Krell KSA 50 PCB

[neychi]

Hi neychi, could I ask another question regarding the KSA with 48V rails, two output devices per channel and 325mV across emitter resistors giving 1.3A bias. I assumed emitter resistors to be 0.5 Ohm. This equates to 27W class A RMS into 8 Ohm, then it transitions into AB. Is this correct, or have I misunderstood?
Best Regards
Alan

That's correct Alan. You meant two pairs, not two, output devices per channel, of course ;-)

 

[gaborbela]

I believe the higher voltage like 48VDC, not KSA50I or KSA50II that is the higher power version.
KSA 100 uses mosfets in the circuit. If you want that much power please look for the KSA100, there is a thread on the forum guys built it successfully.
Be aware lots of power with lots of heat.

 

[Dave murrey]

Hi gaborbela my pcb has printed on it krell ksa mk2 -48v +48v.
I have looked at images of krell ksa 100 and its a differant pcb it has extra transistors.
This pcb looks exactly like genuine krell ksa 50 mk2 pcbs.
I would like to make my amp with same voltages as genuine krell ksa 50 mk2 as I only need 50w a channel but would like 50w of class A like original amp.
Thanks Dave.

 

[gaborbela]

That is not the KSA50!
With that high voltage, I am sure that is a higher power, Krell.
They had an 80W version or 75W, I am not sure exactly.
You should go to ZeroZone site to find out more about your PCB.

 

[alibear]

Thanks for the reply neychi, yes 2 pair per channel.
So given 48v rails 1.3 A bias gives 27 W RMS into 8Ohm, but heat dissipation will be 125W per side ( a lot for 27W ) To get 50 W class A we will be dissipating 173 W per side.
This is assuming the output stage runs at the same supply voltage.
With the lower voltage rails of the earlier version of 37 V , 50 W RMS can be achieved with a dissipation of 133W per side.
So am I correct in thinking that the later Mk2 needed even more cooling? Or is there something I am missing?
Best Regards
Alan

 

[gaborbela]

DualMono Class A

 

[Dave murrey]

ive come across this thread a while ago, since i've owned this amp for 10 years now just a pic;
notice that the ac-wires are thin and the dc-wires are thick
powersupply-voltage +/- 47.7 v
standing current per output device =500 mA(250mv/0.5ohm)


comments and questions r appreciated (i own a dynatek and i know how 2 use it)

Klaas

I no this is a long time ago but do you no what the voltage was at the transformers of your amp please.
thanks dave.

 

[gaborbela]

Yes, you can use your PC boards for KSA50 (It is advertised KSA50II) but it is not fully Class A biassed at 48V rail voltage.
I read some KSA50 and KSA100 restoration and KSA50 uses about 26V transformers.
If you want 48V rail voltage full Class A bias you will need like 1KVA transformers per channel. If You want to build a 50W please you do not use 48V rail voltage because you will need higher V power supply caps etc, it will cost you more.

 

[gaborbela]

Please take a look he uses 35V transformers 1KVA per channel for the KSA100
35V under load will give you 42V or so.
So even the KSA100II does not use 48V rail voltage!
DIY Krell KSA-100 Clone | Flickr

 

[Dave murrey]

Please take a look he uses 35V transformers 1KVA per channel for the KSA100
35V under load will give you 42V or so.
So even the KSA100II does not use 48V rail voltage!
DIY Krell KSA-100 Clone | Flickr

Hi gaborbela Kvholio has a real krell ksa 50 mk2 with +/-47.7v on rails on page 559 of this thread.
Was this krell class A or did they do a ab version.
thanks again Dave.

 

[neychi]

Thanks for the reply neychi, yes 2 pair per channel.
So given 48v rails 1.3 A bias gives 27 W RMS into 8Ohm, but heat dissipation will be 125W per side ( a lot for 27W ) To get 50 W class A we will be dissipating 173 W per side.
This is assuming the output stage runs at the same supply voltage.
With the lower voltage rails of the earlier version of 37 V , 50 W RMS can be achieved with a dissipation of 133W per side.
So am I correct in thinking that the later Mk2 needed even more cooling? Or is there something I am missing?
Best Regards
Alan

Alan, ksa50 is NOT fully biased class A amplifier, simple as that. Dan obvious had his reasons why he put higher rail voltage in later versions instead of lower rail voltage on earlier amps. Higher voltage seems to be more important to him than full bias. Sound quality and ability to drive more demanding speakers perhaps? Between two of my Krells one with higher rail voltage is a true winner.
Btw., two pair of output devices will not survive full bias for 50 W class A on 48V DC rails.

 

[neychi]

Please take a look he uses 35V transformers 1KVA per channel for the KSA100
35V under load will give you 42V or so.
So even the KSA100II does not use 48V rail voltage!
DIY Krell KSA-100 Clone | Flickr

Original Ksa100mk2 had 52 to 56V DC rail voltage. Clone is a different story.

 

[Dave murrey]

Alan, ksa50 is NOT fully biased class A amplifier, simple as that. Dan obvious had his reasons why he put higher rail voltage in later versions instead of lower rail voltage on earlier amps. Higher voltage seems to be more important to him than full bias. Sound quality and ability to drive more demanding speakers perhaps? Between two of my Krells one with higher rail voltage is a true winner.
Btw., two pair of output devices will not survive full bias for 50 W class A on 48V DC rails.

Thank you this explains a lot so KSA 50 MK2 was not full class A but sounds best.
Reviews I have read say that MK2 sounded better than original KSA 50 but it looks like there is a few different versions of this amp.
KSA 50 with pcb that looks more like delta pcb then pcb that looks like my pcb without dale resistors then the one with dale resistors.

