K-Amps, Thanks for doing those calcs! Both you and Stuart have been a big help and you've definately saved this set of output devices! You are right BUT I was ALSO running this thing at +/-43 volt rails, not the +/- 36 volt rails that you thought I was. Could you crunch that one again for me? Jeez, I think was in the KSA-100 realm there for a while but I have now lowered my rails to the normal +/- 36 volts. Your calcs are just about right on the money too! I am measuring 1.85 amps for 400 mv across the .68 ohm emiter resistors. Sink temp is down and device temp is down under 50 deg c. Now I feel better about this whole thing..... BTW: At the normal 630 mv and 36 volt rails as per the old Krell schematic from that other thread I was drawing slightly over 2.8 amps per rail! So is this new 400 mv setting because it is split up between three devices per side and not the normal two as was used?
Thanks!
Mark
Thanks!
Mark
Here u go Mark with 43 rails and also with regular KSA-50 (per your numbers).. Sorry guys about so much data being posted...
Class-A output Calculation: Krell KSA-50 Mk. II(43 rails and 520mV)
Inputs
Number of output devices: 3 Pairs
Voltage rails (per rail): 43 volts
Emitter resistance (per device): 0.68 ohms
Bias voltage per Emmiter resistor: 520 mv
Idle bias per device 764.7 mA
Speaker ohms 8 ohms
Results
Idle bias per device: 0.765 Amps
Total Amplifier bias (per rail)** 2.29 Amps
Total Amplifier bias (both rails) 4.59 Amps
Total Dissipation (per channel) 197.3 Watts
Dissipation per device pair at idle 32.9 watts
Class-A output: Peak 168.4 Watts peak
Class-A output: RMS 84.2 Watts RMS
Efficiency 42.68 %
Here with 400mV
Class-A output Calculation: Krell KSA-50 Mk. II
Inputs
Number of output devices: 3 Pairs
Voltage rails (per rail): 36 volts
Emitter resistance (per device): 0.68 ohms
Bias voltage per Emmiter resistor: 400 mv
Idle bias per device 588.2 mA
Speaker ohms 8 ohms
Results
Idle bias per device: 0.588 Amps
Total Amplifier bias (per rail)** 1.76 Amps
Total Amplifier bias (both rails) 3.53 Amps
Total Dissipation (per channel) 127.1 Watts
Dissipation per device pair at idle 21.2 watts
Class-A output: Peak 99.7 Watts peak
Class-A output: RMS 49.8 Watts RMS
Efficiency 39.22 %
Max Class-AB RMS (1 kVA Toroid p/CH 5% reg) 73.70 Watts RMS
NORMAL KSA-50 Per Mark's numbers:
Class-A output Calculation: Krell KSA-50 Mk. II
Inputs
Number of output devices: 2 Pairs
Voltage rails (per rail): 36 volts
Emitter resistance (per device): 0.68 ohms
Bias voltage per Emmiter resistor: 630 mv
Idle bias per device 926.5 mA
Speaker ohms 8 ohms
Results
Idle bias per device: 0.926 Amps
Total Amplifier bias (per rail)** 1.85 Amps
Total Amplifier bias (both rails) 3.71 Amps
Total Dissipation (per channel) 133.4 Watts
Dissipation per device pair at idle 33.4 watts
Class-A output: Peak 109.9 Watts peak
Class-A output: RMS 54.9 Watts RMS
Efficiency 41.18 %
Max Class-AB RMS (1 kVA Toroid p/CH 5% reg) 73.70 Watts RMS
Class-A output Calculation: Krell KSA-50 Mk. II(43 rails and 520mV)
Inputs
Number of output devices: 3 Pairs
Voltage rails (per rail): 43 volts
Emitter resistance (per device): 0.68 ohms
Bias voltage per Emmiter resistor: 520 mv
Idle bias per device 764.7 mA
Speaker ohms 8 ohms
Results
Idle bias per device: 0.765 Amps
Total Amplifier bias (per rail)** 2.29 Amps
Total Amplifier bias (both rails) 4.59 Amps
Total Dissipation (per channel) 197.3 Watts
Dissipation per device pair at idle 32.9 watts
Class-A output: Peak 168.4 Watts peak
Class-A output: RMS 84.2 Watts RMS
Efficiency 42.68 %
Here with 400mV
Class-A output Calculation: Krell KSA-50 Mk. II
Inputs
Number of output devices: 3 Pairs
Voltage rails (per rail): 36 volts
Emitter resistance (per device): 0.68 ohms
Bias voltage per Emmiter resistor: 400 mv
Idle bias per device 588.2 mA
Speaker ohms 8 ohms
Results
Idle bias per device: 0.588 Amps
Total Amplifier bias (per rail)** 1.76 Amps
