Hi Gr and Curl,
I am not looking for A class amp what curl think about first 10 watt is resonable to me what I like to know is something like JC-1 amp that mr curl said use +/- 90 volt and use 9 pair of output devices. I believe as mr Cur said it use a bipolar cause its reliability in short output mode. in other post mr Curl also said he preferable use all FET ( MOSFET ?) so if He dont mind to shed me a light I want to know what in his mind for Output device,
I am not looking for A class amp what curl think about first 10 watt is resonable to me what I like to know is something like JC-1 amp that mr curl said use +/- 90 volt and use 9 pair of output devices. I believe as mr Cur said it use a bipolar cause its reliability in short output mode. in other post mr Curl also said he preferable use all FET ( MOSFET ?) so if He dont mind to shed me a light I want to know what in his mind for Output device,
I'll save John the trouble: Topology.
Just specifying rails and output devices is a rather haphazard way to go about designing an amp. You might want to consider determining the load and desired wattage, then working backwards from that.
Note that the JC-3 front end will not work at such high voltages.
Grey
Just specifying rails and output devices is a rather haphazard way to go about designing an amp. You might want to consider determining the load and desired wattage, then working backwards from that.
Note that the JC-3 front end will not work at such high voltages.
Grey
GRollins
I'll save John the trouble: Topology.
Just specifying rails and output devices is a rather haphazard way to go about designing an amp. You might want to consider determining the load and desired wattage, then working backwards from that.
Note that the JC-3 front end will not work at such high voltages.
Grey
Hi Mr Grey,
I Believe the front end will not stand for that voltage thats why I ask whats more important topology or devices selection.
for higher voltage I hope we can use IRFD110 or even IRFP9640 or similar
Cascode is out off my list if the topology is in first list.
I'll save John the trouble: Topology.
Just specifying rails and output devices is a rather haphazard way to go about designing an amp. You might want to consider determining the load and desired wattage, then working backwards from that.
Note that the JC-3 front end will not work at such high voltages.
Grey
Hi Mr Grey,
I Believe the front end will not stand for that voltage thats why I ask whats more important topology or devices selection.
for higher voltage I hope we can use IRFD110 or even IRFP9640 or similar
Cascode is out off my list if the topology is in first list.
I use all three, but no tubes. It depends on the situation. The cheapest solution is bipolars. The optimum high voltage solution is complementary mosfets. The elegant solution is jfets. Tubes don't have complements, so they don't work in my topologies.
Mr Curl,
I Looking for +/- 90 volt the JFET like 2SK389 and the complement can not stand with this volt level, you also told so many times you dont like cascode. do you mind to explain the way out ?
Mr Curl,
I Looking for +/- 90 volt the JFET like 2SK389 and the complement can not stand with this volt level, you also told so many times you dont like cascode. do you mind to explain the way out ?
For a high voltage input stage, cascode is the best practical way to do it. All of my Parasound amps have a cascoded complementary differential fet input stage.
Thirty five years ago, when I was first designing this circuit topology with fets, I tried very hard to keep the 'stage count' as low as possible.
'Stage count' was counting each amplifying stage in series with the signal as "1" and every follower or cascode as "1/2". Even a differential pair was "1.5" because it was MORE than a simple single device.
In fact, some of my early amps did not have the differential input, to reduce 'stage count' but I found it too difficult to work with in production.
Therefore, there had to be a darn good reason to add a cascode, follower, etc.
Now let us look at the 'stage count' of a JC-3 power amp: differential pair (1.5) +second stage (1) +Darlington output (.5+.5) = 3.5. Ideally, it could be reduced to 2.5, with a non differential input stage and a mosfet output.
Now take one of Bob Cordell's designs and do a 'stage count'. Better yet a Halcro. See the difference?
Another example: The 'stage count' of a CTC 'Blowtorch preamp is 2.
For amateurs, I do not recommend +/- 90V designs. It is cheaper and easier to buy an amp that is already designed. It takes more knowledge than simply raising the voltage to make the amp safe and stable. The detail work and experience necessary is more than it is worth when you can just go out and buy the equivalent design, and then 'improve' it if you wish.
Thirty five years ago, when I was first designing this circuit topology with fets, I tried very hard to keep the 'stage count' as low as possible.
'Stage count' was counting each amplifying stage in series with the signal as "1" and every follower or cascode as "1/2". Even a differential pair was "1.5" because it was MORE than a simple single device.
In fact, some of my early amps did not have the differential input, to reduce 'stage count' but I found it too difficult to work with in production.
Therefore, there had to be a darn good reason to add a cascode, follower, etc.
Now let us look at the 'stage count' of a JC-3 power amp: differential pair (1.5) +second stage (1) +Darlington output (.5+.5) = 3.5. Ideally, it could be reduced to 2.5, with a non differential input stage and a mosfet output.
