No-one has the same amp with the same new output devices and the same ears as you. Try raising the bias and see what YOU hear.
I got tired of rebuilding/mofiying (cascode input, DC servo, regulated driver supply etc) the KSA 100 and I converted mine into a Borbely Millenium biased at 50 watts Class A - the output Fets Run at 50 C and have lasted a decade.
I got tired of rebuilding/mofiying (cascode input, DC servo, regulated driver supply etc) the KSA 100 and I converted mine into a Borbely Millenium biased at 50 watts Class A - the output Fets Run at 50 C and have lasted a decade.
I don't need answers on the same amp, although that would be nice. Any amp. What have participants of this forum heard when rebiasing ANY amp. You make it sound as if raising and lowering the bias is easy. Maybe for you. I have never done it, so I would need to A: Take this 100 lb amp to a tech, OR B: Buy the proper equipment and get explicit instructions so that I do not damage the amp or myself.
Therefore, before doing any of that, I wanted to hear from others what they heard when they re-biased ANY amp. Does biasing an amp act like a tone control of some sort? Does it increase or decrease the distortion? What is noticeable is what I'm after. Aside from getting the specifications from the manufacturer, what does one listen for to determine correct bias? These are things I want to hear. I figured I would have my best shot at these answers on a Do it Yourself site like this.
No, altering bias does not act like a tone control. It does alter the distortion, both the raw numbers and the 'distortion profile', that is to say the relative amplitudes of 2nd, 3rd, 4th and so on harmonics. Is it noticeable... I dare say it is on some amps, less so on others. You can't 'listen' for correct bias, a design will have an optimal bias for minimum distortion that should ideally be adjusted with a spectrum analyser... but no diy'er does that.
Most amps will 'sound the same' over a huge range of bias... you only need enough to overcome the obvious crossover point and beyond that the non linearity's disappear rapidly. That bias current could be as low 5 milliamps, above which you might hear no difference, no matter how high you go. That is very amp specific though.
Lets try this with a real amp in a simulator.
A Class AB amp correctly biased (that's only 100ma or so bias) can deliver say 0.0005% distortion at 50 watts RMS/8 ohm. Increase the bias to 2.6 amps (one channel of the amp is now dissipating 200 watts of heat compared to around 8 watts in Class AB) The distortion has now fallen to 0.000268. Both are incredibly small numbers.
Now look at the pictures showing the actual harmonics that make up those numbers. Class AB has the characteristic 'spray' of harmonics that the pure Class A avoids.
Can you tell the difference by listening though.
Most amps will 'sound the same' over a huge range of bias... you only need enough to overcome the obvious crossover point and beyond that the non linearity's disappear rapidly. That bias current could be as low 5 milliamps, above which you might hear no difference, no matter how high you go. That is very amp specific though.
Lets try this with a real amp in a simulator.
A Class AB amp correctly biased (that's only 100ma or so bias) can deliver say 0.0005% distortion at 50 watts RMS/8 ohm. Increase the bias to 2.6 amps (one channel of the amp is now dissipating 200 watts of heat compared to around 8 watts in Class AB) The distortion has now fallen to 0.000268. Both are incredibly small numbers.
Now look at the pictures showing the actual harmonics that make up those numbers. Class AB has the characteristic 'spray' of harmonics that the pure Class A avoids.
Can you tell the difference by listening though.
Attachments
Samman11, why were the original output transistors changed? I can not see why they would do that without asking you first. Were any of them blown?
One channel running a little hotter than the other was probably caused by the bias on that channel drifting over time, most likely by components in the bias circuit changing value due to aging.
One channel running a little hotter than the other was probably caused by the bias on that channel drifting over time, most likely by components in the bias circuit changing value due to aging.
Last edited:
All Krell stated was that they were needed. Maybe they change those with all the upgrades. Not cheap. As you can see on the list I posted earlier in the thread, 16 transistors were changed at a price of $300. I believe the cost for each transistor was $18.90 X 16.
They did tell me that these new transistors are quite fast and that's the reason for the ability to run the bias down to 85 mV with the lowest distortion possible. I must say, IT DOES sound fantastic, but from what I have gathered, when others have refurbished this amp, they tend to set the bias at about 230 - 250mV. Dan D'Agostino told me to set it at 650! Another told me that would run the amp way too hot. Maybe Dan isn't aware of the capabilities of the new transistors that Krell is using. I don't know, but I do know that I have never been more confused when the original designer states one figure, Krell states another, and other techs fall somewhere in between in terms of Bias.
