"Rather than use a single driver for multiple parallel CF outputs, use instead multiple drivers, one per MOSFET device. It sounds very brute force, . . . "
This is somewhat analogous the the idea, "Power Slice", Linear Technology promotes for its LT1166. You could build a bunch of output module complete with their own heatsink then daisy chan as many as you need for a devised output. Of cource one goal is to sell as many LT1166's as possible.
This is somewhat analogous the the idea, "Power Slice", Linear Technology promotes for its LT1166. You could build a bunch of output module complete with their own heatsink then daisy chan as many as you need for a devised output. Of cource one goal is to sell as many LT1166's as possible.
tcpip,
My patience is intact, although I can't speak for others. It's just that it's difficult to resolve things like this easily. H&H is as close to an "audio textbook" as we're likely to have in the near future (I am specifically not including Self and Slone--others may feel differently--for what it's worth, I'd take Self over Slone any day of the week). Nelson Pass has produced a series of projects that have instructive value in addition to being working circuits that you can actually listen to. I, myself, get more ideas from reading patents, although the writing is sometimes rather opaque. Lacking a definitive, dedicated "audio" book, you have to learn on your own.
Note the number of posts you see from EEs who find that their courses did not cover some of the things that are considered fundamental in audio. We're off in a very small corner of the electronics field with very little overlap between our area of interest and any other part of electronics, with the arguable exception of radio.
If it was easy, it wouldn't be so much fun as a hobby. Cookbook recipies don't interest me much. I want to roam around a little.
As far as .01 resistors go, you'll have to see what's available. If all else fails, use .1 ohm. You should be able to find those. I used to have a store that stocked a lot of NTE stuff locally. They went out of business last summer, so I'm reduced to mail order now. If my perception of the world is correct, things will get worse before they get better. The world economy will be variable in the near future, poor after that. Go ahead and think mail order if you can't get something locally. You won't need 1% parts. When dealing with 5% or greater parts, I measure the resistance and write on the side of the piece with an indelible marker. Easy to do if your're using a white ceramic resistor. Don't worry about inductance for my test--the bias at idle (i.e. DC) will answer your question. Buy cheap resistors, measure them, write it on the side, and go.
Try not to let the people who sneer at you keep you from trying things. It's a pain the the (*ahem*) sometimes, but there are people who get their jollies from criticizing others. Makes 'em feel important. Just keep plugging and you'll get there.
Incidentally, I'm not saying that H&H is a bad book. It's just that we're in a niche market with our own odd requirements. By all means, keep reading the book, but remember that some things may have to be seen from a different point of view.
Might I suggest the word experience instead of the phrase common sense? As you gain experience, it sometimes begins to seem like common sense, but you have to keep in mind that none of us were born knowing the difference between MOSFETS, tubes, and bipolars. We had to learn. Some people take that learning process for granted. I had to fight too hard to learn what I know. Hopefully the path will be somewhat easier for others.
Grey
My patience is intact, although I can't speak for others. It's just that it's difficult to resolve things like this easily. H&H is as close to an "audio textbook" as we're likely to have in the near future (I am specifically not including Self and Slone--others may feel differently--for what it's worth, I'd take Self over Slone any day of the week). Nelson Pass has produced a series of projects that have instructive value in addition to being working circuits that you can actually listen to. I, myself, get more ideas from reading patents, although the writing is sometimes rather opaque. Lacking a definitive, dedicated "audio" book, you have to learn on your own.
Note the number of posts you see from EEs who find that their courses did not cover some of the things that are considered fundamental in audio. We're off in a very small corner of the electronics field with very little overlap between our area of interest and any other part of electronics, with the arguable exception of radio.
If it was easy, it wouldn't be so much fun as a hobby. Cookbook recipies don't interest me much. I want to roam around a little.
As far as .01 resistors go, you'll have to see what's available. If all else fails, use .1 ohm. You should be able to find those. I used to have a store that stocked a lot of NTE stuff locally. They went out of business last summer, so I'm reduced to mail order now. If my perception of the world is correct, things will get worse before they get better. The world economy will be variable in the near future, poor after that. Go ahead and think mail order if you can't get something locally. You won't need 1% parts. When dealing with 5% or greater parts, I measure the resistance and write on the side of the piece with an indelible marker. Easy to do if your're using a white ceramic resistor. Don't worry about inductance for my test--the bias at idle (i.e. DC) will answer your question. Buy cheap resistors, measure them, write it on the side, and go.