Total Amplifier bias (both rails) 3.53 Amps
Total Dissipation (per channel) 127.1 Watts
Dissipation per device pair at idle 21.2 watts
Class-A output: Peak 99.7 Watts peak
Class-A output: RMS 49.8 Watts RMS
Efficiency 39.22 %
Max Class-AB RMS (1 kVA Toroid p/CH 5% reg) 73.70 Watts RMS
NORMAL KSA-50 Per Mark's numbers:
Class-A output Calculation: Krell KSA-50 Mk. II
Inputs
Number of output devices: 2 Pairs
Voltage rails (per rail): 36 volts
Emitter resistance (per device): 0.68 ohms
Bias voltage per Emmiter resistor: 630 mv
Idle bias per device 926.5 mA
Speaker ohms 8 ohms
Results
Idle bias per device: 0.926 Amps
Total Amplifier bias (per rail)** 1.85 Amps
Total Amplifier bias (both rails) 3.71 Amps
Total Dissipation (per channel) 133.4 Watts
Dissipation per device pair at idle 33.4 watts
Class-A output: Peak 109.9 Watts peak
Class-A output: RMS 54.9 Watts RMS
Efficiency 41.18 %
Max Class-AB RMS (1 kVA Toroid p/CH 5% reg) 73.70 Watts RMS
WOW! That was really fast! And I think these calcs will help others as well.... sorry I'm not an engineer type. I'm set up right now like your second set of calcs and the slo-speed fan has no problem keeping the sinks at about a reasonable 45 deg c. temperature. Seeing those calcs has made things very clear to me now. The amp also sounds very, very good set up this way....
Thanks again!
Mark
P.S. another thing I have noticed is that the driver stage temperature and the output stage as measured on the device cases appear to track each other really well showing that the bias transistor mounted on the driver stage hs works really well.
Thanks again!
Mark
P.S. another thing I have noticed is that the driver stage temperature and the output stage as measured on the device cases appear to track each other really well showing that the bias transistor mounted on the driver stage hs works really well.
output device current sharing...
...is pretty good, basically because we are using relatively large emitter resistors.
But if you keep total idle current the same and add more outputs, the voltage on each emitter resistor decreases proportionally, ie twice as many transistors gives half the voltage...
If it gets too low, like below 0.2v, the adjustment will become twitchy and the sharing between unmatched transistors will degrade. At that point you'd probably want to increase the value of the emitter resistors to the next value, ie from 0.5 to 0.68, or from 0.68 to 1.0, perhaps up to 1.2 or 1.5 if you are seriously increasing the number of outputs... like 8 or more pairs.
IIRC krell used 1ohm on the ksa100, not sure what the 200 had, but it would be interesting to find out, they must have had a lot of transistors in that beast...
Stuart
...is pretty good, basically because we are using relatively large emitter resistors.
But if you keep total idle current the same and add more outputs, the voltage on each emitter resistor decreases proportionally, ie twice as many transistors gives half the voltage...
If it gets too low, like below 0.2v, the adjustment will become twitchy and the sharing between unmatched transistors will degrade. At that point you'd probably want to increase the value of the emitter resistors to the next value, ie from 0.5 to 0.68, or from 0.68 to 1.0, perhaps up to 1.2 or 1.5 if you are seriously increasing the number of outputs... like 8 or more pairs.
IIRC krell used 1ohm on the ksa100, not sure what the 200 had, but it would be interesting to find out, they must have had a lot of transistors in that beast...
Stuart
I also had a KSA-250 that I sold a couple of months ago ;-) It had 1 ohm emitters on all 12 pairs of metal cans. I measured 111mV across the 1ohm emmiters at idle... according to my calcs thats about 28.4 watts pure class-A.... umm it was supposed to be 250 watts class-A.... I guess they used an auto bias scheme.
Mark not a worry, I have a spread sheet and all I do is plug in 3-4 values
cheers!