Samman11
The KSA100 mk1 is a collector's piece and can only increase in value.
Without an expert to assist you I suggest you leave the amp alone. The risk is not worth it.
Rather have a amp you can enjoy for a long period, The amp is close to 40 years old- so be gentle on it.
Jozua
Cape Town
The KSA100 mk1 is a collector's piece and can only increase in value.
Without an expert to assist you I suggest you leave the amp alone. The risk is not worth it.
Rather have a amp you can enjoy for a long period, The amp is close to 40 years old- so be gentle on it.
Jozua
Cape Town
Don't take this out of context. Why do you believe all KSA owners should be aware of this? Were the transistors supposed to last forever? Krell did tell me that these are the transistors that they use in all their current products, so maybe while the amp was there they figured it would be worth doing. I wouldn't say that they were taking advantage of me. They gave me a complete tour of the facility and were quite professional with me. In fact, for the longest time when I wanted to bring in the amp for an upgrade, they kept telling me "if it ain't broke, don't fix it." It was only after the heating differences from one side to the other that they decided, based on age, the bias may be off due to parts getting old. At that point, it was worth a look. The only issue that came up was with this bias setting, and they stand by the setting as being the one with the lowest distortion. I wouldn't imply, however, any shadiness on their part.
The reason they used Class A biasing on the original transistors was to bring down the crossover distortion and it sounded great. They are probably right when they say that with these new transistors you can run it with largely lower biasing and still achieve probably better distortion levels than the originals. In that sense I would also suggest you to leave the amp as it is, as long as the job has been done, and it sounds great. What I am saying is that this is informative to other Krell Ksa 100 mk1 owners. If it is needed should be up to these owners to decide.
Last edited:
So, do transistors age or can they have been kept in the unit? Of course, if it allows the amp to put out less heat at a lower bias, that also may be a plus in terms of longevity on the other parts as well.
All electronics can fail. The failure rate increases as time goes by and heat is the enemy of all electronics. This said, if I was given the choice I would have kept it original as long as it lasted. Renewing the thermal contact to the heat sinks would have been a good idea. A good chance of keeping it alive for 40 more years. You never know.
Last edited:
Are the 16 SIL-PAD T0-3 items listed above in what Krell replaced the "Thermal Contacts" to the heat sinks that you speak of?
Also, I would assume that there were several reasons for Krell to replace the Transistors. As you stated, everything ages. This amp was a low serial number (237), indicating it was from the first batch that Dan made. Since it was in the Krell shop, why take a chance that those items may fail soon when Krell could replace them then and and there. I drove the amp over 700 hundred miles round trip (no original box) and wanted everything brought back to original spec. The other reason, according to Krell, was that the new transistors are better and faster, thereby allowing the amp to run cooler with improved performance.
I would just leave it and enjoy it, you said yourself it sounds fantastic. Forget about it and get back to enjoying musicDon't risk a potentially damaging and expensive adjustment.
The sil pads are the thermal contacts dacen was meaning.
Ahh, that's what I thought the Sil Pads were for. Yes, I agree with you in regards to leaving it alone and enjoying it. But it has been a great learning experience to discuss Bias settings and what that is all about. No learning can take place unless questions are asked. This is a great forum for that purpose.
It seems Ironic that Krell in the early days used to mock other amplifier makes for advertising class A operation when in reality only the first few watts were class A, Krell rightly insisting that their amps were in class A up to rated output.
It seems these days they are happy to play that game themselves.
I echo the advice that if it sounds as good as you hoped it would then be happy , it may well sound a little better biased further into class A but presumably you'd lose the warranty on the work that has been carried out if you altered it after it had left Krell's factory.
In a trawl through my loft I've found 2 Avel Lindberg toroidal transformers that are identical to the one's found in the KSA100, they were owned by someone who had connections with the Avel factory in the UK and I bought them off him about 20 years ago, any ideas as to their
value ?
It seems these days they are happy to play that game themselves.
I echo the advice that if it sounds as good as you hoped it would then be happy , it may well sound a little better biased further into class A but presumably you'd lose the warranty on the work that has been carried out if you altered it after it had left Krell's factory.
In a trawl through my loft I've found 2 Avel Lindberg toroidal transformers that are identical to the one's found in the KSA100, they were owned by someone who had connections with the Avel factory in the UK and I bought them off him about 20 years ago, any ideas as to their
value ?
Hi Samman,
Can you tell us what's the final conclusion about your serviced Krell KSA-100?
