Krell KSA 100mkII Clone

I've measured KSA's at varying bias levels and found that above a fairly low bias the distortion stays relatively unchanged. However, distortion CONSISTENCY did improve with bias i.e. the THD vs. power graph was quite flat, much flatter than most class-AB and class-D amps. So even though the absolute THD percentages aren't too impressive and actually higher in areas than at lower bias levels, the key is that the figure doesn't change much as the power changes, which is IMO more important than varying distortion levels that may be lower at some power levels. Amps exhibiting this behaviour usually have great transient bass response ("grip/slam"), and this amp is no exception.

It should be noted that the power supply needs to be up for the task since increased bias will cause sag and increase 50Hz harmonics, thereby increasing THD. Especially at lower (load) power levels high bias therefore typically results in higher THD due to the large 50Hz components. Separating the output stage power supplies will help with this though since ripple on the input & driver stages affect total distortion far more than just on the output.

Moral of the story: push the bias as high as the power supply and heatsinks will allow. Even though it's always a good idea, if the supply is on the smallish side, get a separate supply for the PCB's. The quality of the power supply is paramount to get proper performance at high bias levels. Note I say quality; a cheap 1kVA toroid and large capacitor bank will not help much if the transformer has poor load regulation and the caps high ESR.
 
How much should one be concerned if the center taps of the large/ little transformers do not show 0v potential difference... would this not induce hum into the amp or maybe DC offset?

I man should one be concerned or just go implement the dual trafo and be able to trim off the differences??
 
oww i had a look at some of them.they are too long..i dont have much time...are you a krell expert ?:) what do you think about the feedback capacitor of krell ksa100? is it very important ?i have bipolar Frolyt 6,3volts 470uf.and 10volts 470uf..i dont know they are quality enough.. i think i will use Blackgate 16 volts 470uf..my friend is gonna use two capacitors as a feedback cap..he says its original and better..i dont agree with him :S
 
A good quality cap is recommended, at the very least something from a reputable manufacturer and not a generic thing. Unless you go Black Gate I'd say rather use two standard electros than a bipolar.

The voltage is very low; you can use 4V too if you like.

I would recommend a single 470uF 6.3V Black Gate NX. If you look at the literature of Black Gate N/NX series it explains that they actually are 2 series caps inside the same package to reduce ESL and space compared to two discrete caps. So using two of them is not only unnecessary but also counter-productive since you're lengthening the signal path and adding extra inductance. Also important if you use one of these BG caps is that the 100nF bypass cap (C4) should be omitted.
 
ow i see..i think its about the inductance? connecting two inductance in series makes a half inductance right? its the best to use two high quality bipolar caps in series connection right?
and question two..is there two normal caps in series inside the bipolar caps?what is the real purpose of bipolar caps?someone says its for not having phase errors..i opened the caps but didnt understand..thanks
 
Hi,
an output stage can be biased to optimum ClassAB (Vre=~20mV)
or
high biased (Vre>25mV)
or
biased to ClassA (Ire=Ioutputmax/2)

All are likely to sound a little different.

If the emitter resistor are reduced in value the bias current must be increased to maintain that Vre~=20mV.
If the output stage has many devices and the bias current through each device is the same as an optimum biased ClassAB then the total bias current approaches the ClassA bias.

2pair with 0r68 will be optimum ClassAB @ ~59mA.
10pair with 0r22 will be optimum ClassAB @ ~909mA.
50W ClassA will be biased @ ~1800mA.
I wonder how different these set ups will sound?
 
PWatts said:
Well the centre taps of two separate transformers will naturally not be at the same potential to each other, and if you left them floating it would be a problem, but since you tie them together it becomes 0V after all.

Typically yes.... but the trafos I have lying around are 1kva types and those do not play well together always when it comes to having 0v across their center taps... anyway this is a uncommon situation...
 
It could be due to they are not bifilar-wounded and because of that the two windings has has different length and therefor they do have different resistance and inductance when feeding a power load so ther will be a miss-mach for real zero (easy to get a buzzing ripple of that damn 50 or 60 Hz).

This is a common problem for standard industrial transformers not made for Hi-Fi products (e.g. standard transformer for standard power supplies). When combining two of them in a Hi-Fi amp the miss-mach could be even worse.
 
PWatts said:
Strange, two voltages that aren't referenced to each other should be able to be tied together without problem.

They are "electro-magnetically" referenced since they share the same mains....


I know I have 2 monoblocks, one with a faulty centertap (it is a few volts off center) and whenever I inset that amp into the system... the rest of the system suddenly experiences an increase in audible hum...
 
Thanks to Ryan, now I can go further with my KSA100 project. Two Diycable transformers are ordered and shipped by now. ;)

Next step will be to drill all holes in the driver sink for the drivers.
But, my drilling machine hasnt arrived yet. Should be at my house early next week.
Then I can drill holes all over the chassis including all those heat sinks needed for the amp.

I think I should order the main chassis from the HiFi2000 in Italy. But its not a good time to order parts all over the world now a days, is it :xeye:

Regards :cool:
 
still4given said:


One of my KSA50's does that. By itself, no hum. Hooked up with any other amps I get a hum from all. What do I check for?

Thanks, Terry


Terry:

with the KSA connected to the mains, (but not to the rest of the system) check the potential difference of it's ground vs. the rest of the system's ground. It should be close, if not then the trafo could be off...

I got 2 identical trafo's of DIYaudio (some guy's Aleph project) and put them in a couple of Forte Model 3's. One hums up the system, the other does not when connected . The one that does, is off by 6vac gnd to gnd.