BTW, Stuart is the expert here, I just try to be a good understudy as my business major did not teach me how to make amps ;-).
PPS: Of the 3 Krells I owned (KSA-100mk.ii, KSA-250 and a KAV-250) I like te KAV-250 THE best by far.... I cannot find the sound wanting in anyway... yes I know it is class-AB but still, the front end is pretty good. Not selling my KAV-250.
Mark not a worry, I have a spread sheet and all I do is plug in 3-4 values
cheers!
BTW, Stuart is the expert here, I just try to be a good understudy as my business major did not teach me how to make amps ;-).
PPS: Of the 3 Krells I owned (KSA-100mk.ii, KSA-250 and a KAV-250) I like te KAV-250 THE best by far.... I cannot find the sound wanting in anyway... yes I know it is class-AB but still, the front end is pretty good. Not selling my KAV-250.
Transformers
One last thing - I've found that big transformers (when loaded with lots of current demand from class-A output stage) tend to buzz. This is mechanical hum, and comes from saturated core (there's always little bit of residual DC in the mains).
I had to put in the mains DC filter (large electrolytic caps back to back in series with primary side, with safety diodes across). Leave some room in your boxes. I had to scramble things in mine because I made everything too tight.
And don't forget the inrush supressor. You will need it - trust me.
Bratislav
One last thing - I've found that big transformers (when loaded with lots of current demand from class-A output stage) tend to buzz. This is mechanical hum, and comes from saturated core (there's always little bit of residual DC in the mains).
I had to put in the mains DC filter (large electrolytic caps back to back in series with primary side, with safety diodes across). Leave some room in your boxes. I had to scramble things in mine because I made everything too tight.
And don't forget the inrush supressor. You will need it - trust me.
Bratislav
tmblack said:
There's not that much current at the primary side (especially with 240V we use in Australia). You need > 16V , > 4700uF, high ripple current caps. Search this forum, there's lots of info already available. I also used common 35A rectifier as protecive diodes across the caps (less likely to fail than a diode).
Bratislav
I read the discussion about railvoltages, bias currents and heat dissipation, and now i am a little worryd about my trafo's.
I use per channel a 2 x 30V toroid 800 VA. That must give a railvoltage of about 43 V. Do i have to worry and look for toroids with a lower voltage, e.g. 2 x 25 V?
I hope not.
I use per channel a 2 x 30V toroid 800 VA. That must give a railvoltage of about 43 V. Do i have to worry and look for toroids with a lower voltage, e.g. 2 x 25 V?
I hope not.
rails
Wim,
The only concern is whether you have big enough heatsinks. Nothing wrong with higher rails, at the same idle current they just require more heat to be dissipated. You can also reduce the idle current to compensate, start at the low end of the idle range and increase it once an hour until you have a temp you can live with. If you draw a couple of amps from the trafos, the voltage will drop some, if you use a couple of big inductors, it will drop more. If you are really worried about it, you can always use a CRC power supply, make R a couple of ohms and you'll get rid of 4volts real fast...
Stuart
Wim,
The only concern is whether you have big enough heatsinks. Nothing wrong with higher rails, at the same idle current they just require more heat to be dissipated. You can also reduce the idle current to compensate, start at the low end of the idle range and increase it once an hour until you have a temp you can live with. If you draw a couple of amps from the trafos, the voltage will drop some, if you use a couple of big inductors, it will drop more. If you are really worried about it, you can always use a CRC power supply, make R a couple of ohms and you'll get rid of 4volts real fast...
Stuart
Finally a proper test
Well, i've had it hooked up and running for quite a while..... about 2 hours now... one set of outputs, about 70mv bias (not much, but it's only 1 pair of outputs and they are on a 2.7c/w heatsink!!!)...
first things, i fixed any "noise" i had while running by adding a star ground, it solved the problem 100%, it's almost DEAD SILENT now when it's not making music... there is an ever so slight 50hz hum, but that will be fixed with proper wiring and less leads running everywhere....
One problem: Maybe someone can answer why - i had it running stable at about 75mv bias (0.68ohm resistors) - however when plugged in the rca connection (obviously connected to the board) into my peavey amp (it has a throughput - peavey amp is off all the time while this is plugged in so it's like connecting it straight to the pre-amp), the bias drops to 6.6mv.... i can then crank it all the way back up again to 70/75mv whatever, but once i d/c the rca input it skyrockets (bias) to 140/150mv depending on where i left it... Why is this? Even with the star ground, without it it does the same... Is there something i missed? The outputs are wired in the most simple way - nothing but wire and emittor resistors... Any ideas?
