That was what I thought when I looked at the B1 Buffer with the caps. I didn't know whether it made a difference or not and it wasn't mentioned so I left it.
Here it is with R5 in the new position. I also added two pads so that the jumpers that are in R3 and R4 can be transferred to where the legs of the pot were. That way everything can be easily connected once the trimmer is bypassed.
Here it is with R5 in the new position. I also added two pads so that the jumpers that are in R3 and R4 can be transferred to where the legs of the pot were. That way everything can be easily connected once the trimmer is bypassed.
Thank you Mike, this is so timely.
I have some extra jfets from an F5 build a while back, and now find myself needing a little buffer for a tube stage.
I have +-28V available in my project right now.
Do I need to cascode? I read the F5 turbo article several times and can’t really understand if it’s required. Seems a matter of faith.
In any event, since you’re accepting wishlist items, and if there’s enough space in your board, could you add pins for the cascode transistors?
I have some extra jfets from an F5 build a while back, and now find myself needing a little buffer for a tube stage.
I have +-28V available in my project right now.
Do I need to cascode? I read the F5 turbo article several times and can’t really understand if it’s required. Seems a matter of faith.
In any event, since you’re accepting wishlist items, and if there’s enough space in your board, could you add pins for the cascode transistors?
Just realized these boards are getting quite large. Part of the charm of the circuit is it’s smallness. I love the small pcb from post 320.
Maybe make a few versions just the gerbers, and people can just order whichever one they want. That way there’s no need for consensus.
I’m just a beggar, so take everything with a heaping of salt. Eternally grateful for the generosity of the community 🙏
Maybe make a few versions just the gerbers, and people can just order whichever one they want. That way there’s no need for consensus.
I’m just a beggar, so take everything with a heaping of salt. Eternally grateful for the generosity of the community 🙏
Last edited:
I think my next step for me is to get the no cap and the output cap boards and test them. If that goes well, I will likely try adding a feature or two as it is fun to play with these PCB board programs. I am relatively new to drawing up circuit boards so I need to test my work so far.
There are a few programs that you can try if you want to play around with adding a cascode circuit. EasyEDA has a good library for components so it is easy to get started with. Kicad is another. Eagle is another program but that costs money I believe.
I can try to figure out how to get the file project file to you if you like. Such a simple circuit like this is a good way to jump in.
I recall reading that the k170/j74 JFETS are good up to 30v. However, maybe in not such a simple circuit as this. That would be a question for the higher-ups.
There are a few programs that you can try if you want to play around with adding a cascode circuit. EasyEDA has a good library for components so it is easy to get started with. Kicad is another. Eagle is another program but that costs money I believe.
I can try to figure out how to get the file project file to you if you like. Such a simple circuit like this is a good way to jump in.
I recall reading that the k170/j74 JFETS are good up to 30v. However, maybe in not such a simple circuit as this. That would be a question for the higher-ups.
Just ensure keeping dissipation at bay. These JFETs are least noisy below 12 volts Vds. 28 is very much higher than spec for this circuit. However, noise in itself should not matter much in this particular application, there is no gain involved.I think my next step for me is to get the no cap and the output cap boards and test them. If that goes well, I will likely try adding a feature or two as it is fun to play with these PCB board programs. I am relatively new to drawing up circuit boards so I need to test my work so far.
There are a few programs that you can try if you want to play around with adding a cascode circuit. EasyEDA has a good library for components so it is easy to get started with. Kicad is another. Eagle is another program but that costs money I believe.
I can try to figure out how to get the file project file to you if you like. Such a simple circuit like this is a good way to jump in.
I recall reading that the k170/j74 JFETS are good up to 30v. However, maybe in not such a simple circuit as this. That would be a question for the higher-ups.
So if you target a max of 200mW per device, subtracting a bit for leverage and say 180mW instead, sure JFETs with an IDSS of 6-ish mA and you should be good.
I would not cascode this circuit. I have it up with only four parts per channel, including JFETs. Simplicity is a part of the gold of going complimentary with the B1, IMO.
I would use a lower rail voltage than 28 volts, but that’s me.
Ok thanks for the feedback. Maybe lowering voltage is easiest (and not that hard actually), if series regulator is not terrible.
One more question: On the F5 schematic, there is between a 600-2k2 resistor atop and below each JFET.
Would this help to drop a few volts or be advantageous?
Also, my JFETS are 7.7ma, purchased from Spencer in 2013. Not sure what his matching tolerance was.
One more question: On the F5 schematic, there is between a 600-2k2 resistor atop and below each JFET.
Would this help to drop a few volts or be advantageous?
Also, my JFETS are 7.7ma, purchased from Spencer in 2013. Not sure what his matching tolerance was.
Not really ideal, if you want the best from this circuit, but if you didn't want to purchase a new transformer then I would probably go RC then Regulator eg
500 Ohms, 15000uF, Regulator
That would drop the voltage about 8 Volts on each rail.
I don't think a new transformer is that expensive though. Just get a new one if you can.
Edit: If you used a mosfet cap multipler (instead of 3 pin regulator) you could drop more voltage so maybe that would be better.
500 Ohms, 15000uF, Regulator
That would drop the voltage about 8 Volts on each rail.
I don't think a new transformer is that expensive though. Just get a new one if you can.
Edit: If you used a mosfet cap multipler (instead of 3 pin regulator) you could drop more voltage so maybe that would be better.
Last edited:
A few questions. The original schematic has no R3 or R4 yet it has appeared in subsequent schematics here. I looked through the entire thread here and saw no discussion of what they are there for. Why are they there? What is the upside and downside of adding them and why was 1k chosen? Many thanks in advance for an elucidating reply.
The original R2 schematic has both R3 and R4. They are part of the power supply filter section, intended for a SMPS. And they are not 1k. I do not use these parts, since i have a regulated linear psu. I don’t think we are talking about the same circuit.
If you are referring to Mike’s revised circuit, the R3-4 1k’s are nominal values to symbolize the possibility of replacing the pot with resistors. Values should end up very much lower, and will take care of matching differences between JFETs normally taken care of by the pot plus apply some degeneration.
Last edited:
Andynor has the right answer. 1k is just the value automatically inserted by the PCB program. I should have probably put a "?" in there or something.
You'll likely end up with a value like 20-30 for each spot but you would have to measure the pot to figure it out. The idea is to get the trimmer out of the signal path.
You'll likely end up with a value like 20-30 for each spot but you would have to measure the pot to figure it out. The idea is to get the trimmer out of the signal path.
That makes more sense. The R3 and R4 I was referring to are in the schematic at post #345 and are not the voltage divider in the PS of the original design. But they have the same numbers. Thanks for the replies, gents.
Thanks Pico. I think the 500 Ohm V dropper sounds simple enough.
Unfortunately, I have no space in my chassis for any additional components or transformers. I have some 28V switchers in place already in place for a solid state power amp stage. Additionally, I have some B+ stuff happening for a tube pre stage. I just need the B1 because I'm using a 100K pot after the tube stage and before the power stage, and need the buffer to clean up some voltage loss because of the impedance mismatch.
As an alternative, does the attached caveman approach work ok, from a B1 input impedance and also performance perspective? I don't care about the wasted heat of the divider. Actually, the SMPS works better (resonant type) with a little extra load on it, and I have power to spare.
Haha. Pumpkin will be there some day. I am on the Iron Pre list too. I like pretexts
