rabstg said:
I got them from
www.conradheatsinks.com
They are here in Australia...
I'm building 2 amps, a 3 channel and a 2 channel, the 2 channel is aircooled, so it has 2 of these heatsinks per channel, the other amp (3 channel) has one per channel and will be fan forced... Be aware i'm going WELL over the original Krell specs and will be running 3.3amp maybe more of bias... That's why i need fan forced
They are pretty heavy sinks also, but they are very very well suited to the job As for the caps, i used a BC components 630v 47pf polypropolene (that's the blue one standing upright behind the first set of drivers)... I'm hoping 47pf is ok, as the 33pf was $4.50 each! I didn't want to spend that much! I think they were silver mica though (as in the original Krells)...
I was not sure if this capacitor will suit - you'll have to power your's up to see if it sounds ok or not! I can't power mine up for proper testing till i have the right toroids and i also need my heatsinks drilled, etc so that parts a while off... That and i have another 2 boards to build!
Aaron
Powered the board alone up...
It did nothing
Mind you, i was running it a standard 12v "lab" supply, just DC with -ve (well, 0v) to the -ve connection and 12v to the other side +ve... I'm not really sure if i would get anything, pretty much a 6-0-6 (???) from what i think... So i didn't think i would get much, i had 3ma (!!) current drain in total, so not enough to do anything...I have no outputs on it either... just the board plain... I wanted to make sure it's not shorted or anything... I don't have a power supply lying around i can think of....
Actually (!) i might!!! I might go looking...
I don't think i do, the only one i can think of is a 34-0-34v toroid (140va!) but i don't think i kept it
It did nothing
Mind you, i was running it a standard 12v "lab" supply, just DC with -ve (well, 0v) to the -ve connection and 12v to the other side +ve... I'm not really sure if i would get anything, pretty much a 6-0-6 (???) from what i think... So i didn't think i would get much, i had 3ma (!!) current drain in total, so not enough to do anything...I have no outputs on it either... just the board plain... I wanted to make sure it's not shorted or anything... I don't have a power supply lying around i can think of....
Actually (!) i might!!! I might go looking...
I don't think i do, the only one i can think of is a 34-0-34v toroid (140va!) but i don't think i kept it
6-0-6...
...isnt going to get the front end biased up, nor the vbe multiplier, so using it to check for shorts is an excellent idea, but there is no way you are going to get music...
You will need ~+/-30v for the front end to get the zeners conducting, though the current will be quite low until you approach the real voltage, the output needs less, perhaps +/- 10v, so if you had a low current (100ma) hi-voltage, and a high current low voltage psu it should get you to the point where you have a music making endeavour.
Stuart
...isnt going to get the front end biased up, nor the vbe multiplier, so using it to check for shorts is an excellent idea, but there is no way you are going to get music...
You will need ~+/-30v for the front end to get the zeners conducting, though the current will be quite low until you approach the real voltage, the output needs less, perhaps +/- 10v, so if you had a low current (100ma) hi-voltage, and a high current low voltage psu it should get you to the point where you have a music making endeavour.
Stuart
wim said:
I couldn't!!
I used metal film 3w (i think they were metal film), 27 ohm ones...
Got them from www.farnell.com
That was all i could find...
Aaron
A nice little sub amp that i had - that has lay dormant for some time - tonight sacrificed itself
I found it had a 28-0-28vac toroid, 330va! Perfect... Almost, at least for testing it's excellent.
I've only done an extremely basic p/s, with rect and caps... straight to board, there's no outputs hooked up at all...
Ran some tests to see how it goes...
ZD101/102 get to about 50deg after about 5mins run time @ 22deg ambient..
Q107/108 hits about 65deg after this time - is that a "safe" running temp for the "pre" drivers? I'm a bit dubious about heatsinking - mainly for space issues - they were reasonably stable after that time, not gaining much in temp... With no outputs i had about 0.4v across the emitter resistors on the drivers... The bias wouldn't turn any lower than that at all (assuming that's what i was adjusting..!)... it didn't seem to affect the temp of the drivers, the drivers heatsink (seen in prev photo's) work very very well, they are pretty much slightly warm...
R103/107 get warm, this may be due to the fact they are right hard against the diodes... (i used 3w diodes btw), there's not much space, they get to about 48deg...
Nothing seemed abnormal in any form... Next step, make up some way of mounting outputs and test it!
Now, some questions for the more talented folk around here
1. Is 65 deg too hot for the pre drivers? (Q107/108)... I don't mind (or care) if they are hot, only if it is a PROBLEM will i heatsink them...Should i?
2. Bias - does that measurement i got sound about right 0.4v across the driver emitter resistors (it was spot on the same for both resistors)
3. DC Offset - how come DC offset adjustment does absolutely nothing whatsoever? I'm guessing because i don't have any outputs - which also lead to the problem of - how do i measure the dc offset with no outputs? And does dc offset contribute to heat problems?
If people could help out with some of the above Would be excellent!!
Aaron
P.S. my 28vac with no load (well, the board alone) turns into about 45vdc (!) which is much higher than i thought... maybe i just need more load!
P.P.S. - i am making the assumption that a safe running temp for all other small transistors/diodes on the board is sub 50deg... they have a rise of approx 25-28deg above amibient, most of the time from what i gather...
I found it had a 28-0-28vac toroid, 330va! Perfect... Almost, at least for testing it's excellent.
I've only done an extremely basic p/s, with rect and caps... straight to board, there's no outputs hooked up at all...
Ran some tests to see how it goes...
ZD101/102 get to about 50deg after about 5mins run time @ 22deg ambient..
