Krell KSA 50 PCB

Jump to Latest
#8,821
there is one basic difference between the earlier delta PCBs and the later Pink's
The early PCB had a switchable bias voltage. It effectively gave a Summer/Winter option
Where Winter is full ClassA and Summer is a reduced bias ClassAB to keep the heat down, especially if you run air con.
 
#8,822
I bought these boards, from a member here, that have KSA-50MKII printed on them but
obviously there are many more parts than are shown in the MKII schematic at the start
of this thread because I think it is actually the KSA-100 MKII board. They are from
jimsaudio on ebay:
http://s22.postimg.org/kg7twdze9/IMG_2809.jpg

It includes FET diff pairs, current source diodes, many more parts as shown in this schematic: Looking for the schematic on here, can't seem to find it.

I like the fact that the layout is very close to the original KSA-50 MKII and that I could use
them to build the KSA-100 if I ever wanted to. It will be easy to build either design.

Does anyone know if the KSA-50 was ever built with the diff pair FETs etc.?
 
#8,823
The latest version of ksa50mk2 was identical to ksa100mk2 except lesser number of driver/output devices, some resistor values and 27V zeners instead of 39V in ksa100mk2.
PC boards were identical for both amps. Extra drivers and output devices were simply omitted in smaller amp.
 
Last edited:
#8,824
Hello, have a nice day.

Thank you very much indeed in advance for reading my mail.

I'm a student and I have the innate passion in building loudspeakers for myself. However, regarding the electronic part, I have not the skills to design an amplifier and I cannot effort the amp that my loudspeakers would deserve.

Trying to find out a solution, I've seen that the Krell Ksa's are well "cloned" amplifiers, but there are also rather a lot of options that I can’t manage. Subsequently I've found this 3d.

I would gently like to ask to the forumers if, for example, the following pcb boards (that could be purchased easily from my country) are good to assemble a stable/good amplifier without issues.


Otherwise, if not, is there a good/well known kit for my purpose?

Douk Audio Class A Power Amplifier 2.0 Channel Stereo Hi-Fi Transistor Amp Board | eBay

One pair KRELL KSA50 Class A Power amplifier board 50W+50W with 2SA1943/2SA5200 | eBay


Thank you.
Davide
 
#8,826
neychi said:
The latest version of ksa50mk2 was identical to ksa100mk2 except lesser number of driver/output devices, some resistor values and 27V zeners instead of 39V in ksa100mk2.
PC boards were identical for both amps. Extra drivers and output devices were simply omitted in smaller amp.
Click to expand...

Does anybody can share the schematic of the latest version of KSA50mk2 that is identical to KSA100mk2.
I want to make my KSA100MK2 that is working great to KSA50MK2 to lower the heat. I have already two 25Vac 620Watts toroidal transformer for mono blocks.
 
#8,834
Guys, Thank you for the data on capacitors. The Epcos's specs are great but 47000/63 are not stoked on either Mouser or Digikey... Unless I parallel some....mmm.

Right now I'm buying parts for a KSA100 Klone. A couple of questions if you don't mind:

1-How important would you say is to use two transformers, I saw photos of real Krell units with just one...KSA100s for example?

2-Planning to use plastic output devices in the 100 Klone. I saw some chinese Klones using 4 pairs of 2SC5200-2SA1943 or NJW0281 NJW0302 . These are 150W devices.
There's another kit using 3 pairs of TTC5200-TTA1943 another 150W device.
However many people say they prefer the MJE15034-35 G which is a 50W device.
Is there a logic limit of the pairs you can use on this design? I'd really like to stay in the safe zone with these heaters. It's my first class A power amp.


Any comments? your input is very valuable to me particularly on this early stage of the process.

Thanks
 
#8,837
the 100W of ClassA Krell KSA100 uses 2000W of output devices in one channel.
It also uses a 1kVA 52-0-52Vac transformer.

If you want to use 1943/5200 you would need to at least match that 2000W. But we know the 1943/5200 have a poor SOA and taking account of that combined with their lower maximum Tj of 150degreesC, I'd suggest you need at least 2500W of devices to get a reliable 100W of ClassA.
That comes to 9pair, instead of 4pair of 250W devices.
 
#8,838
AndrewT said:
the 100W of ClassA Krell KSA100 uses 2000W of output devices in one channel.
That comes to 9pair, instead of 4pair of 250W devices.
Click to expand...

Appreciate the information, Andrew.
Besides the original TO3 devices, are there any plastic substitutes that would be appropriate for this build?

Is it practical to use 9 pairs? The plastic devices are very cheap and easy to wire/mount, but I guess some driver circuit reworking would be needed.
 
Last edited:
#8,840
There are some 230W To264 devices.
But all these plastic devices have a 150degrees C maximum Tj. whereas the To3 can have a 200degrees C Tjmax.
A good SOA 3pair 230W 150C is probably equivalent to a 2pair To3 250W 200C
That brings a Krell 100W ClassA equivalent up to ~5 or 6pair to match reliability. You would need to do some reliability modelling into reactive loading.
 
You must log in or register to reply here.
Top Bottom