Correction: more than 20 years too late...highbias said:
Lets make this is as a rule - everybody should build only the Pass stuff.....
Maybe it's a sliding bias stuff causing you trouble....
But seriously what’s the point of building a Krell clone, and which version? Somebody has said somewhere, that only the few first versions of Krell sound good and the later ones are not....
Argo
"Somebody has said somewhere, that only the few first versions of Krell sound good and the later ones are not.... "
Thats my reason for building one. Keep in mind that the KSA80B WAS one of the first few Krells produced.
"Maybe it's a sliding bias stuff causing you trouble.... "
The KSA-80B didn't have sliding bias...
The 80B was actually a very good sounding amp but very cable and head end fussy, better on the low end than the Pass stuff(but the pass stuff holds its own there) but no where near the fantastic midrange or top end quality of the Aleph 2's I built. Soundstaging and imaging on the Pass is so much superior. I got rid of the Krell after I finished the Alephs, so thats over two years ago now.
Mark
Thats my reason for building one. Keep in mind that the KSA80B WAS one of the first few Krells produced.
"Maybe it's a sliding bias stuff causing you trouble.... "
The KSA-80B didn't have sliding bias...
The 80B was actually a very good sounding amp but very cable and head end fussy, better on the low end than the Pass stuff(but the pass stuff holds its own there) but no where near the fantastic midrange or top end quality of the Aleph 2's I built. Soundstaging and imaging on the Pass is so much superior. I got rid of the Krell after I finished the Alephs, so thats over two years ago now.
Mark
Being currently very broke, and unable to track down some of the input end silicon, I have had a thought. I have a good supply of both types of output devices, and so if anyone else in the UK is building this project and has/ will be buying a surplus of bits, then I am up for doing a swap. Anyone?
Transistor Availability
I was concerned about getting all the recommended transistors. Newark and MCM stock all the parts. They are sister companies.
I ordered a box full last Friday and they all showed up yesterday. The best place to order is www.mcminone.com They stock all.
The output devices can be purchased elsewhere cheaper. But I did look a Digikey and they do not stock all, and the prices were no better.
I think Newark will ship internationally.
George
I was concerned about getting all the recommended transistors. Newark and MCM stock all the parts. They are sister companies.
I ordered a box full last Friday and they all showed up yesterday. The best place to order is www.mcminone.com They stock all.
The output devices can be purchased elsewhere cheaper. But I did look a Digikey and they do not stock all, and the prices were no better.
I think Newark will ship internationally.
George
"I have heard a krell amp many (many) years ago and all i remember is how hot it got and how it was the best thing i'd heard!"
They WERE the best many years ago...... as oddly many of todays designs didn't exist yet. The interesting thing was that my KSA-80B came from a person that had just bought a new Aleph 5..... this was about 1997 or 98. He raved about the Aleph to me and told me about its virtues. Being a big Krell fan back then I didn't believe him, especially since it was a step down in power output.... but even that was many years ago now..... now I know better. But I still look foreword to building the 50 in the very near future. It'll be a fun project to build sometihng vintage....
Mark
They WERE the best many years ago...... as oddly many of todays designs didn't exist yet. The interesting thing was that my KSA-80B came from a person that had just bought a new Aleph 5..... this was about 1997 or 98. He raved about the Aleph to me and told me about its virtues. Being a big Krell fan back then I didn't believe him, especially since it was a step down in power output.... but even that was many years ago now..... now I know better. But I still look foreword to building the 50 in the very near future. It'll be a fun project to build sometihng vintage....
Mark
Stuart
Thanks for the info and willingness to assist. In fact it is a brilliant idea to reduce the bias- i must just cross that "fear" barrier of actually fiddling inside Krell amp.
Here is a link that I think is potential value to this discussion, especially as most of the early Krell circuit diagrams can be found at this link and it might be worthwhile to compare with what this group has designed so far.
http://home.ca.inter.net/~lloyd.maclean/Krell/Krell.htm
Jozua
Thanks for the info and willingness to assist. In fact it is a brilliant idea to reduce the bias- i must just cross that "fear" barrier of actually fiddling inside Krell amp.
