said or asked nearly always gets perverted through reductio ad absurdium.
Last edited:
The amp I'm working on now has a +/- 350 V/uS slew rating. It's a Sansui C-2101.
I've seen ratings as high as +/- 400 V/uS in some of the higher up Sansui models.
I guess the big numbers were being disregarded as unnecessary.
My Sansui AU-X701 is much more dynamic than my Parasound HCA-750A amps. Perhaps some of John Curl's MOS amps would be a better comparison.
I've seen ratings as high as +/- 400 V/uS in some of the higher up Sansui models.
I guess the big numbers were being disregarded as unnecessary.
My Sansui AU-X701 is much more dynamic than my Parasound HCA-750A amps. Perhaps some of John Curl's MOS amps would be a better comparison.
Last edited:
I missed your comment when you first made it. Thanks for sharing! I have subsequently acquired a hk870 to go with the hk825. I am looking forward to the day when I can repair both and try them out. The hk870 has a massive in-a-can power transformer and very few parts, but there are many copper screws. So unusual. My hk825 looks like a small bomb exploded in the power supply area. At least one transistor blew and several components' legs have been eaten through by the corrosive glue they used. What a mess. But it is a challenge and I really want to hear what these two pieces sound like.
No problem, Doug - and neat avatar pic, btw!
And to be honest: So far I haven't dared to have a thorough look inside my own hk825/870 combo to check its current condition - to save myself from the rather nasty surprise (or maybe better "non-surprise" in this case), that would most likely await me. And yup, that problem with the acidic glue is really unpleasant - quite a pity, because otherwise it really was a nice pre-/power-amp combo. I remember its sound as very lively/dynamic and otherwise very "neutral" (in its best sense) - and the hk870 always made the impression, that it could hardly be brought to its knees at all.
Greetings from Munich!
Manfred / lini
And to be honest: So far I haven't dared to have a thorough look inside my own hk825/870 combo to check its current condition - to save myself from the rather nasty surprise (or maybe better "non-surprise" in this case), that would most likely await me. And yup, that problem with the acidic glue is really unpleasant - quite a pity, because otherwise it really was a nice pre-/power-amp combo. I remember its sound as very lively/dynamic and otherwise very "neutral" (in its best sense) - and the hk870 always made the impression, that it could hardly be brought to its knees at all.
Greetings from Munich!
Manfred / lini
Hah! I think your avatar of Peace is cooler than just me in a Russian space program training helmet.
