PCB Maker
Hiii Mike
1
Theres a cheaper PCB maker herein Jakarta
for 1 PCB will take cost US$ 10 but must run
for 100 pcs.
Total order for JC Amp more than 50 pcs
but must use group buy
But i dont know, are we look for the trustable
or we just look for the cheaper ?
2
we must consider about shipping cost
US$ 6-10 for doule pcbs
3
due to 2sk389 and 2sj109 accident maybe
we must find another good supplier
4
Do you have the PSU design for this amp ?
Thanks Mike
Hiii Mike
1
Theres a cheaper PCB maker herein Jakarta
for 1 PCB will take cost US$ 10 but must run
for 100 pcs.
Total order for JC Amp more than 50 pcs
but must use group buy
But i dont know, are we look for the trustable
or we just look for the cheaper ?
2
we must consider about shipping cost
US$ 6-10 for doule pcbs
3
due to 2sk389 and 2sj109 accident maybe
we must find another good supplier
4
Do you have the PSU design for this amp ?
Thanks Mike
Mike,
I think you'll find that 500 Ohms is a pretty high value for R25 unless you're intending to run the front end really, really lean. Assuming that you're going to be using BL parts for the front end, a fixed resistor somewhere between 10 and 50 Ohms would probably suffice depending on the Idss of the actual parts. If you're using a pot, something along the lines of 20, 50, or 100 Ohms will do nicely.
Won't make any difference in your layout, though.
You might want to consider an option on your PCB to run the front end on a separate, regulated rail. John's self-biasing complementary differential has only so-so rejection of rail nasties. I'm using separate, regulated rails for both input and output. My input rails are higher than the output rails--something that you won't have to do.
I wish you luck on your design. I'm in the process of building a second channel so as to be able to listen to my version. Hope to have two channels running by the end of next week or perhaps the week following. The PCBs are ready; I need to stop and do metal work so as to have heatsinking for the output stages. Either that or pay someone else to do it and hope they get it right. Neither option appeals to me. Ugh.
Grey
I think you'll find that 500 Ohms is a pretty high value for R25 unless you're intending to run the front end really, really lean. Assuming that you're going to be using BL parts for the front end, a fixed resistor somewhere between 10 and 50 Ohms would probably suffice depending on the Idss of the actual parts. If you're using a pot, something along the lines of 20, 50, or 100 Ohms will do nicely.
Won't make any difference in your layout, though.
You might want to consider an option on your PCB to run the front end on a separate, regulated rail. John's self-biasing complementary differential has only so-so rejection of rail nasties. I'm using separate, regulated rails for both input and output. My input rails are higher than the output rails--something that you won't have to do.
I wish you luck on your design. I'm in the process of building a second channel so as to be able to listen to my version. Hope to have two channels running by the end of next week or perhaps the week following. The PCBs are ready; I need to stop and do metal work so as to have heatsinking for the output stages. Either that or pay someone else to do it and hope they get it right. Neither option appeals to me. Ugh.
Grey
Split at R26 & R31? Yeah, looks good.
65V front end and 55V outputs? Uh...perhaps I'd better come up there and help you install a water-cooled system like I have here.
On a more serious note, I'm using passive heatsinks for the single-ended 25W version of my circuit, but will probably go water-cooled for the bridged 100W. I don't have access to enough monster heatsinks to handle the heat, otherwise.
Are you going to try yours bridged?
Grey
65V front end and 55V outputs? Uh...perhaps I'd better come up there and help you install a water-cooled system like I have here.
On a more serious note, I'm using passive heatsinks for the single-ended 25W version of my circuit, but will probably go water-cooled for the bridged 100W. I don't have access to enough monster heatsinks to handle the heat, otherwise.
Are you going to try yours bridged?
Grey
They're predicting rain/freezing rain/sleet/snow for tonight. If it happens at all (unlikely) it'll be the first frozen stuff we've had in a couple of years. I'm not holding my breath--no accumulation is expected. You can send me some of yours if you're tired of it. My wife's birthday is Monday and she'd love to see snow.
At this point it's around 55 degrees or so down in the Dungeon. Any and all heat is welcome. Class A is ideal. I could run a duct or two from the heat pump into my listening room but I so rarely get a chance to listen that it's not worth the expense to heat the room on the off chance that I might get in there once in a blue moon.
One of the things I've done with my front end board is leave space for a dedicated series high pass capacitor. That way I can toss in a cap and have a tweeter amp with only the addition of a level control.
