Hi,
You may be able to manage with just one pair (the Hitachi HCA7500 did, though it used feedback of course).
You can use two or three pairs quite successfully, unless you must killerwats into 0.1 Ohm loads or such.
Bipolars or Vertical Fet's may also require multiple devices to handle dissipation and currents required.
There are no "magic parts".
For me it is a question of being lazy.
Having already an amp that runs the frontend on +/-68V and the bipolar output stage on +/-56V it is easier (and in terms of time also cheaper) to throw the bipolars, emitter resistors etc. out, stick a bunch of matched Hitachi Fets in instead and be done, while having significant upgrades where I want them. I even already have the emitter follower needed to drive multiple pairs of them...
Now if I was developing a commercial amplifier to be made in 100's of pieces, I would probably look at designing a circuit that keeps the combined transconductance of the classic Class AB Bipolar output stage with emitter resistors constant at all current levels and avoids switching glitches. It can be done (the concept is quite trivial, but involves control loops to make it work), but it's hard work... I don't see the point for a one-off.
Ciao T
You may be able to manage with just one pair (the Hitachi HCA7500 did, though it used feedback of course).
You can use two or three pairs quite successfully, unless you must killerwats into 0.1 Ohm loads or such.
Bipolars or Vertical Fet's may also require multiple devices to handle dissipation and currents required.
There are no "magic parts".
For me it is a question of being lazy.
Having already an amp that runs the frontend on +/-68V and the bipolar output stage on +/-56V it is easier (and in terms of time also cheaper) to throw the bipolars, emitter resistors etc. out, stick a bunch of matched Hitachi Fets in instead and be done, while having significant upgrades where I want them. I even already have the emitter follower needed to drive multiple pairs of them...
Now if I was developing a commercial amplifier to be made in 100's of pieces, I would probably look at designing a circuit that keeps the combined transconductance of the classic Class AB Bipolar output stage with emitter resistors constant at all current levels and avoids switching glitches. It can be done (the concept is quite trivial, but involves control loops to make it work), but it's hard work... I don't see the point for a one-off.
Ciao T
Groove T, I feel that you have said something very useful.
It is true that fuses are non-linear devices. It fact it would be almost impossible to make them completely linear, as the internal temperature rise will effect virtually any metal and change its resistance over changes in current. Circuit breakers may be a partial solution, and they well may be more linear. They have their problems too, but not as bad, especially with high current situations.
It is true that fuses are non-linear devices. It fact it would be almost impossible to make them completely linear, as the internal temperature rise will effect virtually any metal and change its resistance over changes in current. Circuit breakers may be a partial solution, and they well may be more linear. They have their problems too, but not as bad, especially with high current situations.
And how exactly would this manifest itself in the output of a given circuit?
Keeping in mind that the typical linear power supply is drawing current in 120 Hz haversine pulses and all the associated harmonics that produces.
se
From the power generation station to the power transformer the resistance should be less than .33 0hms as per the NEC. So how much resistance does the fuse and fuse holder add?
Luvin this thread... most of the discussion revolves around half-truths, leaving out data not supporting a given position... etc... a microcosm of the testosterone side of the internet for sure...
so a 12a fuse would add <2% to the impedance seen by a given device using it?? This assumes, of course, that the powerline feeding the xfrmr is properly maintained (not a given)
so a 12a fuse would add <2% to the impedance seen by a given device using it?? This assumes, of course, that the powerline feeding the xfrmr is properly maintained (not a given)
I doubt it. This is because a fuse is deliberately made nonlinear in order to increase to its peak temperature as quickly and accurately as possible.
This would be a worthwhile project for someone. It has been done, over the years, but usually concerning using a fuse in series with a loudspeaker, for example.
The problem is: That fuses, like light bulbs, change their resistance with current and time. This is a perfect vehicle for nonlinearity.
This would be a worthwhile project for someone. It has been done, over the years, but usually concerning using a fuse in series with a loudspeaker, for example.
The problem is: That fuses, like light bulbs, change their resistance with current and time. This is a perfect vehicle for nonlinearity.