Here is a link that I think is potential value to this discussion, especially as most of the early Krell circuit diagrams can be found at this link and it might be worthwhile to compare with what this group has designed so far.
http://home.ca.inter.net/~lloyd.maclean/Krell/Krell.htm
Jozua
very helpful info...
Hi Jozua,
I've been following the other Krell threads, and checking the postings on the website you mention. So far everything we are doing here seems to align very well with the info there. Both Aaron & I asked Mr Mclean the same questions regarding bias and rail voltages etc.
If you have a multimeter adjusting the bias is very easy, one direction on the pot will raise it, the other lower it. Obviously you make initial changes slowly, then once you have an idea about which direction the pot goes in, you set it, let the amp get to a stable temp, check it and adjust again. Since the ksa100 has more transistors, the voltages across the emitter resistors will not match the numbers for the ksa50, you will want to see the voltages listed for the ksa100, reduced by about 40%.
Good luck
Stuart
Hi Jozua,
I've been following the other Krell threads, and checking the postings on the website you mention. So far everything we are doing here seems to align very well with the info there. Both Aaron & I asked Mr Mclean the same questions regarding bias and rail voltages etc.
If you have a multimeter adjusting the bias is very easy, one direction on the pot will raise it, the other lower it. Obviously you make initial changes slowly, then once you have an idea about which direction the pot goes in, you set it, let the amp get to a stable temp, check it and adjust again. Since the ksa100 has more transistors, the voltages across the emitter resistors will not match the numbers for the ksa50, you will want to see the voltages listed for the ksa100, reduced by about 40%.
Good luck
Stuart
Two more questions!!
The krell KSA-100 boards show that it has MJE15030 as a bias transistor...... can this be directly substituted for the 2SC3955 (i think it was) that is proposed to be used in this kit? (it's harder to get and i've got a stack of MJE15030's!) I'm hoping that nothing has to be changed to do this?
Also the output's on the KSA-100 are driven by 2 sets of MJE15030+31.... Will the MJE15032+33 enough to drive more outputs? The drivers will be on the main heatsink i think due to plenty of heatsink area... So will they be ok running, say, 16 output transistors or 12? or 8? Stock should be 4, but i'm going to at least double the load on them... Or will i need to modify it to add another set of drivers? I don't mind having to test it to find out - but i'd prefer to know before blowing it!
Speaking of which i have 25 output pairs here now and more on the way They are quite heavy surprisingly... Instead of using output "boards" i might make up a copper buss-bar system to allow PLENTY of current.... How come we have collector (i think) resistors and the original schematics do not? (going off the top of my head, pretty sure they were collector ones!)
Thanks heaps!
Aaron
The krell KSA-100 boards show that it has MJE15030 as a bias transistor...... can this be directly substituted for the 2SC3955 (i think it was) that is proposed to be used in this kit? (it's harder to get and i've got a stack of MJE15030's!) I'm hoping that nothing has to be changed to do this?
Also the output's on the KSA-100 are driven by 2 sets of MJE15030+31.... Will the MJE15032+33 enough to drive more outputs? The drivers will be on the main heatsink i think due to plenty of heatsink area... So will they be ok running, say, 16 output transistors or 12? or 8? Stock should be 4, but i'm going to at least double the load on them... Or will i need to modify it to add another set of drivers? I don't mind having to test it to find out - but i'd prefer to know before blowing it!
Speaking of which i have 25 output pairs here now and more on the way
They are quite heavy surprisingly... Instead of using output "boards" i might make up a copper buss-bar system to allow PLENTY of current.... How come we have collector (i think) resistors and the original schematics do not? (going off the top of my head, pretty sure they were collector ones!)Thanks heaps!