Is it still Class A up to its 100 Watt rating or did they indeed modify it into a AB one?
Did you ever measure something like the millivolts over the 1 Ohm emitter resistors?
You could even place a energymeter in front in your AC outlet and then see how many Watts the amp draws when idling.
Let's finalize this matter!
Can you tell us what's the final conclusion about your serviced Krell KSA-100?
Is it still Class A up to its 100 Watt rating or did they indeed modify it into a AB one?
Did you ever measure something like the millivolts over the 1 Ohm emitter resistors?
You could even place a energymeter in front in your AC outlet and then see how many Watts the amp draws when idling.
Let's finalize this matter!
Hello there. Well, it's been two years since I posted here. No, I never did measure the millivolts over the 1 ohm emitter resistor. I'm not a tech in any sense of the word. I do like the idea of the "energymeter", since that appears to be something any layman like myself could utilize. Where do I get one of these? How much are they?
I would have to say that based on the heat, or lack of it, coming from the amp, Krell must have re-biased it to class AB. The sound is great, and Krell states that they biased it for the lowest possible distortion. They stated they were able to bias it lower with the newer and updated parts used during the refurbishing. One thing is certain, the sound is the best I've heard from my system.
Hi Samman,
You can buy it everywhere.
It's main purpose is to monitor your energyconsumption so that your 'environment awarness' heightens or something.
I always used them for monitoring the Class A consumption of my amps.
1. If the intake stays steady then you're still playing music in Class A.
2. If the manufacturer specifies a certain AC consumption, Krell KSA-100 should consume around 8 Amps at idle (according to the Krell KSA-100 owner's manual) meaning around 960 Watts at a 120 Volts AC line.
So with an energymeter you can dial the biastrimmer to that value.
Personally I think that's to much (be aware that other components than the outputtransistors) draw current also so you can't conclude that the 8 Amps are a starting point for calculating your Class A output.
I'm buying a original Krell KSA-100 Mk2 tomorrow for my Watkins woofers and will check it's specs for you so probably at tuesday so when there should be .640 Volts over the 1 Ohm emitters as Dan D' Agostino said, the I will give you the other figures as a reference.
If your sounds okay to you I wouldn't worry to much.
I'll ask that guy at Krell what exactly is now the procedure for a refurbishing of a Krell KSA-100, what kind of outputtransistors they use and why they lower the bias from 640 mV to 87 mV.
http://www.diyaudio.com/forums/attachment.php?attachmentid=594999&stc=1&d=1485512325
You can buy it everywhere.
It's main purpose is to monitor your energyconsumption so that your 'environment awarness' heightens or something.
I always used them for monitoring the Class A consumption of my amps.
1. If the intake stays steady then you're still playing music in Class A.
2. If the manufacturer specifies a certain AC consumption, Krell KSA-100 should consume around 8 Amps at idle (according to the Krell KSA-100 owner's manual) meaning around 960 Watts at a 120 Volts AC line.
So with an energymeter you can dial the biastrimmer to that value.
Personally I think that's to much (be aware that other components than the outputtransistors) draw current also so you can't conclude that the 8 Amps are a starting point for calculating your Class A output.
I'm buying a original Krell KSA-100 Mk2 tomorrow for my Watkins woofers and will check it's specs for you so probably at tuesday so when there should be .640 Volts over the 1 Ohm emitters as Dan D' Agostino said, the I will give you the other figures as a reference.
If your sounds okay to you I wouldn't worry to much.
I'll ask that guy at Krell what exactly is now the procedure for a refurbishing of a Krell KSA-100, what kind of outputtransistors they use and why they lower the bias from 640 mV to 87 mV.
http://www.diyaudio.com/forums/attachment.php?attachmentid=594999&stc=1&d=1485512325
Last edited:
I looked up their latest amps on their website. The "Duo 125" for example has an idle power dissipation of just under 60W ( they quote it in BTU/hr which is understandable if you are a heating installer but daft otherwise). From what I can glean your KSA-100 is now idling a little higher than that. So at least you can have peace of mind that your amp is no less class A than the latest models.
Technically, it is not the class A itself that makes Krells sound good. That's just how the original Krells were implemented with the technology and know-how at the time. Also a good marketing device. If the Krell factory tells you it is better now and it sounds better then who are you or we to argue with them?
But I take your point if what you actually wanted was a restored KSA-100 with original parts for the sake of authenticity. It sounds like the Krell folk thought you wanted it brought up to a higher sonic standard and be less wasteful of energy to boot.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.