It is the nonlinear source resistance and the nonlinear load resistance multiplied together that gives a clue as to why fuses are so critical. Non linear load resistance MAKES linear output voltage, so it does not necessarily add distortion. However, the nonlinear loading of the fuse, makes the problem worse than the fuse itself.
Critical how?
You keep talking about the mite on the *** of the flea on the *** of the dog, but can't seem to describe how that mite changes the dog's behavior.
se
Please quantify as well as qualify these effects. Ed does some nice work but things at -150dB don't do much for me. Follow this effect back to an equivalent input signal, people do make pico-voltmeters with ordinary fuses.
All this shows the wisdom of turning AC mains into DC, turning it into AC again, but at a much higher frequency, tranning it into the voltages you need, then putting it into DC again and finally filtering the trash out. Which, because it happens at much higher frequencies, is much easier than at 60Hz. On top of that, you might even apply some tricks to get a stable DC voltage that is load independant!! A trick no linear PS can do, even using fuses wired with oxygen free dollshitium.
In short, SMPS.
Now, these things are not perfect yet, which is good news. Because a lot of the issues that have been discussed in this thread have landed in an area of ever decreasing returns of investment and effort. To develop a truly audiophile SMPS might put everything on a higher level.
In short, SMPS.
Now, these things are not perfect yet, which is good news. Because a lot of the issues that have been discussed in this thread have landed in an area of ever decreasing returns of investment and effort. To develop a truly audiophile SMPS might put everything on a higher level.
Last edited:
vac, Many years ago, I worked for a distributor of a tube preamp with a switching supply like that. Stereophile Class A, back when Gordon Holt was doing those ratings (rather than the current crop of dermatologists and ad copywriters); just sayin' since some folks think that means something. It was a tough sell because it was light and ran cool. Prospective customers would heft it, heft an SP-3, then buy the Audio Research since it was satisfyingly heavy.
I learned a lot from that. There's more to selling "high end" equipment than performance. You're buying an "experience."
I learned a lot from that. There's more to selling "high end" equipment than performance. You're buying an "experience."
Dear SY, that's the way it might work and that's also why I think putting the Blowtorch into a solid block of machined aluminum is such a brilliant idea. It is a vehicle to sell a piece of no holds barred electronics that otherwise might not have found its way to the market.
At the same time, this thread is also about getting the best of the best even better. So, why not address one of the major issues that completely obviates the influence any power cord or fuse could ever have. Everybody who has ever put a scope on one of the rails of a working power amplifier knows how much is going on there, even at low power levels. And this is completely unnecessary, given the state of the art in this field.
High end is seriously running behind in the way it sources its currents. A thread like this could help, by the way, to change opinions. We might first need a 10K SMPS to do this, but why not?
In short, I take your words for a reinforcement of my position that much is to be gained in this area, with a caution added that market acceptance might be an issue.
regards,
Vac
Well, everyone, some people make successful audio products that seem to sell better than some others, and I suspect because they concern themselves with any potential problem, large or small. In power supplies, I would not, at first, suspect that fuses would be a great problem, but I know of hi fi modifiers who REMOVE fuses and get better sound, so maybe, higher quality fuses might be a better solution.
We addressed this problem with the BLOWTORCH by using solid state slow blow fuses that we auditioned before we used them. Their advantage is that they are HARD WIRED into the input of every transformer, removing all the rest of the 'built for economy' connections typical in a fuse holder. We have also used circuit breakers to replace input fuses on some models of Parasound. A real improvement, but then I do hear differences between amplifiers, even of my own design. What a concept! '-)
We addressed this problem with the BLOWTORCH by using solid state slow blow fuses that we auditioned before we used them. Their advantage is that they are HARD WIRED into the input of every transformer, removing all the rest of the 'built for economy' connections typical in a fuse holder. We have also used circuit breakers to replace input fuses on some models of Parasound. A real improvement, but then I do hear differences between amplifiers, even of my own design. What a concept! '-)
- Status
- Not open for further replies.