Double everything except voltages when working out resources for two channels.
I can't see how stating two different values for Total output bias current helps anyone make good decisions.
I can't see how stating two different values for Total output bias current helps anyone make good decisions.
Build it using ~+/- 47 volt rails.
scale and cool drive stage accordingly.
Some early buiders were unaware of the fact that Krell fitted the driver stage with a substantial heatsink.
cool and scale output-stage accordingly. btw, it functions fine in a/b also
scale and cool drive stage accordingly.
Some early buiders were unaware of the fact that Krell fitted the driver stage with a substantial heatsink.
cool and scale output-stage accordingly. btw, it functions fine in a/b also
Last edited:
"Build it using ~+/- 47 volt rails."
Heck, you may as well just build the KSA-100 at that point because that's where you are at that rail voltage!
Well, I used three pairs of output devices per channel in all my KSA-50's and 38 volt rails. I remember us settling on this some time back, have constructed a half dozen amps and I've never used 4 pairs per channel and I've never had an amp fail, not even blow a fuse. The reliability has been amazing. It would be great if you guys could post a working version of the spread sheet again as it would be fun to experiment with more/less OP devices and different bias settings.
Also, I have built up a pair of amps on some gold plated boards a friend had made in China, and I am about to test them on the bench. I want to just power up the main board and driver stage first. I remember doing this when we did the KSA-100 and that board was capable of directly driving an 8 ohm speaker. But has anyone fired up a KSA-50 minus it's output stage? I would think it would be all happy being powered up this way for bench testing...
Lastly, if anyone needs some boards I do still have a small stack of the Delta Audio KSA-50 boards that I'm never going to use. Those boards work fine after the corrections are made, but the driver arrangement is just too tight for me...
I recently rebuilt a Flame Linear 700 using the White Oak Audio Labs board. It is an early Gen 1 Model 700 from probably 1972 and I converted it to full complimentary output. All I can say is WOW! My KSA-100 has some serious competition now. It is nothing like the sound of the original 700 and is just amazing.
Highly recommended if you guys are looking for a new project. I think the 700 cost me 120 bucks on Ebay, the driver board and new devices and hardware was another 250. It clips at 425 watts per channel into 8 ohms which is quite a bit better than the 380 watts the original 700 put out! And the best part is it's completely stable and reliable unlike the old version.
Mark
Heck, you may as well just build the KSA-100 at that point because that's where you are at that rail voltage!
Well, I used three pairs of output devices per channel in all my KSA-50's and 38 volt rails. I remember us settling on this some time back, have constructed a half dozen amps and I've never used 4 pairs per channel and I've never had an amp fail, not even blow a fuse. The reliability has been amazing. It would be great if you guys could post a working version of the spread sheet again as it would be fun to experiment with more/less OP devices and different bias settings.
Also, I have built up a pair of amps on some gold plated boards a friend had made in China, and I am about to test them on the bench. I want to just power up the main board and driver stage first. I remember doing this when we did the KSA-100 and that board was capable of directly driving an 8 ohm speaker. But has anyone fired up a KSA-50 minus it's output stage? I would think it would be all happy being powered up this way for bench testing...
Lastly, if anyone needs some boards I do still have a small stack of the Delta Audio KSA-50 boards that I'm never going to use. Those boards work fine after the corrections are made, but the driver arrangement is just too tight for me...
I recently rebuilt a Flame Linear 700 using the White Oak Audio Labs board. It is an early Gen 1 Model 700 from probably 1972 and I converted it to full complimentary output. All I can say is WOW! My KSA-100 has some serious competition now. It is nothing like the sound of the original 700 and is just amazing.
Highly recommended if you guys are looking for a new project. I think the 700 cost me 120 bucks on Ebay, the driver board and new devices and hardware was another 250. It clips at 425 watts per channel into 8 ohms which is quite a bit better than the 380 watts the original 700 put out! And the best part is it's completely stable and reliable unlike the old version.
Mark
Last edited:
Thanks Pink! I will fire these up later today that way first. Resistance measurements all check ok, and the same on both boards...