Now take one of Bob Cordell's designs and do a 'stage count'. Better yet a Halcro. See the difference?
Another example: The 'stage count' of a CTC 'Blowtorch preamp is 2.
For amateurs, I do not recommend +/- 90V designs. It is cheaper and easier to buy an amp that is already designed. It takes more knowledge than simply raising the voltage to make the amp safe and stable. The detail work and experience necessary is more than it is worth when you can just go out and buy the equivalent design, and then 'improve' it if you wish.
"Upgrading" an amp is something best done with a good game plan and decent electronic skills.
True story:
I set out to buy a dead Hafler DH-200, purely for the chassis, heat sinks, and transformer. Nailed one on ebay. It came in. Some nincompoop had tried to modify the thing, but left bad solder joints. I fixed the solder joints and the amp worked perfectly. I couldn't bring myself to gut a functional amp, so I put it on my AV system (wherein in sounded leagues better than my la-de-da Pioneer Elite...draw any conclusion you like) and bought another dead DH-200. Well, this one came in with some bizarre huge metal can film caps substituted for the input DC blocking caps. Punchline? The metal cans had long braided wire leads and were dangling loose inside the chassis, shorting things right and left. I removed the stupid metal cans and mirable dictu the amp came back to life. Like the first, I put it in my AV system and bought another "dead" DH-200.
I'll cut the story short--it took four of the bloody things before I found one that was actually dead. Frankly, I had no idea that there were that many incompetent "mods" out there. I have a far lower opinion of the ordinary fella's ability to modify a circuit than I used to.
Perhaps I should have politely suggested that they all come hang out here for a while and learn a thing or two.
Naaah...elseways, I'd have to pay more for DH-200s.
Grey
P.S.: My AV system sounds worlds better than it used to. I didn't set out to improve my AV system, but at least the story had a happy ending.
P.P.S.: As I hinted earlier--perhaps not strongly enough--if you run IRFP240s and IRFP9240s at +-90V rails you're going to have big problems.
True story:
I set out to buy a dead Hafler DH-200, purely for the chassis, heat sinks, and transformer. Nailed one on ebay. It came in. Some nincompoop had tried to modify the thing, but left bad solder joints. I fixed the solder joints and the amp worked perfectly. I couldn't bring myself to gut a functional amp, so I put it on my AV system (wherein in sounded leagues better than my la-de-da Pioneer Elite...draw any conclusion you like) and bought another dead DH-200. Well, this one came in with some bizarre huge metal can film caps substituted for the input DC blocking caps. Punchline? The metal cans had long braided wire leads and were dangling loose inside the chassis, shorting things right and left. I removed the stupid metal cans and mirable dictu the amp came back to life. Like the first, I put it in my AV system and bought another "dead" DH-200.
I'll cut the story short--it took four of the bloody things before I found one that was actually dead. Frankly, I had no idea that there were that many incompetent "mods" out there. I have a far lower opinion of the ordinary fella's ability to modify a circuit than I used to.
Perhaps I should have politely suggested that they all come hang out here for a while and learn a thing or two.
Naaah...elseways, I'd have to pay more for DH-200s.
Grey
P.S.: My AV system sounds worlds better than it used to. I didn't set out to improve my AV system, but at least the story had a happy ending.
P.P.S.: As I hinted earlier--perhaps not strongly enough--if you run IRFP240s and IRFP9240s at +-90V rails you're going to have big problems.
Hi, Rembulan,
Use TRANSNOVA topology (Hafler/Rockford). With this topology, the front end can be as low as +/-15V while the output stage can be as high as+/-110V. The K389 can be used without cascode at all and still gives full control of the output stage.
One topology becomes usefull if you want to use low voltage frontend (like K389 pairs) with high-voltage output stage.
Use TRANSNOVA topology (Hafler/Rockford). With this topology, the front end can be as low as +/-15V while the output stage can be as high as+/-110V. The K389 can be used without cascode at all and still gives full control of the output stage.
Thank Lumanauw,toquito, and other member who give sugestion.
I ask to Mr John about the topology and etc cause I like to understand in many many post john said he do not like to have a cascode input so i wondering what he uses in Parasound JC1, but he already answer it.
Honestly I still a have at least one question to Mr john about his design in somewhere at this web you said, you do not like to have a capacitor in the feedback, maybe I can understand that but in JC-3 he put the condensator in input pin that make me confuse. do you mind to explain why you dont want a capacitor in feedback but put it in input ?
I ask to Mr John about the topology and etc cause I like to understand in many many post john said he do not like to have a cascode input so i wondering what he uses in Parasound JC1, but he already answer it.
Honestly I still a have at least one question to Mr john about his design in somewhere at this web you said, you do not like to have a capacitor in the feedback, maybe I can understand that but in JC-3 he put the condensator in input pin that make me confuse. do you mind to explain why you dont want a capacitor in feedback but put it in input ?