My apologies as I have no clue. I don't see where the emitter resistors were changed on the product sheet that I was charged for.
So, no other thoughts or opinions on this matter. Has Krell's answers satisfied everyone as to why I have a refurbished Class A amp that runs cool with no heat whatsoever? It sounds great, but I guess Krell has come up with a way to run this beast cool. Please add any useful comments so that I can understand this better. Thank you.
Can't be. No way a "few watts" is going to move my Hales at the volume levels I have been playing it at with the bass response I'm getting (chest moving) What you are saying is that it is now a low powered 3 to 4 watt amp????? Krell stated that it is class A all the way to 100 watts. Sure sounds like it.
Whose emails? You mean Krell's emails with the 85mv idle bias that they stated? Explain to me how this translates to 1.8 watts.
This is the classic example of buyers remorse, you pay your money and it sounds fabulous but for what ever reason your still not satisfied?
OK I get it, so to solve the last remaining issues and bring you around full circle, put a voltmeter on the big emitter resistor closest to each output tranny and note what it is in MV.
Now start turning up the bias pot until you feel what ever warmth will satisfy you and listen again. Check the output for DCO, DC offset, at the new level, there is a pot for that too. Adjust output for lowest level
Use protected clip leads probes so you don,t smoke anything adjacent. Their saying they adjust for lowest distortion is probably right, but doesn't necessarily mean another higher bias setting wouldn't,t suit you better. Distortion is a one dimensional test but music is a 3 dim experience.
So have at it and report back. It's easy enough to go back to the original setting
Enjoy
David
OK I get it, so to solve the last remaining issues and bring you around full circle, put a voltmeter on the big emitter resistor closest to each output tranny and note what it is in MV.
Now start turning up the bias pot until you feel what ever warmth will satisfy you and listen again. Check the output for DCO, DC offset, at the new level, there is a pot for that too. Adjust output for lowest level
Use protected clip leads probes so you don,t smoke anything adjacent. Their saying they adjust for lowest distortion is probably right, but doesn't necessarily mean another higher bias setting wouldn't,t suit you better. Distortion is a one dimensional test but music is a 3 dim experience.
So have at it and report back. It's easy enough to go back to the original setting
Enjoy
David
David, it has nothing to do with "Buyer's Remorse." I had the thing upgraded at Krell. Yes, it sounds nice, but no heat whatsoever, so based on what I know about the history of this amp, I inquired all around at how the unit could work and produce no heat. Why the hell would Krell bias down an amp to only 1.8 watts per channel in Class A mode on a classic Class a amp, based on what the other poster stated. That's why I'm confused. Buyer's remorse doesn't enter into it and checking it out as you stated is not an option for me. I would not feel comfortable doing that. Why would Krell give me back an amp at only 1.8 watts per channel in class A. Doesn't make any sense at all or sound like it.
Power is IsquaredR. Lets take your amp originally with 2.6A bias and 8 Ohms. I2R is 54W. Its push pull so we can double that and get 108W peak.
Now do the same at 340mA and you get 1.85W peak.
For RMS divide by 1.4, and you get 77W and 1.3W RMS
none of which matters a jot if you love the sound. Class AB is not necessarily evil.
It matters if I'm not being told the truth by Krell. Here's what they said:
Sam, The unit is running as it should be in Class A operation. If it were an A/B design the bias would be adjusted to about 7 mv at idle rather than 85mv.
Ray
Seems to me that all this discussion hinges on Krells definition of Class A and the accepted definition as used and accepted by engineers. Sliding bias, dynamic bias, bias that changes in "anticipation" of the signal etc etc etc are not Class A.
As to being told the truth, well it sounds like you are being told that it conforms to Krells definition of Class A. Whether that would stand up if challenged I have no idea.
As to being told the truth, well it sounds like you are being told that it conforms to Krells definition of Class A. Whether that would stand up if challenged I have no idea.
I have an early KSA 100 with Motorola 15024 / 15025 outputs and it had 1 ohm emitter resistors. (Dan liked Motorola semiconductors - later it was Japanese ones). During it's life time Krell went from 2 watt 5% carbon composition to 5 watt 1% non-inductive wire wound for the 1 ohm emitter resistors and had a batch of output devices (a million was the rumor) run at Motorola with a narrow gain group - labeled Krell A and AA.