Try not to let the people who sneer at you keep you from trying things. It's a pain the the (*ahem*) sometimes, but there are people who get their jollies from criticizing others. Makes 'em feel important. Just keep plugging and you'll get there.
Incidentally, I'm not saying that H&H is a bad book. It's just that we're in a niche market with our own odd requirements. By all means, keep reading the book, but remember that some things may have to be seen from a different point of view.
Might I suggest the word experience instead of the phrase common sense? As you gain experience, it sometimes begins to seem like common sense, but you have to keep in mind that none of us were born knowing the difference between MOSFETS, tubes, and bipolars. We had to learn. Some people take that learning process for granted. I had to fight too hard to learn what I know. Hopefully the path will be somewhat easier for others.
Grey
To expand on the post by GRollins, it becomes obvious (to me anyway) after reading this form and a couple of books that there is more than one way to build a good amplifier. However, you have to deduce that from information, often contradictory, provided by people who believe they have found the one true path.
Most important build SOMETHING! You won't really learn what works until you (try to) power it up. I don't think there is any source, book or forum poster, that is 100% right about all things audio, but excluding the obvious snakeoil salesmen nearly all have some gem of informatiom. Some have more gems than others.
Most important build SOMETHING! You won't really learn what works until you (try to) power it up. I don't think there is any source, book or forum poster, that is 100% right about all things audio, but excluding the obvious snakeoil salesmen nearly all have some gem of informatiom. Some have more gems than others.
I stand with one foot in the tube camp and the other in the solid state camp, so I can agree that there are more ways than one to build a good circuit. People who insist that there's only one way and it's their way are best ignored. They may be right some of the time, but it's not always worth the trouble to sort through the bluster in search of useful information.
I certainly don't claim to know everything about electronics. If I did I'd get bored and move on to something else where I could learn things. I'm only truly happy when I'm learning something.
Pay forward sounds good to me. It shouldn't be so confounded hard to learn this stuff, but we're reduced to picking up crumbs here and there and it takes literally years to get even the rudiments of the audio trade under control.
The irony is that reproduced music touches nearly everyone in the world sooner or later. Yes, there are aboriginal tribes in Australia and the Amazonian rain forest who may not have even a boom box, but expressed as a percentage, the vast majority of humanity uses some sort of electronic device to listen to music. Given that large a potential customer base and high demand (home, car, and personal portables, not to mention pro gear for concerts), audio designers should be more in demand. You'd think that universities would be churning out scores of them every time graduation rolled around. The reality is that there are very few people involved. A few mega corporations such as Sony and Pioneer control the lion's share of the market. They sell a large number of units of a limited number of models and their design teams are comparatively small. Then there's the sad fact that they design to a price point instead of designing for performance. The high end designers try to take up the slack, but their product is expensive. If you have sufficient discretionary income you can have a very fine system indeed.
The rest of us have to learn...and do...on our own.
Grey
I certainly don't claim to know everything about electronics. If I did I'd get bored and move on to something else where I could learn things. I'm only truly happy when I'm learning something.
Pay forward sounds good to me. It shouldn't be so confounded hard to learn this stuff, but we're reduced to picking up crumbs here and there and it takes literally years to get even the rudiments of the audio trade under control.
The irony is that reproduced music touches nearly everyone in the world sooner or later. Yes, there are aboriginal tribes in Australia and the Amazonian rain forest who may not have even a boom box, but expressed as a percentage, the vast majority of humanity uses some sort of electronic device to listen to music. Given that large a potential customer base and high demand (home, car, and personal portables, not to mention pro gear for concerts), audio designers should be more in demand. You'd think that universities would be churning out scores of them every time graduation rolled around. The reality is that there are very few people involved. A few mega corporations such as Sony and Pioneer control the lion's share of the market. They sell a large number of units of a limited number of models and their design teams are comparatively small. Then there's the sad fact that they design to a price point instead of designing for performance. The high end designers try to take up the slack, but their product is expensive. If you have sufficient discretionary income you can have a very fine system indeed.
