Since the output voltage is only 11V max, I've decided to use this as the output coupling cap in my amp:
Panasonic OS-CON 16SEPF1000M
16V, 1000uF, 10x13mm.
Going to lay it on its side the way I drew it up in Post #148.
Say, X, have you tried listening to the Pocket Class A using just OS-CON output caps without the WIMA film bypass? I came across this post on Gearslutz wherein one guy sings the praises of OS-CONs and says they give the best treble transparency with no bypass at all.
https://www.gearslutz.com/board/8949963-post16.html
Panasonic OS-CON 16SEPF1000M
16V, 1000uF, 10x13mm.
Going to lay it on its side the way I drew it up in Post #148.
Say, X, have you tried listening to the Pocket Class A using just OS-CON output caps without the WIMA film bypass? I came across this post on Gearslutz wherein one guy sings the praises of OS-CONs and says they give the best treble transparency with no bypass at all.
https://www.gearslutz.com/board/8949963-post16.html
Last edited:
I can't say that I have tried without. However, it's easy enough to try on the next one I build to leave it off first. My hunch is that it's tough to beat a good Wima film cap on treble with an electrolytic. The OSCON is there for the bass - low ESR needed. When using plain elco's for full range and treble, many people like Nichicon Gold, or Elna Silmic. I have nver tried those.
It's interesting how on Gearslutz post the guy says it's 2-3% more noise! I somehow doubt that to be possible because my noise floor measures -123dB.
Last edited:
Snazzy. The way you turned one 90 degrees gives a bit more breathing room than the way I had imagined.
I am ready to order all my parts but am hung up on the LED, lol. Yes, I am a noob. Digikey has 650 different 3mm Red LEDs, and I have no idea which ones will work in this circuit. Can somebody link a cheap one that'll work, from either Digikey or Mouser? Or just tell me what parameters to look for. I don't like them too bright. Also, do I need to get a resistor to go in front of it, or is that already part of the BOM?
I am ready to order all my parts but am hung up on the LED, lol. Yes, I am a noob. Digikey has 650 different 3mm Red LEDs, and I have no idea which ones will work in this circuit. Can somebody link a cheap one that'll work, from either Digikey or Mouser? Or just tell me what parameters to look for. I don't like them too bright. Also, do I need to get a resistor to go in front of it, or is that already part of the BOM?
Any 3mm LED (green, red, yellow, orange) can work. You can also get the one Agdr uses on the CMOY with a plastic mount, but unfortunately the legs are rotated 90deg. It can be made to work though by bending the pins and covering one with shrink tube,
Mouser Part 638-MV5464.MP4B (Everlight green LED)
The PCB already has R23 which I recommend to use 12k for the LED.
Mouser Part 638-MV5464.MP4B (Everlight green LED)
The PCB already has R23 which I recommend to use 12k for the LED.
Please add your name and number of boards.
Turbon - 1 board
Tryptamin - 2 bords
Schlomoff - 5 boards
It fits just fine.
Btw, in measuring my entire collection of BF862's in the special jig to simulate the circuit, and then using the circuit itself as a jig (with bare MSOFET pads) to measure the ZVN4306's (I record the bias current in volts across the four source resistors), I am getting some phenomenal results (I never bothered to match my own MOSFETs for this amp before). I am getting a match of bias current to better than 0.25% between left and right channels now! So just want to let you know if you order the matched FETs, they will work really well. The way I measured them in-situ in the actual circuit that they will be used in is really the best.
Attachments
Last edited:
Matching FETs
To match FETs do not solder JFETs or power MOSFETs up at first - use the actual circuit to test and match (if you are intererested in doing do for perfect channel balance).
This is kind of tricky but easy once you get the idea: use the actual circuit as the FET testing jig at actual conditions. There is no better way to match FETs than actual circuit it will be used in.
Take tiny wires (30 ga) and solder across test points for measuring V across R3. Take a plastic ESD safe tweezer and press the little JFET BF862 onto the pads of a live circuit while monitoring the DVM across the 1k resisitor (R3A or R3B) This gives you Idss for the exact operating condition. Do this for the about 12 JFETs you have and find two that are close - should get readings like 3.5v to 4.0v etc. Match would be 3.85v and 3.88v or more variance is Ok even.
Once you have matched BF862's install them with solder. Now repeat same operation but press ZVN4306 in place while monitoring voltage across the four parallel resistors (R7A or R7B). Depending on how strict you are, you may need a dozen. But I find that I can match 3 pairs for 12 - within very acceptable tolerances. Again, there is not a better way to match as it is running live at the actual bias conditions.
Once you have your matched pair of ZVN4306's, solder them in place. Now test your bias current by measuring voltage drop across R7. The left and right channels should basically be identical within less than 1%.
The next step is to adjust the source resistor of the JFET for optimal bias current. Measure the voltage across R7 and determine the bias current using I = V/R. It it is a lot bigger than 55mA, reduce by one value and if less DNA 50mA increase the resistance by one value (33R to 47Rbfie example). When bias is between 50mA and 55mA with freshly charged batteries, the bias is set correctly. It will work fine with default 33R value but optimal odd about 59mA.