The original bias per device was 625 ma (626 mv on a 1 ohm resistor).
This level of bias eventually baked the thermal compound into dust and every output stage had to be rebuilt and often more than once. Even if the new output devices had a better "efficiency" that only means less voltage drop across the output devices, the current to get 100 watts class A is still 625 ma per device - they've turned the bias down and 85 mv is only a couple of watts class A bias - the output stage will last much longer and if it sounds good then great.
I converted mine a decade ago into a Borbely Milenium, put 700,000 uf in a CLC filter into the chassis and run it at 50 watts Class A for a decade now - very happy with the sound and haven't lost a semiconductor in 10 years.
The original bias per device was 625 ma (626 mv on a 1 ohm resistor).
This level of bias eventually baked the thermal compound into dust and every output stage had to be rebuilt and often more than once. Even if the new output devices had a better "efficiency" that only means less voltage drop across the output devices, the current to get 100 watts class A is still 625 ma per device - they've turned the bias down and 85 mv is only a couple of watts class A bias - the output stage will last much longer and if it sounds good then great.
I converted mine a decade ago into a Borbely Milenium, put 700,000 uf in a CLC filter into the chassis and run it at 50 watts Class A for a decade now - very happy with the sound and haven't lost a semiconductor in 10 years.
I sent Krell another email. Bottom line is that I did not pay $1100 to have them turn my amp into a 1.8 watt per channel amp. That means that at some point it turns into a class ab, NOT WHAT I HAVE A KRELL FOR. I bought this based on the reports that it was a true class a THROUGHOUT its full operating range of 100 watts. I'm quite upset by all of this as I wanted the amp brought back to "original" specs. I did not have the original box and had to drive it into Krell, a 700 mile round trip from my house. Why would they take a pure class a amp and change it from Dan's original design? I feel I'm not hearing what it is capable of. What if I change speakers to a tough load like Apogees? I wanted an amp that would drive anything and this was it. Now Krell tells me that the higher bias would increase distortion. So are they saying that Dan's original design had a lot of distortion in it???? Damn, I hate when options aren't explained in full to the customer. Damn..
They lied to me then. They stated that it was not in class AB, as that would be 7Mv at idle.
If the OP does not want to open up the amp, would it help to use
something like this:
Kill A Watt - Wikipedia, the free encyclopedia
to measure the power draw at idle and load?
something like this:
Kill A Watt - Wikipedia, the free encyclopedia
to measure the power draw at idle and load?
I really don't see the reason to do that when Krell themselves stated that they set the idle bias to 85 mV, combined with the fact that the internal fans are blowing cold air out of the amp after it has driven my speakers for over 6 hours. No, I do believe Krell that they turned the bias WAY down. The BIG question is why. According to Krell, the use of newer replacement parts (transistors) being more efficient and LESS distortion with the bias turned down. So, it appears they have turned the amp into a nice Class AB design. It sounds good, but the question remains as to why Dan D'Agostino design the original KSA 100 to run in full Class A with all 100 watts, not just the first 1.8 as they now have my amp set to. Was there a lot of distortion with the original. I don't think so based on the reviews. The heat of the original design was a by product of what Dan achieved in his design, and that was pure Class A with ALL 100 watts. If he could have the same sound with less heat, I have to believe he would have done so. No need then for fans or large heat sinks. But he didn't do that and that's what is bothering me. Unless Krell has come up with a way to get the same sound without biasing in Full Class A. That's what they tell me is happening, but they won't explain it. The last email I received from Krell was somewhat curt. They stated " If you want a toaster with high distortion, bring the amp back in and will turn up the bias." So am I to think that they have found a way to make this classic amp sound better? Do all of their new amps run cold? Yes, it does currently sound good, but then again I don't have an original KSA 100 with the original bias to compare it to. I just want answers that make sense, not "Don't worry about it. We are Krell and we know what we're doing."
Is it possible to produce a sliding Class A amp ? One that is set at idle to produce a few watts of pure Class A but also self adjusts its bias to remain in Class A all the way up to P Max. The common thinking is that a pure Class A amp is ready for anything and is biassed accordingly might be a thing of the past.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.