What I am trying to achieve at the moment I have a yaqin mc100b valve amp I am pleased with switchable class a /ab but sounds best in class A.
Really big sound stage good with voices but lacks that treble sparkle up top
and the deepest bass ( sometimes I play bass cds)
I remember going to a demo at a hifi shop in 1984/85 when my father was looking for speakers a krell KSA sounded amazing on some high end keff speakers I am guessing it was a KSA 50 mk2.
He played a demo disc of a Everest double glazing advert it sounded so real the slam in the bass was amazing the sound stage and imaging was great.
I borrowed my brothers creek amp it had the bass and treble but did not have the big fill the room sound stage.
My speakers are Ruark crusader 2 and I am also making a Kondo m77 clone kit to go with krell as I dont have a pre amp.
Sorry for the long story but what I am asking is should I go for 48v not full class a or lower voltage class A or would full class A version have better valve like sound stage.
I don,t play classical music mostly pop music/vocal stuff Nora Jones ect.

 

[neychi]

In that case go for the higher voltage version, Dave. After all, you listened to it and you liked it

 

[alibear]

Just a thought---- supply the PCB with the +-48V to satisfy the opinion that running at this voltage sounds better and supply the output stage with say 37 V as per the original. That way you could bias to run 50W class A with still manageable heat dissipation. Any opinions,or is is nonsense.
Best Regards
Alan

 

[neychi]

The design of the amplifier determines a idling current "sweet spot", which means that at a certain amount of idling current, amplifier has the best sound and lowest, or satisfactory, amount of harmonic distortion. So basically, high bias is not a guarantee for excellent result...optimal bias is what you need. You can, of course, go either way, high bias/lower rail voltage or lower bias/higher rail voltage as long as power dissipation is within the limits allowed by output devices and cooling ability.

 

[Blitz]

May I ask those who built it and compared it with other builds like Pass F5 or other class A SS power amps: Is it still worth it to build it or are other models superior...given that this was invented in the 80s...and yes, I wanted one of them as well when I was a student and could not afford them...so it is easy to convince me to build one, but really...should I ?

 

[Medisinmannen]

Do it so you can stop thinking about it! There will be time to build Pass after KSA.

I have just started building, so i cannot share opinions on sound etc. I thought about Pass designs, but went for the KSA just because it is considered legendary with its "slamming bass". KSA can handle any speaker, some Pass cannot.

I have collected parts and can share BOM's(Mouser for most parts, TME for cheap toroids). KSA is a costly build. I have issues with sourcing second hand parts, so all from vendors. Boards eBay(20), Mouser; driverboard(100) output(100) caps(140), toroids tme(120). Prices(USD).

Heatsinks is ebay, small size but cheap. Uncertain if they will cool enough.

 

[Blitz]

Well, what is the recommended board version ?

I understand that the KsA-50 Mk2 is recommended with fet CCS and higher voltage...but is there as well a clear opinion if bipolar vs. mosfet output devices are to prefer ?

I have build Pass F4 and F5....but I am not that impressed with them to be honest. My 300B PP amp beats them easily. I am disappointed on the Pass stuff to sound too warm, lush, by far not as holografic and not owning the fine resolution or micro-dynamics of my tube gear....and I build the 300B-PP amp to sound very high resolution, not the warm, fluffy way you can as well build.

I read on Counterpoint power amps (hybrid) and Michael Elliott chose beginning of the 90s intentionally bipolar power transistors over Mosfets as he did not like the too warm and not clear enough sound of mosfets...maybe a myth, but I understand that the KSA-50 was as well build initially with bipolar and later with mosfets, correct ?

Has anyone built/heard both versions ?

When I look at aliexpress I find:

- Versions with 2SA1943: YJ KSA50 Klasse A Krell 50 Watt + 50 Watt 2SA1943/2SC5200 MJE15034/MJE15035 Abgeschlossen verstarkerplatine in YJ KSA50 Klasse A Krell 50 Watt + 50 Watt 2SA1943/2SC5200 MJE15034/MJE15035 Abgeschlossen verstarkerplatine aus Verstarker auf AliExpress.com | Alibaba Group

- Versions with MJ15024G: YJ KSA50 Klasse A Krell 50 Watt + 50 Watt MJ15024G/MJ15025G MJE15034/MJE15035 Abgeschlossen verstarkerplatine in YJ KSA50 Klasse A Krell 50 Watt + 50 Watt MJ15024G/MJ15025G MJE15034/MJE15035 Abgeschlossen verstarkerplatine aus Verstarker auf AliExpress.com | Alibaba Group


...tbc...

 

[henryve]

Well, what is the recommended board version ?

I understand that the KsA-50 Mk2 is recommended with fet CCS and higher voltage...but is there as well a clear opinion if bipolar vs. mosfet output devices are to prefer ?

I have build Pass F4 and F5....but I am not that impressed with them to be honest. My 300B PP amp beats them easily. I am disappointed on the Pass stuff to sound too warm, lush, by far not as holografic and not owning the fine resolution or micro-dynamics of my tube gear....and I build the 300B-PP amp to sound very high resolution, not the warm, fluffy way you can as well build.

I read on Counterpoint power amps (hybrid) and Michael Elliott chose beginning of the 90s intentionally bipolar power transistors over Mosfets as he did not like the too warm and not clear enough sound of mosfets...maybe a myth, but I understand that the KSA-50 was as well build initially with bipolar and later with mosfets, correct ?

Has anyone built/heard both versions ?

I don't know of ANY KSA 50 that had MOSFET outputs.