Hi,
two or so of you have said that increasing the output bias current improves the bass.
A thought ocurred while dish washing tonight.
Much of the EF/VAS, driver and output is very similar to D Self. So is it stretching it to suggest that some of Self's conclusions on distortion would also apply to KSA Klone?
Self says that a properly biased ClassAB stage has a distortion minima and that biasing above or below this level increases distortion until at very high bias the classAB transforms into a ClassA power amp. Then the crossover is eliminated and the VAS loading improves.
Our KSA at low bias falls into Self's high bias type and as such has a higher level of distortion than the minimum. Much of this will be crossover distortion. As the bias is raised we end up with a true ClassA output stage.
If raising the bias reduces crossover distortion then is it audible?
I now suggest that the reduction in distortion products reduces the hi frequency content of the music signal and gives more prominence to each of the fundamentals present in the music. This would equate to that impression of more bass.
Could this be the effect you are hearing?
At the assumed low level of distortion present I would be surprised that the effect would be more bass, since I had always thought that crossover sounded more like hi frequency grunge added to our music.
Any further thoughts?
two or so of you have said that increasing the output bias current improves the bass.
A thought ocurred while dish washing tonight.
Much of the EF/VAS, driver and output is very similar to D Self. So is it stretching it to suggest that some of Self's conclusions on distortion would also apply to KSA Klone?
Self says that a properly biased ClassAB stage has a distortion minima and that biasing above or below this level increases distortion until at very high bias the classAB transforms into a ClassA power amp. Then the crossover is eliminated and the VAS loading improves.
Our KSA at low bias falls into Self's high bias type and as such has a higher level of distortion than the minimum. Much of this will be crossover distortion. As the bias is raised we end up with a true ClassA output stage.
If raising the bias reduces crossover distortion then is it audible?
I now suggest that the reduction in distortion products reduces the hi frequency content of the music signal and gives more prominence to each of the fundamentals present in the music. This would equate to that impression of more bass.
Could this be the effect you are hearing?
At the assumed low level of distortion present I would be surprised that the effect would be more bass, since I had always thought that crossover sounded more like hi frequency grunge added to our music.
Any further thoughts?
Snap-in caps
I found some 27,000uf 63V caps for those using the PSU boards...
Panasonic - Digi-Key part# ECE-T1JA273FA $14.28 each for orders less than 10
This guy on ebay has
Nichicon for $24 a pair plus 12$ shipping per pair (from Hong Kong to US). I e-mailed for accurate shipping quote for 3 pairs.
I also saw that Roederstein and BHC both make caps of this value but couldn't find a supplier. Not sure where I'm getting mine from, or if I should just put the PSU boards aside and go with the apexjr screw terminal caps.
~Brad
I found some 27,000uf 63V caps for those using the PSU boards...
Panasonic - Digi-Key part# ECE-T1JA273FA $14.28 each for orders less than 10
This guy on ebay has
Nichicon for $24 a pair plus 12$ shipping per pair (from Hong Kong to US). I e-mailed for accurate shipping quote for 3 pairs.
I also saw that Roederstein and BHC both make caps of this value but couldn't find a supplier. Not sure where I'm getting mine from, or if I should just put the PSU boards aside and go with the apexjr screw terminal caps.
~Brad
Re: Snap-in caps
Well, I was planning to use the BrianGT PSU boards, but I was under the impression that it took one board for each channel with a plus and minus rail. That would allow 6 caps per channel, and somewhere in the neighborhood of 120k to 162k per channel depending on the caps. But I don't have my boards yet, so I'm figuring kinda blind here.
Any suggestions?....am I mistaken about using two PSU boards along with two xfmrs for stereo?
Robert
googler said:
Well, I was planning to use the BrianGT PSU boards, but I was under the impression that it took one board for each channel with a plus and minus rail. That would allow 6 caps per channel, and somewhere in the neighborhood of 120k to 162k per channel depending on the caps. But I don't have my boards yet, so I'm figuring kinda blind here.
Any suggestions?....am I mistaken about using two PSU boards along with two xfmrs for stereo?
Robert
I guess I should take a look at the boards a little closer, for some reason I'm thinking that one side of the board is 3 caps in parallel which are in series with the other side... I thought the design allowed for both channels to be connected from one board, and that the reason for two seperate boards was to keep with the dual mono theme of the originals.
I have been and very well could be wrong...
~Brad
I have been and very well could be wrong...
~Brad
The BrianGT pwr supply boards CAN supply both channels.
The MUR860 diodes are capable of 8 amps continuous (properly heat sunk) and you can get main supply caps big enough to support two channels from the single board. I just happen to have lots of smaller caps and ordered 4 pcb's so will do mono.
You will either need to jumper the resistor pads, use inductors, or place the resistors in the boards to parallel the 2 rear caps to the front one.
The MUR860 diodes are capable of 8 amps continuous (properly heat sunk) and you can get main supply caps big enough to support two channels from the single board. I just happen to have lots of smaller caps and ordered 4 pcb's so will do mono.
You will either need to jumper the resistor pads, use inductors, or place the resistors in the boards to parallel the 2 rear caps to the front one.
Mjl4302/4281
Hi,
If I use these transistors in place of the 1302/3281, will my heat dissipation go up? I assume that I can lower the bias so I get the same power dissipation, but the 4302/4281 has a higher fT and a lower junction-to-case thermal resistance. The hFE are all the same from 100mA to 5A, but the 4302/4281 starts to decrease after 5A, while the 1302/3281 goes until 7A before it starts going down. Anyone tried this? Digi-key has them for 6.90. Only a bit more than the spec'ed pair. Thanks.
EDIT:
oops... did anybody ever give an opinion on the sound quality? Doing the search I found some discussion, but no rulings. Thanks.
Hi,
If I use these transistors in place of the 1302/3281, will my heat dissipation go up? I assume that I can lower the bias so I get the same power dissipation, but the 4302/4281 has a higher fT and a lower junction-to-case thermal resistance. The hFE are all the same from 100mA to 5A, but the 4302/4281 starts to decrease after 5A, while the 1302/3281 goes until 7A before it starts going down. Anyone tried this? Digi-key has them for 6.90. Only a bit more than the spec'ed pair. Thanks.
EDIT:
oops... did anybody ever give an opinion on the sound quality? Doing the search I found some discussion, but no rulings. Thanks.
pinkmouse said:
???
Could you elucidate? It's early in the morning here, and my brain can't quite get around that sentence.
Sorry it was late and I was rambling... What I meant to say is aren't C3,C5,and C7 in parralel? as well as C4,C6, and C8?
Is it correct that the odd numbered caps (c3,c5,c7) are in series with the even numbered caps (c4,C6,c8)?
~Brad