I have hade the parts and PCBs for this (the 10 x BF862 version) since 2017 but never assembled it. I finally found some time now and started matching the jfets. Getting Idss between 23 and 29mA I quickly realized these much be fake. I bought 120 of them from Farnell and did not expect them to have anything other that the real thing. Anyhow, since I had started I decided to continue so I matched them all and assembled the boards. I fired it up yesterday and it sounds amazing! I only have a Schiit Magni 2 to compare with but that was an easy win. I wonder how good it would have sounded with genuine BF862s.
The datasheet for BF862 says max Idss 25mA and since most of mine were higher than that I assumed they are fake. Well, the most important thing is that it sounds good.
I do not feel an urge to do matching of any more SOT23 devices now so I will just leave it as is and focus on the next project, probably a F4 amp.
Thank you xrk971 for your great contribution to this forum!
I do not feel an urge to do matching of any more SOT23 devices now so I will just leave it as is and focus on the next project, probably a F4 amp.
Thank you xrk971 for your great contribution to this forum!
I've assembled mine with tightly matched jfets and Wima MKP and Elna Silmic II's, takman metal film resistors...shortly the best parts I could get within reason.
I haven't tested it yet though as my chronic pain decide what gets done and when.
My comparison will be a Lehmann BCL clone(with the best parts within reason) and an upgraded version of the AD797/BUF634 headamp with a C-L-C filter amongst other things I've changed. I also have some SSB01 populated to test as replacement for the BUF634.
Power supply is, for the BF862 amp, the super simple cap-MX, and for the other two either parallel LT3042 supplies, TPS7A4700/TPS7A3301 or the the version 3 of P/A's super Buffer.
It's nice to hear that there is some good feedback on the (many) BF862 headamp as it is a PITA to match those buggers.
I haven't tested it yet though as my chronic pain decide what gets done and when.
My comparison will be a Lehmann BCL clone(with the best parts within reason) and an upgraded version of the AD797/BUF634 headamp with a C-L-C filter amongst other things I've changed. I also have some SSB01 populated to test as replacement for the BUF634.
Power supply is, for the BF862 amp, the super simple cap-MX, and for the other two either parallel LT3042 supplies, TPS7A4700/TPS7A3301 or the the version 3 of P/A's super Buffer.
It's nice to hear that there is some good feedback on the (many) BF862 headamp as it is a PITA to match those buggers.
Mayday, great to see you are still around. Sorry to hear that you are still battling with back pain. Looking forward to your listening impressions. I am a great fan of the (unfortunately now obsolete) BF862 little bugger jfets – I have them in my PCA headamp, Pass B1 buffer and PCA based Aksa Lender preamp. Hope your amp sounds great too!
It will be very interesting to hear your opinion Mayday when you get the chance to test it. You have better reference amps than I do.
I used Silmic II, Nichicon UFG and Dale CMF resistors. Have so far only used battery power but will soon test with a SMPS with cap mx, mostly because I am curious to test that setup after a lot of different linear supplies to my projects during the years.
I used Silmic II, Nichicon UFG and Dale CMF resistors. Have so far only used battery power but will soon test with a SMPS with cap mx, mostly because I am curious to test that setup after a lot of different linear supplies to my projects during the years.
Of course I meant the P-A Sjostrom super regulator V3 as supply, not the buffer.
Thanks
It's chronic and degenerative, so not much to do except not aggravate it and the(inadequate) pain relief I get.
I'm a big fan of the 2SK170, and from what I've read, there are many who consider the BF862 better sounding.
I'm very familiar with both the BCL clone and the AD797/BUF634 headamps, so I should be able to give a reasonable impression of the "many jfet" compared to those two at least.
I too have some BF862, when I ordered mine I asked the retailer to make sure I got all from the same roll.
I have maybe 60 or 70 left. I'm going to keep those as well.
I also have some, very tightly matched, 2SK170BL that I got from Spencer and I am keeping those as well.
I got very nice sound using 4 of those as replacement for the opamp in the BCL clone (with two trimpots and a couple of resistors soldered to a DIP8 socket.
I have soo many parts and PCB's at this point that I forget. I use InvenTree database now to keep it easier to remember what I have.
I have maybe 60 or 70 left. I'm going to keep those as well.
I also have some, very tightly matched, 2SK170BL that I got from Spencer and I am keeping those as well.
I got very nice sound using 4 of those as replacement for the opamp in the BCL clone (with two trimpots and a couple of resistors soldered to a DIP8 socket.
I have soo many parts and PCB's at this point that I forget. I use InvenTree database now to keep it easier to remember what I have.
The 2sk209 comes in GR abs BL grades. Either way, you will need to run the simulation to see what the best bias current is. The resistor setpoints will be different than BF862.
Typically use the bias current set to about 70% of Idss. So for 20mA Idss you can run each at about 14mA. 14mA x 5 pairs is 70mA overall bias current.
So then tweak setting in sim to get 14mA across ea JFET.
Typically use the bias current set to about 70% of Idss. So for 20mA Idss you can run each at about 14mA. 14mA x 5 pairs is 70mA overall bias current.
So then tweak setting in sim to get 14mA across ea JFET.
I have a roll (100) of BF862 bought from Aliexpress about 3 years ago and they worked fine in the class-A headamp and have the 2AW marking. I just got another roll from the same seller this week also with the 2AW marking and finally got to measure a single one from each roll. I set my power supply to 12V and 50mA limit. Measured across a 100R resistor, the new roll jfet was 23mA and the old roll was 19mA. I only measured one from each but it looks like it's still possible to get real ones or at least ones that still fall within the listed 10-25mA range.
X, this is a continuation of the discussion we've had on the Fistful of Solder thread about the BF862/IRF610 schematic that isn't working for me. I'll post a photo of the pcb that I created when I have a chance tomorrow. Here is the BF862/IRF610 schematic sim that I used for laying out the board.
Ok, I see it’s the basic PCA/DCA. I have literally built over 100 of these. They always worked and when they didn’t , it was either a cold solder joint or a zapped MOSFET. There’s not much to go wrong.
if you are not getting 15v at the drain of JFET then either R3, R4, or R5 are incorrect value, open circuit, or shorted.
if you are not getting 15v at the drain of JFET then either R3, R4, or R5 are incorrect value, open circuit, or shorted.
Thanks X. I'll have an update tomorrow night with pics as the kids came home today with homework and I'm busy helping my daughter get through math. I'll recheck for bad solder joints and reflow them to be sure. The board is laid out for thin film resistors except for the 4 x 270R resistors. The drain at the jfet is 13.9V, so off by a volt. Not sure if there is some critical characteristic that a thin film has that is not suitable for this circuit.