What is the cheapest amp of commercial origin you might bother to use ? NAD 3020 me. To be honest the list ends there. Rotel is better but less flexible. Kenwood mostly were good and some Sansui ( AU 101 and AU 222 ). Yamaha A300 the poormans Quad.
I have sent out my friend out to buy me one as he goes to the charity shops. 3020's are so useful. Two even better as the PSU is then split. One for pre and one for power.
I once heard a 3020 with Linn Sara speakers. For once something could drive them. As someone said the designer turned a pile of junk into a reasonable amp. Others turn reasonable components into a pile of junk.
Come to think of it only Sony TA5650 beats it. A one off fluke from my un-favourite maker.
Sugden A48 is very good. I would not recommend my most favourite Quad 33/303 as most people contrive to make it fail. I would not recommend Dynaco A70 as the hype makes it impossible to live up to it's reputation. I detest the Quad 2/22. I mostly dislike the Quad 405. Quad 33/303 is like driving a 1960's RR, bettered by many cheap amps but never in it's class way of doing things. That class is also not to waste money. A modern Alps pot would be better.
The other would be any of the Crimson amps. To be honest after that speakers do the job for me. I have heard more super amps than most have had holidays. Most don''t do much for me. The $$ are always too big to make me take them seriously.
The amplifier I most disliked Radford HD 250. Daft as it is the type of circuit I like best. That horse must have fallen at the last fence somehow.
I have sent out my friend out to buy me one as he goes to the charity shops. 3020's are so useful. Two even better as the PSU is then split. One for pre and one for power.
I once heard a 3020 with Linn Sara speakers. For once something could drive them. As someone said the designer turned a pile of junk into a reasonable amp. Others turn reasonable components into a pile of junk.
Come to think of it only Sony TA5650 beats it. A one off fluke from my un-favourite maker.
Sugden A48 is very good. I would not recommend my most favourite Quad 33/303 as most people contrive to make it fail. I would not recommend Dynaco A70 as the hype makes it impossible to live up to it's reputation. I detest the Quad 2/22. I mostly dislike the Quad 405. Quad 33/303 is like driving a 1960's RR, bettered by many cheap amps but never in it's class way of doing things. That class is also not to waste money. A modern Alps pot would be better.
The other would be any of the Crimson amps. To be honest after that speakers do the job for me. I have heard more super amps than most have had holidays. Most don''t do much for me. The $$ are always too big to make me take them seriously.
The amplifier I most disliked Radford HD 250. Daft as it is the type of circuit I like best. That horse must have fallen at the last fence somehow.
The Cool Sound of Tubes - IEEE Spectrum
Nice look at tubes sent to me. Some small surprises that shouldn't be.
Nice look at tubes sent to me. Some small surprises that shouldn't be.
Perhaps so, but that doesn't make him any less good for that.
Attachments
Toshiba SB-420 (on some markets, under the name of Aurex).
A surprisingly capable amplifer for peanut money.
Good reading, to be sure, but ...
The choice of any semiconductor and/or tube for comparison is always highly debatable.
As is the topology of the stage itself. May deteriorate the S/N ratio, but vastly improve THD performance, or vice versa, or something in between. Too many variables.
Nice to see the old codger still can do the job ( tube) .
The JLH amp is fascinating. I suspect P type IGBT is no more . I was interview for a job doing plastic film wrapping using HT IGBT . The guy who interviewed me told me straight I would hate the job and worst still get an Opel car to drive( yep). Then I met the boss. I knew him from VOX amplifiers. His first statement was " you will get no design work here". How sad he chose money over respect. Do any of you know how good a VOX AC30 is ? I do.
These were to only amps I remember as being truly better than NAD's . Kendle Perry introduced me to them. He now writes about electronics and heads some electronic company somewhere. They were kits also.
The big DH MOS FET Hafler also I like.
About Crimson Products Limited
The big DH MOS FET Hafler also I like.
About Crimson Products Limited
There is slow, steady movement forward ... this review, Lampizator Level 7 DAC | Hifi Pig, conveys very nicely the type of sound one should be chasing - to get there one just does whatever it takes, whatever the means at one's disposal ...
I've been mulling over the idea of a mixed bag output stage, and I must say, I just don't see it happening.
The drive and general behavior characteristics of BJTs and MOSFETs are so different, I see no way to put them one after the other in a single stage whose whole point is to work as much in unison as possible. MOSFETs want voltage drive, BJTs need current drive. MOSFET switches off in like 100 nS, a BJT needs something likr 450 nS, MOSFET switches on in 45 nS, BJT needs something like 350 nS at best, and so forth.
It's like mixing apples and watermelons and calling it an applemelon pie.
The drive and general behavior characteristics of BJTs and MOSFETs are so different, I see no way to put them one after the other in a single stage whose whole point is to work as much in unison as possible. MOSFETs want voltage drive, BJTs need current drive. MOSFET switches off in like 100 nS, a BJT needs something likr 450 nS, MOSFET switches on in 45 nS, BJT needs something like 350 nS at best, and so forth.
It's like mixing apples and watermelons and calling it an applemelon pie.
SIM jockey's --- make a Sziklai (or similar topology) using both fet and bipolar.
THx-RNMarsh
THx-RNMarsh
Last edited:
That was what I meant, but it came out kinda wierd - my apologies.
My point was, the drive requirements are very different for BJT and MOSFET, if you want them to perform at maximum capability each. Marrying that into one system would the Devil's own work.
For example, how would you go about ensuring that each output pair delivers exactly the same portion of the signal? So that none pulls more than the other?
Scott, I wonder?
