Krell KSA 100mkII Clone

I built one channel with the original KSA100 voltages and found the heat generated almost impossible to deal with, so IIRC Neil and I decided to run 600va per channel, with the final, fully biased voltage being 43V @ 2.5A to the output stage. The measured class A & B output into 8ohms was 100W rms. The output into lower impedances was ultimately defined by the various safety 'features' we built in, which included primary fuse, overheating of the heat sink etc.

After deciding to reduce the heat I rewound the secondaries to get the lower voltage, but I had the exact equivalent of 4 of these sitting around:

http://www.antekinc.com/an-6445-600va-45v-transformer/

HTH

Stuart
 
Last edited:
I built one channel with the original KSA100 voltages and found the heat generated almost impossible to deal with, so IIRC Neil and I decided to run 600va per channel, with the final, fully biased voltage being 43V @ 2.5A to the output stage. The measured class A & B output into 8ohms was 100W rms. The output into lower impedances was ultimately defined by the various safety 'features' we built in, which included primary fuse, overheating of the heat sink etc.

After deciding to reduce the heat I rewound the secondaries to get the lower voltage, but I had the exact equivalent of 4 of these sitting around:

AN-6445 - 600VA 45V Transformer - AnTek Products Corp

HTH

Stuart

Hi Stuart,

If you want +-43v rails wouldn't one like this 32-0 32-0 600VA transformer be about right? Knowing Antek, most of their transformers are set for 115v primary so you generally get a little higher than rated output. I'm not sure how much you lose when fully biased but I would think this would get you close.
i still have a pair of the Mark Gulbrandsen boards. One of these days I will either have to build them or sell them. A shame for them just to sit.

Blessings, Terry
 
Last edited:
Terry,

You are quite right, if I had known I was going to build the krells when I bought the transformers I would have chosen the lower voltage, as it was I had to improvise. The current rating of the copper windings was greater than I needed, so rewinding was straightforward...If I hadn't needed a center tap, I could have simply unwound ;-)

Stuart
 
The amp does not pull anything like 12A per channel, it simply provides a very low impedance source for the current it does pull.

At full power, I think we saw maybe 2-3A per channel, and used a 5A primary fuse. Of course the class A nature of the beast means it pulls that current all the time until something drives it into class B, when it can pull more :)

HTH

Stuart
 
Not sure if you are addressing me or/and where your current number originates... but I already owned the transformers, I modified them to provide a little less voltage, which has the side effect of increasing the current capacity. To follow your analogy, if my boat and trailer weighed weighed 3000lbs, but my truck had capacity for 6000, should I go buy another truck?

Unlike trucks, transformers are more efficient and generally better in all other parameters (except weight) as they increase in size ;-)

I used each 600w transformer in a role where it was expected to provide (2 * 43 * 2.5) 200+W constantly, with potential for higher peaks. I don't think that level of overrating is quite as bad as your commentary would imply ;-)

Others ran the transformers without rewinding and would have used even more power (2 * 57 * 2.5==300W) at idle, so 600va+ per channel is well in the ballpark of what prudent engineering would suggest.

Stuart
 
I have a pair of 40vac 1kVA transformers which gives me 53v at the rails and don't ask me why, but with a .630 bias, the highest temperature I get at the output transistors, MJ15003/4, are ~40*C. Stuart gave me some monster heat sinks and I have AC fans facing down atop each heat sink which I will later put below facing up. For now, I run the amp at .400 bias with no difference to the sound although my hearing is so bad that I may not notice the difference. It is loud with no distortion on the oscilloscope--with a ~20 step resistor attenuator, I can't go beyond 4 or 5 steps fearing I will blow out the windows and neighbors. The temps do not rise over time especially with the MJE15033/4s. I don't think I have more than 20 hours of burn in time yet. By the way, each transformer weigh 22 pounds.
 
Not sure if you are addressing me or/and where your current number originates... but I already owned the transformers, I modified them to provide a little less voltage, which has the side effect of increasing the current capacity. To follow your analogy, if my boat and trailer weighed weighed 3000lbs, but my truck had capacity for 6000, should I go buy another truck?

Unlike trucks, transformers are more efficient and generally better in all other parameters (except weight) as they increase in size ;-)

I used each 600w transformer in a role where it was expected to provide (2 * 43 * 2.5) 200+W constantly, with potential for higher peaks. I don't think that level of overrating is quite as bad as your commentary would imply ;-)

Others ran the transformers without rewinding and would have used even more power (2 * 57 * 2.5==300W) at idle, so 600va+ per channel is well in the ballpark of what prudent engineering would suggest.

Stuart

Hi Stuart,

No I was addressing the post I quoted where two Antek 10440 transformers were suggested. Those are 18lb 1000VA 40VAC transformers, They are rated at 12.5A. If the amp pulls 2.5A quiescent Class A, don't you think 5A would be more than enough? I guess the Krells are all about excess but 5X the required amperage seems a bit much to me.
 
Last edited:
Well, the 2.5A is for 100W class A into 8ohms, there were people shooting for full class A power into 4ohms and lower, ie idle current of 5A ... that's serious commitment ;-)

I don't think I've ever seen a KSA200 in the flesh, but by running the numbers I assume it must be a monstrous beast.

To each, his own :)
 
I have a 1500 VA transformer in my KSA-50 clone. That's a little overkill but the amp realy sounds great. I am going to end up making my 100 as KMA-100 mono blocks and I have forced air heat sinks as in the original Krell so heat isn't going to be an issue. I will shoot for a 1kva transformer then and try to arrive as close to the original rails as possible. I suspected most on here were using Antek transformers. I don't think many of us could do these projects if they didn't exist. Plitron is way too expensive. But thanks to all of you who have given me some insight into the transformers currently being used.

You guys may not realize that I am Mark Gulbrandsen. After spending the last seven years getting D-Cinema in place in my 5 State area I came back but couldn't remember my old password. So some time back I re-registered using my middle name since the system wouldn't allow me to re-register with a previously used name.

P.S. I probably have about 15,000 on hours on the KSA-50. It has never ever so much as even flinched or burped.

Mark
 
Last edited:
Hi Mark,

I wondered what happened to you. I still have a pair of your boards. They sit on top of my parts bin, right in front of me when I work at my bench. I keep getting tempted to do something with them but the thought of 40lbs of transformers and all that wasted heat always stops me in my tracks. I still listen to my KSA 50 from time to time. It is a better winter amp.

Blessings, Terry