To match FETs do not solder JFETs or power MOSFETs up at first - use the actual circuit to test and match (if you are intererested in doing do for perfect channel balance).
This is kind of tricky but easy once you get the idea: use the actual circuit as the FET testing jig at actual conditions. There is no better way to match FETs than actual circuit it will be used in.
Take tiny wires (30 ga) and solder across test points for measuring V across R3. Take a plastic ESD safe tweezer and press the little JFET BF862 onto the pads of a live circuit while monitoring the DVM across the 1k resisitor (R3A or R3B) This gives you Idss for the exact operating condition. Do this for the about 12 JFETs you have and find two that are close - should get readings like 3.5v to 4.0v etc. Match would be 3.85v and 3.88v or more variance is Ok even.
Once you have matched BF862's install them with solder. Now repeat same operation but press ZVN4306 in place while monitoring voltage across the four parallel resistors (R7A or R7B). Depending on how strict you are, you may need a dozen. But I find that I can match 3 pairs for 12 - within very acceptable tolerances. Again, there is not a better way to match as it is running live at the actual bias conditions.
Once you have your matched pair of ZVN4306's, solder them in place. Now test your bias current by measuring voltage drop across R7. The left and right channels should basically be identical within less than 1%.
The next step is to adjust the source resistor of the JFET for optimal bias current. Measure the voltage across R7 and determine the bias current using I = V/R. It it is a lot bigger than 55mA, reduce by one value and if less DNA 50mA increase the resistance by one value (33R to 47Rbfie example). When bias is between 50mA and 55mA with freshly charged batteries, the bias is set correctly. It will work fine with default 33R value but optimal odd about 59mA.
pcgab - 2 boards
zman01 - 1 board
stellarelephant - 2 boards
samoloko - 1 pcb
Morde - 2 board
cyteen - 2 boards
funch - 2 boards
syyma - 2 boards - SHIPMENT RECEIVED
av-trouvaille - 2 boards - SHIPMENT RECEIVED
Mituisho - 2 boards
bogde - 2 boards
faltinsen - 2 boards
vanofmonks - 1 board
M_Balou - 2 boards
dwpeterson - 2 boards
agdr - 1 board
walangalam - 2 boards
gwrskien - 2 boards
redarc12- 2 boards
dw1narso - 2 boards
zimmer64 - 2 boards
shambala - 2 boards
HaroldHill - 2 boards
gig - 2 boards
hui13 - 2 boards
pranavkr -1 board
JimS- 1 board
Kumori - 1 board
tubediy - 1 board
AKSA - 1 board
zman01 - 1 board
stellarelephant - 2 boards
samoloko - 1 pcb
Morde - 2 board
cyteen - 2 boards
funch - 2 boards
syyma - 2 boards - SHIPMENT RECEIVED
av-trouvaille - 2 boards - SHIPMENT RECEIVED
Mituisho - 2 boards
bogde - 2 boards
faltinsen - 2 boards
vanofmonks - 1 board
M_Balou - 2 boards
dwpeterson - 2 boards
agdr - 1 board
walangalam - 2 boards
gwrskien - 2 boards
redarc12- 2 boards
dw1narso - 2 boards
zimmer64 - 2 boards
shambala - 2 boards
HaroldHill - 2 boards
gig - 2 boards
hui13 - 2 boards
pranavkr -1 board
JimS- 1 board
Kumori - 1 board
tubediy - 1 board
AKSA - 1 board
pcgab - 2 boards
zman01 - 1 board
stellarelephant - 2 boards
samoloko - 1 pcb
Morde - 2 board
cyteen - 2 boards
funch - 2 boards
syyma - 2 boards - SHIPMENT RECEIVED
av-trouvaille - 2 boards - SHIPMENT RECEIVED
Mituisho - 2 boards
bogde - 2 boards
faltinsen - 2 boards
vanofmonks - 1 board
M_Balou - 2 boards
dwpeterson - 2 boards
agdr - 1 board
walangalam - 2 boards
gwrskien - 2 boards shipment received
redarc12- 2 boards
dw1narso - 2 boards
zimmer64 - 2 boards
shambala - 2 boards
HaroldHill - 2 boards
gig - 2 boards
hui13 - 2 boards
pranavkr -1 board
JimS- 1 board
Kumori - 1 board
tubediy - 1 board
AKSA - 1 board
zman01 - 1 board
stellarelephant - 2 boards
samoloko - 1 pcb
Morde - 2 board
cyteen - 2 boards
funch - 2 boards
syyma - 2 boards - SHIPMENT RECEIVED
av-trouvaille - 2 boards - SHIPMENT RECEIVED
Mituisho - 2 boards
bogde - 2 boards
faltinsen - 2 boards
vanofmonks - 1 board
M_Balou - 2 boards
dwpeterson - 2 boards
agdr - 1 board
walangalam - 2 boards
gwrskien - 2 boards shipment received
redarc12- 2 boards
dw1narso - 2 boards
zimmer64 - 2 boards
shambala - 2 boards
HaroldHill - 2 boards
gig - 2 boards
hui13 - 2 boards
pranavkr -1 board
JimS- 1 board
Kumori - 1 board
tubediy - 1 board
AKSA - 1 board