Q107/108 hits about 65deg after this time - is that a "safe" running temp for the "pre" drivers? I'm a bit dubious about heatsinking - mainly for space issues - they were reasonably stable after that time, not gaining much in temp... With no outputs i had about 0.4v across the emitter resistors on the drivers... The bias wouldn't turn any lower than that at all (assuming that's what i was adjusting..!)... it didn't seem to affect the temp of the drivers, the drivers heatsink (seen in prev photo's) work very very well, they are pretty much slightly warm...
R103/107 get warm, this may be due to the fact they are right hard against the diodes... (i used 3w diodes btw), there's not much space, they get to about 48deg...
Nothing seemed abnormal in any form... Next step, make up some way of mounting outputs and test it!
Now, some questions for the more talented folk around here
1. Is 65 deg too hot for the pre drivers? (Q107/108)... I don't mind (or care) if they are hot, only if it is a PROBLEM will i heatsink them...Should i?
2. Bias - does that measurement i got sound about right 0.4v across the driver emitter resistors (it was spot on the same for both resistors)
3. DC Offset - how come DC offset adjustment does absolutely nothing whatsoever? I'm guessing because i don't have any outputs - which also lead to the problem of - how do i measure the dc offset with no outputs? And does dc offset contribute to heat problems?
If people could help out with some of the above
Would be excellent!!Aaron
P.S. my 28vac with no load (well, the board alone) turns into about 45vdc (!) which is much higher than i thought... maybe i just need more load!
P.P.S. - i am making the assumption that a safe running temp for all other small transistors/diodes on the board is sub 50deg... they have a rise of approx 25-28deg above amibient, most of the time from what i gather...
my 2c...
Aaron,
65c is not too hot, but it is likely that the temp will be much hotter once the boards etc are encased. I'd use a heatsink, some people think transistors sound better hot, in which case you can always take them off later, but even a small piece of AL bolted to the tab will double the very limited dissipation of the package, giving the transistor a longer life.
With 36volt rails the zeners have to dissipate about a 1/6W, if that makes the temp reach 25c above ambient, things will be OK, 45 volts rails raise the zener dissipation to 1/3W, above this point the resistors will need adjusting or things will begin to pop.
The emitter resistors will need to have ~2v each by the time the amp is fully biased up, but you won't get a completely accurate adjustment until the outputs are attached.
At this point I'm not sure why you can't adjust the DC offset. From the schematic it looks like the adjustment should have some effect, outputs attached or not...I don't have the circuit board in front of me, or I'd look, but make sure the feedback point is connected to the union of the driver emitter resistors, and also check the output offset pot has approx 0.4v from the top to bottom, ie +0.2 to -0.2, and that adjusting it gives you this range of voltage offset to the bases of q101 and 102...
As to the effects of DC offset, within the amp it has little or no effect, up to a point. If it is really bad, as in volts, it will reduce the maximum output by asymetric clipping in the output voltage. Your speakers are likely to hate it, it will overheat the voicecoils of any DC connected driveunits, and may phyically damage them as a side effect of the cone having gross physical displacement.
Let us know how it goes...
Stuart
Aaron,
65c is not too hot, but it is likely that the temp will be much hotter once the boards etc are encased. I'd use a heatsink, some people think transistors sound better hot, in which case you can always take them off later, but even a small piece of AL bolted to the tab will double the very limited dissipation of the package, giving the transistor a longer life.
With 36volt rails the zeners have to dissipate about a 1/6W, if that makes the temp reach 25c above ambient, things will be OK, 45 volts rails raise the zener dissipation to 1/3W, above this point the resistors will need adjusting or things will begin to pop.
The emitter resistors will need to have ~2v each by the time the amp is fully biased up, but you won't get a completely accurate adjustment until the outputs are attached.
At this point I'm not sure why you can't adjust the DC offset. From the schematic it looks like the adjustment should have some effect, outputs attached or not...I don't have the circuit board in front of me, or I'd look, but make sure the feedback point is connected to the union of the driver emitter resistors, and also check the output offset pot has approx 0.4v from the top to bottom, ie +0.2 to -0.2, and that adjusting it gives you this range of voltage offset to the bases of q101 and 102...
As to the effects of DC offset, within the amp it has little or no effect, up to a point. If it is really bad, as in volts, it will reduce the maximum output by asymetric clipping in the output voltage. Your speakers are likely to hate it, it will overheat the voicecoils of any DC connected driveunits, and may phyically damage them as a side effect of the cone having gross physical displacement.
Let us know how it goes...
Stuart
Are you talking F, or C on the temp?
I powered up my driver boards last night from +/- 20 volts. With the bias adjust set for about 50 mils per rail the drivers got warm but not nearly hot and its been running all morning now. 1khz Square wave looks really clean but 10khz looks a bit rounded off making me wonder if those 330 pf caps are indeed right or wrong! I will install a pair of 39pf caps on this board and see if that improves the hf a bit before I fire up the complete circuit.
BTW: I am pretty sure that I have enough heat sinking having made a duplicate of the Krell driver heat sink. Later today I will fire up the complete amp... hopefully not really fire though.......
Stuart,
My DC offset has no effect at this point either..... 1.1 volts across the drivers emmitter resistors.....
Mark
I powered up my driver boards last night from +/- 20 volts. With the bias adjust set for about 50 mils per rail the drivers got warm but not nearly hot and its been running all morning now. 1khz Square wave looks really clean but 10khz looks a bit rounded off making me wonder if those 330 pf caps are indeed right or wrong! I will install a pair of 39pf caps on this board and see if that improves the hf a bit before I fire up the complete circuit.
BTW: I am pretty sure that I have enough heat sinking having made a duplicate of the Krell driver heat sink. Later today I will fire up the complete amp... hopefully not really fire though.......
Stuart,
My DC offset has no effect at this point either..... 1.1 volts across the drivers emmitter resistors.....
Mark