Mark
Mark
One more question for you guys so I don't have to tear into my KSA-50 to find out. I have one set of main board and driver board hooked up and run up to 20 volts per rail on my small Tektronix Split bench supply. This amp oscillates at quite a high frequency! I was able to stop it by adding another 330pf across C-105 and C-106, base to collector. Its specified as 390 pf on the Delta schematic, 330 PF in the Delta parts list and only 39 pf on all the other schematics of the KSA-50 that I have. So whats the correct value? I also note that C-107 & C-108 (100 pf) do not have locations on the board, so if need be those can be soldered across the bottom of the driver PCB.... My amp does not have C-107 & C-108 and without substantial tear down I have no idea what value cap I installed at C-105 & C-106. Do any of you guys remember if you ran into this oscillation and what value did you use at C-105 & 106? Also, did you even bother with C-107 & 108?
Thanks in advance for any clues...
Mark
Thanks in advance for any clues...
Mark
Attachments
Mark,
12 years ago you asked the same question:
http://www.diyaudio.com/forums/solid-state/31077-krell-ksa-50-pcb-33.html#post576524
And this was the only straight answer 2 months later:
http://www.diyaudio.com/forums/solid-state/31077-krell-ksa-50-pcb-67.html#post625808
In my build I used 39 pF dipped mica and had no problems.
12 years ago you asked the same question:
http://www.diyaudio.com/forums/solid-state/31077-krell-ksa-50-pcb-33.html#post576524
And this was the only straight answer 2 months later:
http://www.diyaudio.com/forums/solid-state/31077-krell-ksa-50-pcb-67.html#post625808
In my build I used 39 pF dipped mica and had no problems.
Hi Grimberg,
Thanks! And yes... it's been a very long time. Mine is so old I should probably think about recapping it by now.
And good job finding that post in all those pages of KSA-50! I was not able to turn up anything with the search engine.
Best,
Mark
Thanks! And yes... it's been a very long time. Mine is so old I should probably think about recapping it by now.
And good job finding that post in all those pages of KSA-50! I was not able to turn up anything with the search engine.
Best,
Mark
the feedback around the VAS is a bit unusual on this.
It has a double EF VAS/TIS.
normally we see the feedback cap from second collector to first base.
In this we see the cap from second collector to second base. That will behave differently.
Might be worth trying the "back to first base" NFB.
It has a double EF VAS/TIS.
normally we see the feedback cap from second collector to first base.
In this we see the cap from second collector to second base. That will behave differently.
Might be worth trying the "back to first base" NFB.
The Delta Audio BOM had c-107/8 as 100 pf
The Pink M boards had none nor the wiki BOM
I used 33 pf for c-105/6 ( PM pcb )
The zobel was always there at power up and never pulled to see if an oscillation was present
Unlike my 150 mono block that went crazy without it
Regards
David
The Pink M boards had none nor the wiki BOM
I used 33 pf for c-105/6 ( PM pcb )
The zobel was always there at power up and never pulled to see if an oscillation was present
Unlike my 150 mono block that went crazy without it
Regards
David
I switched C-105 & 106 to 39 pf and I still have the strange oscillation issue, but I am able to cure it by removing C-105 and reinstalling the 330 pf at C-106. I am also able to quench it by jumping from the emitter of both Q-107 & Q108 to the feedback line with a 39 pf cap. I'm afraid I have a device or two that seem to be very high frequency sensitive... For now I am going to leave it this way.
Mark
Mark
Strange that you have this issue.
I am using 56pF cap and run without any Zobel network on the output. I looked at my square wave on the output and there is no hint of overshoot.
I am running the Pink Mouse boards. The amp is in daily use since the day I finished it many moons ago.
I am using 56pF cap and run without any Zobel network on the output. I looked at my square wave on the output and there is no hint of overshoot.
I am running the Pink Mouse boards. The amp is in daily use since the day I finished it many moons ago.
It was more an issue of what I had in stock at the time. Too lazy to experiment later on.
Wow, so good to see posters from a good while back still showing interest and giving advice on this thread. Great forum, super community.
Regards
Alan
Regards
Alan