The 100W version will get a tryout on my midranges somewhat later in the process.
Grey
At this point it's around 55 degrees or so down in the Dungeon. Any and all heat is welcome. Class A is ideal. I could run a duct or two from the heat pump into my listening room but I so rarely get a chance to listen that it's not worth the expense to heat the room on the off chance that I might get in there once in a blue moon.
One of the things I've done with my front end board is leave space for a dedicated series high pass capacitor. That way I can toss in a cap and have a tweeter amp with only the addition of a level control.
The 100W version will get a tryout on my midranges somewhat later in the process.
Grey
Mine's stable as a rock. I don't see any reason it should be any worse than any other circuit regarding stability. Actually, mine's probably more stable than most because it's inherently rather low gain and there's no feedback, hence no phase-shifted stuff coming in the back door.
Once Mike builds a prototype, we'll see how he fares.
(C'mon cheering squad...Go Mike! Go Mike! Go Mike!)
Grey
Once Mike builds a prototype, we'll see how he fares.
(C'mon cheering squad...Go Mike! Go Mike! Go Mike!)
Grey
Just in case instability occurs
Altough it is certainly adding distorsion I think that the solution posted back here might take care of this problem... plus a list of several others: eliminates the input coupling cap, allows the use of a potentiometer at the input, isolates the influence of preamp output Z on gain and of course should also work as servo. As is the schematic is not updated with John's usefull recomandations.
Unfortunately as far as I'm concerned, after the dualFET GB mishap, I had to postphone this project. Reason: I can't use the singles because the supplies I had in mind are a bit too high: regulated +/- 27 for input and driver, +/-25 unregulated for output stage.
dimitri said:
Altough it is certainly adding distorsion I think that the solution posted back here might take care of this problem... plus a list of several others: eliminates the input coupling cap, allows the use of a potentiometer at the input, isolates the influence of preamp output Z on gain and of course should also work as servo. As is the schematic is not updated with John's usefull recomandations.
Unfortunately as far as I'm concerned, after the dualFET GB mishap, I had to postphone this project. Reason: I can't use the singles because the supplies I had in mind are a bit too high: regulated +/- 27 for input and driver, +/-25 unregulated for output stage.
My circuit is DC coupled, hence no caps in the signal path unless you want to use one as a high pass. Each input is connected to ground (at present by 100k) setting the Zin and making things behave.
To the extent that you want to more faithfully reproduce John's original circuit I don't believe too much violence would be done to the concept if you were to use a large resistor to ground at the input. Yes, it would change the feedback somewhat, but it would ground reference the thing.
Running feedback back to the input isn't the most common way to do things, but it works. (I do it in my main tube amps.) In this case, the cap allows the DC offset pot to control the circuit independent of anything else going on out in the world. If you're brave and happen to have a tube preamp (hence already cap coupled unless you're using something like the Atma-Sphere) or a solid state preamp that happens to be cap-coupled (like some of Nelson's designs), then you can begin by deleting the input cap on the JC-3. If you're certain that there's no DC at the output of your direct-coupled preamp, you can go one step further and see how things go without the cap.
Me? I'd stick in a high value ground referencing resistor, anyway, just to be sure. 1M, 2.21M, 3.32M, something big so as not to load the feedback any more than necessary. At these values, even the minor Gate currents in JFETs can begin to cause offsets, but you should be able to adjust things with the pot.
Grey
To the extent that you want to more faithfully reproduce John's original circuit I don't believe too much violence would be done to the concept if you were to use a large resistor to ground at the input. Yes, it would change the feedback somewhat, but it would ground reference the thing.
Running feedback back to the input isn't the most common way to do things, but it works. (I do it in my main tube amps.) In this case, the cap allows the DC offset pot to control the circuit independent of anything else going on out in the world. If you're brave and happen to have a tube preamp (hence already cap coupled unless you're using something like the Atma-Sphere) or a solid state preamp that happens to be cap-coupled (like some of Nelson's designs), then you can begin by deleting the input cap on the JC-3. If you're certain that there's no DC at the output of your direct-coupled preamp, you can go one step further and see how things go without the cap.
Me? I'd stick in a high value ground referencing resistor, anyway, just to be sure. 1M, 2.21M, 3.32M, something big so as not to load the feedback any more than necessary. At these values, even the minor Gate currents in JFETs can begin to cause offsets, but you should be able to adjust things with the pot.
Grey