As for stability - i don't have a scope - but it's stable, no oscillations that i can tell... i added little bits of aluminium to the pre-drivers, it is all they need, they stay MUCH cooler now (less than 50deg)... considering it's 25ambient here, that's plenty of room...
I'm running 43v rails too and it seems ok, i might just stick to 28vac toroids...
Now - important part - sound!
It sounds excellent to my ear - from what i've heard so far, it's excellent, crystal clear and very smooooth... The bass especially, it doesn't sound 'sharp' and 'rough' it sounds very nice... quite authoritative, but still very very clear... sounds different to my old peavey!!! (which will be on the sub once this proj is done!)...
Just thought i'd drop in some thoughts and see what people think...
As a side note - is everyone using output protection? the KSA-50 had none, AFAIK it had no relays, etc on the outputs either did it?
Aaron
Well, i've had it hooked up and running for quite a while..... about 2 hours now... one set of outputs, about 70mv bias (not much, but it's only 1 pair of outputs and they are on a 2.7c/w heatsink!!!)...
first things, i fixed any "noise" i had while running by adding a star ground, it solved the problem 100%, it's almost DEAD SILENT now when it's not making music... there is an ever so slight 50hz hum, but that will be fixed with proper wiring and less leads running everywhere....
One problem: Maybe someone can answer why - i had it running stable at about 75mv bias (0.68ohm resistors) - however when plugged in the rca connection (obviously connected to the board) into my peavey amp (it has a throughput - peavey amp is off all the time while this is plugged in so it's like connecting it straight to the pre-amp), the bias drops to 6.6mv.... i can then crank it all the way back up again to 70/75mv whatever, but once i d/c the rca input it skyrockets (bias) to 140/150mv depending on where i left it... Why is this? Even with the star ground, without it it does the same... Is there something i missed? The outputs are wired in the most simple way - nothing but wire and emittor resistors... Any ideas?
As for stability - i don't have a scope - but it's stable, no oscillations that i can tell... i added little bits of aluminium to the pre-drivers, it is all they need, they stay MUCH cooler now (less than 50deg)... considering it's 25ambient here, that's plenty of room...
I'm running 43v rails too and it seems ok, i might just stick to 28vac toroids...
Now - important part - sound!
It sounds excellent to my ear - from what i've heard so far, it's excellent, crystal clear and very smooooth... The bass especially, it doesn't sound 'sharp' and 'rough' it sounds very nice... quite authoritative, but still very very clear... sounds different to my old peavey!!! (which will be on the sub once this proj is done!)...
Just thought i'd drop in some thoughts and see what people think...
As a side note - is everyone using output protection? the KSA-50 had none, AFAIK it had no relays, etc on the outputs either did it?
Aaron
Never mind about the dc offset problem, all i needed was a star ground!!! It fixed everything (a proper star ground)...
Ok, i agree now, the mounting that is needed was definately left off the boards!! I just used the blue terminal blocks, works 100%!
Now i'm happy, the amp has stabilised and with the star ground the bias isn't going wacko when i unplug the rca!
I've settled at 60mv bias for now (from memory) as the heatsink idles @ 50deg... That's hot enough!
Not much class-a power, but i can tell when it leaves class A, the temp on the heatsink rises, notably... considering it sits a solid 50deg, when it hits 51/52 you know you are in class a/b...
I'll post up a photo in a minute to show my "setup"
Ok, i agree now, the mounting that is needed was definately left off the boards!! I just used the blue terminal blocks, works 100%!
Now i'm happy, the amp has stabilised and with the star ground the bias isn't going wacko when i unplug the rca!
I've settled at 60mv bias for now (from memory) as the heatsink idles @ 50deg... That's hot enough!
Not much class-a power, but i can tell when it leaves class A, the temp on the heatsink rises, notably... considering it sits a solid 50deg, when it hits 51/52 you know you are in class a/b...
I'll post up a photo in a minute to show my "setup"
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
*and yes, 3 multimeters are much better than 1!!! I must admit, having 3 multimeters has proven completely invaluble to me on sooooooo many occasions!