Yes, what Adason said. If you are putting the preamp circuit in the power amp, only one cap is needed.
On the buffer, don’t build it yet. I’m trying other configs. Almost had it yesterday but it didn’t keep the sound over time.
On the buffer, don’t build it yet. I’m trying other configs. Almost had it yesterday but it didn’t keep the sound over time.
Thanks, Rahul.
I am making a MOUSER order - should I go ahead and get extra J113s? Or could that change?
Back to the caps - I am using B- instead of B+ with the TDV THF51 - you seemed to think that would make a difference a couple of years ago - sorry to belabor it - would that make any difference?
I am making a MOUSER order - should I go ahead and get extra J113s? Or could that change?
Back to the caps - I am using B- instead of B+ with the TDV THF51 - you seemed to think that would make a difference a couple of years ago - sorry to belabor it - would that make any difference?
Last edited:
Yeah, a few extra are nice to have on hand. I haven’t yet done the math on that, though it looks like we’ll need to select just the J113s, not the others. All the JFETs in this design are cheap.
The B- on the TdV won’t make a difference to the cap.
The B- on the TdV won’t make a difference to the cap.
Last edited:
So I am trying the lower voltage B+ again.
From what I can tell it seems the 10K pot is less than half of the value needed.
The resistor values between the "original" schematic and the new are quite different. I do not know how that would effect one's ability to set the bias. New - 100K and 47K - the old 47K and 1M respectively.
I used a voltage divider calculator and the result were close enough to the results I am seeing.
With about 7K to ground I would have 14 volts at the junction of the two fets with 130 volts B+. With 78 volts I cannot get anywhere close to 14 volts even with 10K.
Seems to me resistor attached to ground (the pot) needs to be, at least, 20K and closer to 22K.
Does this make sense?
Anyone tried the FET version yet with nominally 80 volts B+?
From what I can tell it seems the 10K pot is less than half of the value needed.
The resistor values between the "original" schematic and the new are quite different. I do not know how that would effect one's ability to set the bias. New - 100K and 47K - the old 47K and 1M respectively.
I used a voltage divider calculator and the result were close enough to the results I am seeing.
With about 7K to ground I would have 14 volts at the junction of the two fets with 130 volts B+. With 78 volts I cannot get anywhere close to 14 volts even with 10K.
Seems to me resistor attached to ground (the pot) needs to be, at least, 20K and closer to 22K.
Does this make sense?
Anyone tried the FET version yet with nominally 80 volts B+?
Using a 15K resistor added to a 5K trimmer - don't think I am using much of the trim pots resistance - I get 40 volts and 10 volts.
I did not change the other resistors. In my case adding the resistor was the easy path.
Now to listen to it.
This side has the reversed connection for J175.
I did not change the other resistors. In my case adding the resistor was the easy path.
Now to listen to it.
This side has the reversed connection for J175.
It shouldn’t change much. The CCS looks like a very high value resistor to the power supply, so 130V or 80V shouldn’t sound different. But you never know 😉
I was worried it would not sound as good!
Not the case. Getting rid of the redundant capacitor does make a difference. The bass has much more character.
From what I heard in a few hours I think placing the line stage at the amp, allowing one less high pass filter, makes a difference.
So far, it seems a substantial difference.
Whether the buffer would negate this remains to be seen.
Not the case. Getting rid of the redundant capacitor does make a difference. The bass has much more character.
From what I heard in a few hours I think placing the line stage at the amp, allowing one less high pass filter, makes a difference.
So far, it seems a substantial difference.
Whether the buffer would negate this remains to be seen.
Converted to low voltage which allows me to make the supply choke input. Thankfully that worked with my transformer. About eight volts across the regulator.
All went fine.
Then I tried substituting the J113 in the current source and, as before, I cannot get it to work.
Is there something to doing this I am missing?
Is there a sonic difference doing this? I figure there must be just needed to be motivated. I have ordered additional 2535s.
I suspect the problem is trying to replace the 2535 - in my hands whenever I do anything with them they go away.
On the stage I first tried this I should have replaced the 2535 at the same time.
There is nothing lost with the lower voltage. I continue to think getting rid of the extra capacitor is the big difference.
Thanks, Rahul
All went fine.
Then I tried substituting the J113 in the current source and, as before, I cannot get it to work.
Is there something to doing this I am missing?
Is there a sonic difference doing this? I figure there must be just needed to be motivated. I have ordered additional 2535s.
I suspect the problem is trying to replace the 2535 - in my hands whenever I do anything with them they go away.
On the stage I first tried this I should have replaced the 2535 at the same time.
There is nothing lost with the lower voltage. I continue to think getting rid of the extra capacitor is the big difference.
Thanks, Rahul
Have you noticed improvements addind the choke when you already have the CCS?
Can you describe what happens? Are you using the J113 as CCS not cascoded?
Is this on lows only as you previously wrote, or on the full range of frequencies?
Last edited:
I started out with CLCRC so I cannot compare to the standard supply.
I do think there is some improvement with the choke input but that is probably because "I KNOW" it is better - otherwise known as suspect assumptions. I had no choice but to go choke input to get to the proper voltage. I am lucky it worked out. I bought two fancy TOROIDY transformers and they had to be used!
I do think there is something gained by losing an output capacitor - no question the bass is perceived as more dynamic which makes sense to me with one less LP filter in the chain. Not sure I hear anything similar with the tweeter channel. I was using two DUELUND caps and now they are in parallel at the amplifier.
I have not listened to this long enough to make an assessment AND I am only listening to one channel.
