post8700
I don't know if you have the old spreadsheet, or the corrected version.
The old excel gave some wrong answers.
I don't know if you have the old spreadsheet, or the corrected version.
The old excel gave some wrong answers.
The early PCB just before I discovered the Thread had a summer/winter setting, using two different bias setting Variable resistors and a switch that added the second resistor in parallel to the first to increase the bias voltage and thus the bias current for a winter setting (or all year in the UK & Scandinavia).
The newer PCB which was designed to fit inside the Krell original if one were ever needed had a single "all year round" bias setting.
John is right, 185mVre on 3pr 0.22Re is indeed 50W of ClassA into 4r0.
One needs to change the .xls to an accepted format for the Forum.
just adding .txt to the full excell export string usually works.
Then when we download we remove the .txt and save the save the file to be opened later with excell or another spreadsheet reader.
But I used Libre Office recently and I could not open my own download. When I used Open Office, I found that the method worked.
Last edited:
Hey Alan, Glad to have stumbled here again. I have moved to the Middle East for a few years and here, there's little chance of me having a workshop to indulge... maybe in another decade, I will get back 😉
Wonderfull Guys thanks a ton
i will love to have the excel sheet, cause im planning to run it at 8 ohm only 3 pairs of A1943 and C5200.
so i can have the right values
my question was about the board silkscreen where i have 3 holes for a pot where its labeled low bias an another one labeled high bias, im using jan`s boards the ones from the pdf and gerber files.
the DC bias its the very firs pot arround the input stage?
K amps couldyou please send me the sheet via email or pct message please¡¡
Regards guys and thanks a ton ill post pics later
I used Pinkmouse's boards. I saw Jan's boards and now remember the dual post for bias.
Some advice:
Build it, checking your work as you go.. start it up one channel at a time, with a variac... at a certain voltage the amp will wake up... two voltmeters are needed - one for bias, and one on the speaker outputs showing DC offset. Keep Bias low and slowly ramp up the variac until your at mains voltage. No smoke is good.
Slowly dial up the bias, wait for the sinks to get nice and hot, trim offset to zero, wait, go up another 20 mv on bias, again and again, until 125 - 130F on the sinks - its an iterative process, all the while trimming your DC offset while raising bias. You will notice ramping up the bias will disrupt your DC offset and vice versa. Do the other channel same way. Close the case all up and check bias and DC offset one more time, as the two channels enclosed will affect final bias and DC Offset.
Note your final Rail voltage, and bias across emitters, and plug into the excel spreadsheet to see where you landed.
Crack open a beer and enjoy.
Some advice:
Build it, checking your work as you go.. start it up one channel at a time, with a variac... at a certain voltage the amp will wake up... two voltmeters are needed - one for bias, and one on the speaker outputs showing DC offset. Keep Bias low and slowly ramp up the variac until your at mains voltage. No smoke is good.
Slowly dial up the bias, wait for the sinks to get nice and hot, trim offset to zero, wait, go up another 20 mv on bias, again and again, until 125 - 130F on the sinks - its an iterative process, all the while trimming your DC offset while raising bias. You will notice ramping up the bias will disrupt your DC offset and vice versa. Do the other channel same way. Close the case all up and check bias and DC offset one more time, as the two channels enclosed will affect final bias and DC Offset.
Note your final Rail voltage, and bias across emitters, and plug into the excel spreadsheet to see where you landed.
Crack open a beer and enjoy.
But if you remember there was an "error" with Jan's boards. He used a one off pot so the boards (traces) had to be modified.
And I believe the low and high bias settings were switch controlled not pot variable.
And I believe the low and high bias settings were switch controlled not pot variable.
Yes there was an error on Jans boards concerning the pot, but very easy to rectify. I also carried out a mod to the layout to enable driver transistors to be more easily mounted to their heat sink, again easy to do.
My KSA has been running for about a year, bias and offset rock steady. Some component values were altered ( many thanks to forum members) to run with rails of +-25V, biased to give25W RMS at 8 Ohms with passive cooling, seperate supply for front end.
Now thinking about using two spare boards to construct another with 50W output. Cooling is the major problem as I would like it to be passive.
Alan
My KSA has been running for about a year, bias and offset rock steady. Some component values were altered ( many thanks to forum members) to run with rails of +-25V, biased to give25W RMS at 8 Ohms with passive cooling, seperate supply for front end.
Now thinking about using two spare boards to construct another with 50W output. Cooling is the major problem as I would like it to be passive.
Alan
Thanks Guys.
How can i rectify boards error due i have the channels working signals look beautiful but my concern its bias now.
and thanks for the excel sheet
regards and thanks wolf
I am afraid I can't remember it myself. You would have to go through the thread to find it. Don't worry, I have had to do it twice now. And will have to do it again as I am building 3 sets of KSA50 boards for a friend. One with Jan's boards, one with PinkMouse's boards, and one with boards from a guy in Poland.
I always forget to make notes somewhere. There used to be a Wiki where all the info was available. Can't find in anymore.
I always forget to make notes somewhere. There used to be a Wiki where all the info was available. Can't find in anymore.
Wolf, I tried to view the print side of my PCBs, but I could not see clearly due to how they are mounted in the chassis. My memory is not too good but I am thinking that the error only affected the DC offset pot ( near the input end of the board )and the fix is very straight forward.
Alan
Alan
i would appreciate your advice in order to solve the issue and bias them correctly also the correct dc offset.
thanks alibear, would be nice if you remember the fix, i must also try the boards in the scope with dc coupling in order to see the dc offset
regards wolf
@wolf...
Refer to the information in the wiki that Harry has posted. It has enough info on the correct biasing procedure.
I will try to find the fix for Jan's PCB.
Refer to the information in the wiki that Harry has posted. It has enough info on the correct biasing procedure.
I will try to find the fix for Jan's PCB.
Here you go, took me 3 minutes to find it using the Search facility. I just typed in "Jan PCB fix".
http://www.diyaudio.com/forums/solid-state/31077-krell-ksa-50-pcb-15.html#post571699
http://www.diyaudio.com/forums/solid-state/31077-krell-ksa-50-pcb-15.html#post571699
I can post K-Amps biasing spread sheet if you can't find it. I will check back shortly. I always found Jan's boards a real PIA in regards to mounting the drivers. You pretty much have to mount them on their own heat sink like Krell did. That was why the Pinkmouse board came to be.
I still have a few of the prototype pink mouse boards in my parts bins if anyone needs them. They were made by Advanced Circuits and I believe that I have six of those. There is no mask or silkscreen.
I recently rebuilt a friends KSA-50 that I originally constructed with Jan's boards with some gold plated Pinkmouse boards that are also sans silkscreen and mask. Those look fantastic.
Mark Gulbrandsen
I still have a few of the prototype pink mouse boards in my parts bins if anyone needs them. They were made by Advanced Circuits and I believe that I have six of those. There is no mask or silkscreen.
I recently rebuilt a friends KSA-50 that I originally constructed with Jan's boards with some gold plated Pinkmouse boards that are also sans silkscreen and mask. Those look fantastic.
Mark Gulbrandsen