The rest of us have to learn...and do...on our own.
Grey
But then there are different ways to experience music. The way that a lot of my friends listen to music does not require all that realistic a reproduction. So I guess the demand for discerning audio system designers will always be low, don't you think?GRollins said:
So true.
I came into diyaudio because I have more taste than money. But now that I'm in, I feel that I'd stick to diy audio even if I had lots of money. There are things which commercial designers simply don't attempt, purely for pragmatic reasons. For instance, how many active multi-amped speakers exist commercially? And this in spite of most people firmly believing that it's easier to build good line-level xo than high-current ones.GRollins said:
that's if everthing else is held constant. Unfortunate, economy of scale and fierce competition will drive out specialty designers, or anyone who failed to expand. That's why we had so many designers / audio companies back then than we do now.
Plus, technology has advanced to such a point that inexpensive boxes today can compete in performance with expensive boxes back then and now.
GRollins said:
Because I think we are approaching the limit of human hearing, with even the consumer grade boxes. I am one of those who don't believe in 0.0000007% THD, or egyptian maple boxes or golden feet, magic sound stages, 'positively gray', or 'shortest signal path', to name a few. I think most of us, if put to an objective test, will not be able to detect 99% of the things we claim to be able to detect.
The fastly dwindling audio designer circle is a testiment to that, in my view.
I wonder about this, though.millwood said:
I too am the kind of person you describe. I bought Van den Hull cables and couldn't hear any difference with my home-grown cables. I can't hear a difference between an NE5532, a TL082, and an OPA2134. And in terms of approach, I find Randy Slone and Doug Self's approach the one that makes sense. (As far as I have seen, they believe in measurements, comparison testing, listening tests, etc, but are unwilling to entertain things which do not have an explanation in theory, e.g. the directionality of audio cables.)
So, to that extent, I am the kind of person you describe: I can't hear many of the differences that many others on this forum seem to.
However, the difference between a simple home-grown pair of speakers and the ones that come with combo systems (which 99.99% of users seem to buy for their homes today) is so dramatic that I (and all my non-diy non-audiophile friends) can hear it. Thus it seems that there are major improvements possible.to the sound of modern music systems. Why don't we see more of it happening in commercial audio systems?
I actually have a theory about why not.
I believe that there's such a thing as "adequate quality" in free markets. This means that the market will not always opt for a better product, even if the price increase is affordable. The market often seems to have a concensus about a "good enough" level, and the bulk of the demand comes at that level, not higher. I've seen it in modems. Good modems could hold connections much longer with India's noisy telephone lines, but today no one buys anything other than the cheapest ones, even though noisy lines are common even now. And I'm seeing it in music systems.
Amplifiers are one thing, speakers are another. Even if a perfect speaker were definable, it will sound different in different rooms or different locations in the same room.
An additional thing to think about is that as one ages, hearing declines. Think about what that means as far as harmonics are concerned. If you calculate THD with a limit on bandwidth, yiou start with HD beyond 10k mmeaning little and each year the cut-off gets lower and lower. THD is probably more important because it is and indirect indicator of IM which may be less convenirnt to measure.
Other things I've found in commercial amps that could be avoided include buzzing transformers, off-on thumps, ground loops, ripple contamination the signal and inability to drive Magnepans (low sensitivity and loww impedance) with out suffereing premature death. So there are still some thing you can go for in diy amps that add to the quaity of listening even if THD hads been puhed below audibility.
One other thing about THD and THD+N. It is cummulative. Your amp in itself may well have performance beyond the audible, but is is not beyond the realm of possibility that it could add just enough to signal it takes as input to cross the threshold of audibilty. If I were designing an amp I would want it's noise and distortion far enough below the audible that it could not be guilty of being the "final straw".
An additional thing to think about is that as one ages, hearing declines. Think about what that means as far as harmonics are concerned. If you calculate THD with a limit on bandwidth, yiou start with HD beyond 10k mmeaning little and each year the cut-off gets lower and lower. THD is probably more important because it is and indirect indicator of IM which may be less convenirnt to measure.