I am using the J113 as Rahul has it drawn.
Every time I have tried to change anything - other than the current set resistor - with the current source I lose the 2535 almost every time. These are TO220 packaged. I figure there is something I should be doing. Or something I am doing wrong when desoldering. Is there a best way to work with these?
Luckily they are not expensive but it is frustrating when you do not have any.
I have put some on order.
Thanks for your interest in my (stupid) problems!
Take care,
I do think there is some improvement with the choke input but that is probably because "I KNOW" it is better - otherwise known as suspect assumptions. I had no choice but to go choke input to get to the proper voltage. I am lucky it worked out. I bought two fancy TOROIDY transformers and they had to be used!
I do think there is something gained by losing an output capacitor - no question the bass is perceived as more dynamic which makes sense to me with one less LP filter in the chain. Not sure I hear anything similar with the tweeter channel. I was using two DUELUND caps and now they are in parallel at the amplifier.
I have not listened to this long enough to make an assessment AND I am only listening to one channel.
I am using the J113 as Rahul has it drawn.
Every time I have tried to change anything - other than the current set resistor - with the current source I lose the 2535 almost every time. These are TO220 packaged. I figure there is something I should be doing. Or something I am doing wrong when desoldering. Is there a best way to work with these?
Luckily they are not expensive but it is frustrating when you do not have any.
I have put some on order.
Thanks for your interest in my (stupid) problems!
Take care,
To figure out the issue you can use a 24V PSU and a 1kOhm 2W resistor. Connection is: 24V > CCS > resistor > gnd.
This way you reduce the dissipated power on the 2535 to values that do not need heatsink, and at the same time you easily read the current of the CCS reading the volts across the resistor: 1V = 1 mA. On this project you need to stay around 14-16 mA, so you still have 8-10V across the CCS.
Once you have figured out the resistor you need to target the right current, you can keep it as a starting point on the full circuit at full voltage.
Last edited:
I was used to say that those are the frequencies where the chest is your ear and the the hem of the trousers is your vu-meter. 🙂
I have no trouble getting the CS current set or getting the dual DN2535 to work - when I have inserted the J113 the thing does not work. No voltage across the CS. Before when I was using 130 volts B+ that was not surprising but now with 78 volts it seems like it should work without any drama. Current is set for 15 mA on the three that are working with two DM2535. It is my intention to set the J113/DN2540 to the same level.
I use four line stages - I can always get them to work well with two DN2535s - when I attempt to use the J113 I can never get them to work. I know I am repeating myself - wanted to make it clear where I am running into trouble.
The "bass channel" in my system is from 40 Hz to 400 Hz. Below that are subwoofers. I could get them the shake the floor and rattle my trousers but I do not find that sounds good. Overwhelms the room. I do not listen to EDM or rap which might make a difference. Rare that a recording moves that much air. All cabinets are sealed.
Has anybody, other than Rahul, used the DN2540/J113 current source?
I use four line stages - I can always get them to work well with two DN2535s - when I attempt to use the J113 I can never get them to work. I know I am repeating myself - wanted to make it clear where I am running into trouble.
The "bass channel" in my system is from 40 Hz to 400 Hz. Below that are subwoofers. I could get them the shake the floor and rattle my trousers but I do not find that sounds good. Overwhelms the room. I do not listen to EDM or rap which might make a difference. Rare that a recording moves that much air. All cabinets are sealed.
Has anybody, other than Rahul, used the DN2540/J113 current source?
I am not having trouble getting the proper current - would think the same resistor I am using with the two DN2535s would get me in the general vicinity (95R). It is simply that nothing flows through the current source!
Tomorrow I get additional 2535s - I will try with unused components and see what happens. I hope it was something I did, which is, of course, the most likely thing.
At first I tried using these with the 130 volts supply and that was never going to work. I did not realize it at the time. I was using an early schematic and the voltages were high on the schematic. I thought with the lower voltage it would work. So I simply replaced the lower 2535 with a J113 - I have found that the 2535s IN MY HANDS when I attempt to re-solder them after installation are fragile. Even with the two 2535 setup - if something went wrong I had to replace them both or one would ruin the other. I have no idea which one ruined the other, unfortunately.
This is why I wondered if there was something I should do when de-soldering these devices. If there is a protocol of actions for desoldering them. At this point I have given up on re-using them.
I appreciate the help - I do not think I have asked my questions clearly enough to get what I am wanting.
Tomorrow I get additional 2535s - I will try with unused components and see what happens. I hope it was something I did, which is, of course, the most likely thing.
At first I tried using these with the 130 volts supply and that was never going to work. I did not realize it at the time. I was using an early schematic and the voltages were high on the schematic. I thought with the lower voltage it would work. So I simply replaced the lower 2535 with a J113 - I have found that the 2535s IN MY HANDS when I attempt to re-solder them after installation are fragile. Even with the two 2535 setup - if something went wrong I had to replace them both or one would ruin the other. I have no idea which one ruined the other, unfortunately.
This is why I wondered if there was something I should do when de-soldering these devices. If there is a protocol of actions for desoldering them. At this point I have given up on re-using them.
I appreciate the help - I do not think I have asked my questions clearly enough to get what I am wanting.
You can even test the CCS using a 9V battery. Adjust the resistance until you hit the value and then transfer the same parts to the main circuit. Two DN2535s should work just as well. The second device raises the CCS impedance further and over a wider bandwidth but it is a marginal improvement, i.e., even with one device the CCS impedance is pretty high.