Have you ever tried using a medium power transistor (e.g. 2SC3503/2SA1831) as the first stage of current drive, followed by discrete Darlington transistor packs, so to speak, each one consisting of a say MJE 15030 driving its own say MJ 21196? That way, you eliminate the drive problem with the big MJ trannies since the composite current gain will be quite a lot, well over 60 dB, and you could be seeing current capability which would make even Wayne weep for joy.
That kind of power sharing (since each has its own driver, there is no need for any one single driver to do the whole chore) can produce amazing clarity and especially convincing bass notes. I tried it once, just for the hell of it, and for the life of me, I don't know why I didn't pursue the idea further.
As ever, there is one snag - one needs to keep the two in a pack as close together as possible, for thermal reasons.
Have you ever tried using a medium power transistor (e.g. 2SC3503/2SA1831) as the first stage of current drive, followed by discrete Darlington transistor packs, so to speak, each one consisting of a say MJE 15030 driving its own say MJ 21196? That way, you eliminate the drive problem with the big MJ trannies since the composite current gain will be quite a lot, well over 60 dB, and you could be seeing current capability which would make even Wayne weep for joy.
That kind of power sharing (since each has its own driver, there is no need for any one single driver to do the whole chore) can produce amazing clarity and especially convincing bass notes. I tried it once, just for the hell of it, and for the life of me, I don't know why I didn't pursue the idea further.
As ever, there is one snag - one needs to keep the two in a pack as close together as possible, for thermal reasons.
Last edited:
I would not expect much joy. The Ixys models are frequency specific. That does not bode well for a wideband circuit simulation. It also suggests the parts are not simple transfer functions. The notes suggest the are more like SCR's that don't saturate than linear devices.
<<< Listening Sense & Measurements Sense | The Necessary Non-Related Truth >>>
Click, Watch & Listen => Think you know what your music sounds like ? Think again... - YouTube
__________
"Perhaps more than any other discipline, audio engineering involves not only purely objective characterization but also subjective interpretations. It is the listening experience, that personal and most private sensation, which is the intended result of our labors in audio engineering. No technical measurement, however glorified with mathematics, can escape that fact." - Richard Heyser
__________
<<>> Insider with Robert Harley -- What a test sounds like | The Absolute Sound
Click, Watch & Listen => Think you know what your music sounds like ? Think again... - YouTube
__________
"Perhaps more than any other discipline, audio engineering involves not only purely objective characterization but also subjective interpretations. It is the listening experience, that personal and most private sensation, which is the intended result of our labors in audio engineering. No technical measurement, however glorified with mathematics, can escape that fact." - Richard Heyser
__________
<<>> Insider with Robert Harley -- What a test sounds like | The Absolute Sound
Are we describing the same topology? Ixys models? we want the advantages of the fet and the advantages of the bjt in one device or as two devices configured to give those IGBT benifits. Such as a, Sziklai or compound-complimentary topology. ??
THx-RNmarsh
Yes, why don't people pursue such, more?? 😕
It's obvious I don't fiddle around with the sort of subtleties being discussed here: FET versus BJT, OPS topologies, etc, etc. Because, IME this has very little to do with achieving convincing sound - what I'm after can quite largely be achieved with quite basic, cheap designs - the extra "goodness" brought to to the party by having exactly the right amplifier toplogy doing the work is just a bigger dollop of cream on the essential dish ...
Fetlingtons were once common devices and plenty written in magazines. I suspect a compound discrete device. From the little I known well liked in UK TV's. The expert my brother was is on the other side which means the easy source of info is lost to me. The net seems very uninformative on this.
With a little imagination a hybrid that simplifies the design of a monster amp is possible. A monster that can tread lightly. A monster with no heat problems affecting bias settings.
My ears tell me that overcoming the most obvious defect of MOSFET's might not be the best solution. Why people demand full power at 20 kHz is beyond me. I suspect it is because they like a mouth to a flame have learnt how to control a bipolar amp. It seems unreasonable to unlearn it. I suspect it is this defect that makes these amps sound so much more like real music. As my friend who used the PA versions said none of us want HF full power . If my way of seeing things is right it is only in extreme situations where we can reliably spot this problem and then rejoice that we have it. Same with tube amps and musicians. Say what you like they won't change and it isn't about clipping or a box of crayons . It is about nasty unpleasant sounds just when it was becoming interesting if trying to use the run of the mill hi fi amps with it's musical chlorine as a guitar amps. The Marshall amps played at home music levels are not unlike hi fi. There was a plan to build a hi fi Marshall to a Kondo design. I think John Curl has solved this problem if I read him correctly when he spoke of this some months back.
With a little imagination a hybrid that simplifies the design of a monster amp is possible. A monster that can tread lightly. A monster with no heat problems affecting bias settings.
My ears tell me that overcoming the most obvious defect of MOSFET's might not be the best solution. Why people demand full power at 20 kHz is beyond me. I suspect it is because they like a mouth to a flame have learnt how to control a bipolar amp. It seems unreasonable to unlearn it. I suspect it is this defect that makes these amps sound so much more like real music. As my friend who used the PA versions said none of us want HF full power . If my way of seeing things is right it is only in extreme situations where we can reliably spot this problem and then rejoice that we have it. Same with tube amps and musicians. Say what you like they won't change and it isn't about clipping or a box of crayons . It is about nasty unpleasant sounds just when it was becoming interesting if trying to use the run of the mill hi fi amps with it's musical chlorine as a guitar amps. The Marshall amps played at home music levels are not unlike hi fi. There was a plan to build a hi fi Marshall to a Kondo design. I think John Curl has solved this problem if I read him correctly when he spoke of this some months back.
To me the reason for doing this is force the design to be more competent than it would otherwise be; an analogy would be to make a car totally roadworthy at 150mph, fully stable and well behaved at extreme speeds. This then guarantees at much more modest performance levels a complete competence in every area, it is always at its ease ...
- Status
- Not open for further replies.