Other things I've found in commercial amps that could be avoided include buzzing transformers, off-on thumps, ground loops, ripple contamination the signal and inability to drive Magnepans (low sensitivity and loww impedance) with out suffereing premature death. So there are still some thing you can go for in diy amps that add to the quaity of listening even if THD hads been puhed below audibility.
One other thing about THD and THD+N. It is cummulative. Your amp in itself may well have performance beyond the audible, but is is not beyond the realm of possibility that it could add just enough to signal it takes as input to cross the threshold of audibilty. If I were designing an amp I would want it's noise and distortion far enough below the audible that it could not be guilty of being the "final straw".
Hi,
be careful you don't mix up GDS(V & 201) and GSD(L).
Change your source polarity to suit either N or P type. The diagram is actually the same.
The difference between lateral and vertical types does not affect the measurement technique, only the voltage reading you get as a result.
Your 201s (Y grade) will have a Vgs voltage closer to a vertical type. O grade have a lower Vgs nearer Lateral type.
When carrying out the testing try to keep the Drain current constant. Either use a current limiter based on a 3pin regulator or use the adjustable current limit in a regulated PSU.
I use a plugboard for the connections. The fets don't fit into the holes but on mine can just touch the top of the recepticles.
I tested at 100mA (+- some unknown error) and found that 1058s matched very well out of the tube. The parallel quads measure very close when installed with 0r22 source resistors.
However the Pchannel 162s had a wider spread of Vgs. One out of the sixteen tested was significantly different. When installed the source current is about 70% higher than it's three partners. 60mA to 65mA vs 105mA for the low Vgs Fet. Target for each was 75mA. The measured Vgs @100mA was within 20mV from highest to lowest for any group of four except on the errant group where one Pch was 40mV low:- 1260mV, 1300mV, 1300mV, 1310mV. My heatsinks get too hot at Iq=400mA so had to settle for Iq=300mA at +-69V=41W dissipation.
Hope the testing procedure and background help.
be careful you don't mix up GDS(V & 201) and GSD(L).
Change your source polarity to suit either N or P type. The diagram is actually the same.
The difference between lateral and vertical types does not affect the measurement technique, only the voltage reading you get as a result.
Your 201s (Y grade) will have a Vgs voltage closer to a vertical type. O grade have a lower Vgs nearer Lateral type.
When carrying out the testing try to keep the Drain current constant. Either use a current limiter based on a 3pin regulator or use the adjustable current limit in a regulated PSU.
I use a plugboard for the connections. The fets don't fit into the holes but on mine can just touch the top of the recepticles.
I tested at 100mA (+- some unknown error) and found that 1058s matched very well out of the tube. The parallel quads measure very close when installed with 0r22 source resistors.
However the Pchannel 162s had a wider spread of Vgs. One out of the sixteen tested was significantly different. When installed the source current is about 70% higher than it's three partners. 60mA to 65mA vs 105mA for the low Vgs Fet. Target for each was 75mA. The measured Vgs @100mA was within 20mV from highest to lowest for any group of four except on the errant group where one Pch was 40mV low:- 1260mV, 1300mV, 1300mV, 1310mV. My heatsinks get too hot at Iq=400mA so had to settle for Iq=300mA at +-69V=41W dissipation.
Hope the testing procedure and background help.
As already stated earlier mosfets have negative tempco only above certain current level. Mostly when books claim that mosfets can be paralled just like that without matching they are talking about SMPS-stuff where mosfets are operated at high currents switching on or off but not in linear mode.tcpip said:
Okay, so there seems to be a disagreement between you and "The Art of Electronics". They say power MOSFETs can be freely paralleled without any source resistors (sorry about the drain vs source goof-up). They don't mention any need for matching. In fact they make it clear that it's the negative tempco which allows this to be done, thus (I presume) making the matching unnecessary.
I have the book at home.... I can quote giving page number if anyone wants. I found this comment mentioned in two places, Chapters 3 and 6.
Now I'm in a jam. Who do I believe?
For linear use like audio amplifier normal v-mosfets never operate at sufficiently high current levels to have negative tempco. So matching and current balancing resistors are good to have.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.