Aaron
drivers
Aaron,
The schematic for the early ksa100 I have only shows 1 pair of 30/31 drivers and 4 pairs of outputs. These are quite beefy transistors, and while they won't die they are not going to drive 10 or more output pairs to maximal output without a recalculation and recalibration of the various currents involved. On the other hand, with a 36 volt rail, 5 pairs of the 15003/4 outputs can drive 100amps for short periods, and half that for long enough to dim the lights, pop the breaker etc. Are you sure you need more? If you have big enough heatsinks I don't think the number of output transistors will be a major limiting factor, and if you dont have big enough heatsinks, no amount of outputs will compensate...
I certainly plan on using the '30 as the bias transistor, it doesn't have a terribly taxing job to do, and the 2sc3955 seems a largely unexceptional transistor. The ksa100 schematics I have show, variously 2sc3955, mje15033 and mje340 as the vbe multiplier.
The mje15032/33 pairs look like higher voltage versions of the mje15028/29/30/31, so given the low (!) voltages we are using, I suspect any will do...
Stuart
Aaron,
The schematic for the early ksa100 I have only shows 1 pair of 30/31 drivers and 4 pairs of outputs. These are quite beefy transistors, and while they won't die they are not going to drive 10 or more output pairs to maximal output without a recalculation and recalibration of the various currents involved. On the other hand, with a 36 volt rail, 5 pairs of the 15003/4 outputs can drive 100amps for short periods, and half that for long enough to dim the lights, pop the breaker etc. Are you sure you need more? If you have big enough heatsinks I don't think the number of output transistors will be a major limiting factor, and if you dont have big enough heatsinks, no amount of outputs will compensate...
I certainly plan on using the '30 as the bias transistor, it doesn't have a terribly taxing job to do, and the 2sc3955 seems a largely unexceptional transistor. The ksa100 schematics I have show, variously 2sc3955, mje15033 and mje340 as the vbe multiplier.
The mje15032/33 pairs look like higher voltage versions of the mje15028/29/30/31, so given the low (!) voltages we are using, I suspect any will do...
Stuart
Hi Stuart,
I was looking at the link previously posted for the site about the krell repairs, etc the pictures actually show it using an 15030 as bias and also has 4 drivers! Which surprised me as i hadn't seen that on any schematics before (other than the ones at the site)... I wasn't PLANNING on using 10+ outputs, more rather was trying to get an idea if they'd run 6/8 which is pretty much the limit i'm going to be at anyway (i don't need THAT much power!)
Thanks
Aaron
I was looking at the link previously posted for the site about the krell repairs, etc the pictures actually show it using an 15030 as bias and also has 4 drivers! Which surprised me as i hadn't seen that on any schematics before (other than the ones at the site)... I wasn't PLANNING on using 10+ outputs, more rather was trying to get an idea if they'd run 6/8 which is pretty much the limit i'm going to be at anyway (i don't need THAT much power!)
Thanks
Aaron
"On the other hand, with a 36 volt rail, 5 pairs of the 15003/4 outputs can drive 100amps for short periods, and half that for long enough to dim the lights, pop the breaker etc."
Those that would go this route would be wise to get a transformer primary that is set for 220 volts as the likelyhood of popping breakers would be far less likely. In large cinema applications I prefer to power the subs (BGW M2200) off 220 volts. There is an audible difference!!
Mark
Those that would go this route would be wise to get a transformer primary that is set for 220 volts as the likelyhood of popping breakers would be far less likely. In large cinema applications I prefer to power the subs (BGW M2200) off 220 volts. There is an audible difference!!
Mark
Hi All-
I had posted an example of an output board and did not hear anything on it.
Also, I was planning on ordering the boards from the place Jan mentioned because they were cheaper than I have been able to find here.
I am more than happy to handle the order and mailing of the boards but I am not sure I have been effective in finding a good price.
I had posted an example of an output board and did not hear anything on it.
Also, I was planning on ordering the boards from the place Jan mentioned because they were cheaper than I have been able to find here.
I am more than happy to handle the order and mailing of the boards but I am not sure I have been effective in finding a good price.