transformer?
What say ye about these?
http://www.diycable.com/main/product_info.php?cPath=140_141&products_id=597
Ryan
What say ye about these?
http://www.diycable.com/main/product_info.php?cPath=140_141&products_id=597
Ryan
Very strange.. the original KSA was also wired that way and I did SNR and spectra readings on very crudely wired prototypes without any problem - if there were discernable demodulation taking place it would have shown, so your case may just be a combination of factors. Proper wiring and an aluminium chassis would also have helped the RFI, or a ferrite bead between the caps and the Q2/Q4 bases.
The feedback return (Q2/Q4) of an amplifier with global feedback is sensitive to a possible noise source. If R16 is placed before the caps, this susceptable area will be quite small, and any noise pickup by the large capacitors acting as antennae will be effectively at ground level and damped by R16. In our setup the sensitive antenna area (the node at Q2/Q4, R15, R19 and the caps) is quite large due to the cap size.
For most people this should not be a problem; it certainly wasn't for the original Krell or any of my experiments. Those wishing to be safe can do the swop, the method to get the smallest antenna would be to:
- cut the trace going to J3 as close to the main node as possible
- short out R16's pads
- put R16 vertically in C2's place
- use a 470uF bipolar cap for C3
- do not use C4's pad; tack-solder C4 underneath the board on C3's pins (for Black Gates, do not use C4 whatsoever)
A more elegant method would be to put a small 1206 SMD ferrite bead in series between the junction and the caps, with some ingenuity it can be invisible. Unlike an inductor a ferrite bead is damped and acts more linearly like a resistor at high frequencies.
For most people this should not be a problem; it certainly wasn't for the original Krell or any of my experiments. Those wishing to be safe can do the swop, the method to get the smallest antenna would be to:
- cut the trace going to J3 as close to the main node as possible
- short out R16's pads
- put R16 vertically in C2's place
- use a 470uF bipolar cap for C3
- do not use C4's pad; tack-solder C4 underneath the board on C3's pins (for Black Gates, do not use C4 whatsoever)
A more elegant method would be to put a small 1206 SMD ferrite bead in series between the junction and the caps, with some ingenuity it can be invisible. Unlike an inductor a ferrite bead is damped and acts more linearly like a resistor at high frequencies.
Pierre thanks. If that’s the case I will use that as the general rule for zobels and snubber circuits as well i.e. R-C to earth, instead of the other way around.
On another note, has any one designed an output board with 4 or 8 - TO264 output transistors “similar” to the Pinkmouse output boards but with bi-pass caps etc?
On another note, has any one designed an output board with 4 or 8 - TO264 output transistors “similar” to the Pinkmouse output boards but with bi-pass caps etc?
For a zobel it wouldn't matter since the gain is already fixed - the noise source should be inside the feedback loop, otherwise that would have made your speaker cable and speakers etc. all part of the antenna. Anyway, zobels should preferably be placed on the speaker terminals, I've seen some nasty stuff happen when they're incorrectly referenced.
By-pass Caps on the Power Supply
Gentlemen
After a long absence I have started following this thread again.
Six months ago I replaced the the 16-18 years old main PSU caps in my Krell 100.2 and 50.2 as a preventative measure. The new caps were Sprague versions sourced from Mouser. It amazing how a relatively cheap item in America transforms itself to an expensive item when it arrives at your doorstep.
The amps immediately sounded different and I presumed that the caps simply needed to "play-in". The 50.2 that was my favourite amp lost it's magic and suddenly sounded rather average.
I then started playing with a KSA150 and put the KSA100, the 100 clone and 50.2 on the backburner.
This week I fired up the 100.2 after I had added some 6 yF by-pass caps on the PSU. The results can only be described as exceptional - especially when the cost of the caps are taken into consideration. In fact I have never heard my system sound so good. My gut feeling is that the KSA100 now outperforms the 150 which now sounds "aggressive".
The implication of the above "upgrade" suggest to me that:
a) The KSA100.2 is an incredible amp. Possibly grossly under-performing due to the limitations of it's components quality. If people just knew how easily they can transform this amp !!
b) My KSA50 clone project needs revisiting.
c) I need to experiment with different brands of by-pass caps. I have been told that the Mundorf Gold and Auricaps are well suited to use as a PSU by-pass caps.
d) The PCB by-pass enhancements Pierre has done on the 100 clone PCB suggest to me that the 100 clone should sound better than the original.
e) One possibly needs to revisit the design of the PSU of the clone project and be more careful in terms of the quality of caps you use.
f) There could be a difference in the caps Mouser sell and their counterparts that Krell sell ?
Regards
Jozua
Gentlemen
After a long absence I have started following this thread again.
Six months ago I replaced the the 16-18 years old main PSU caps in my Krell 100.2 and 50.2 as a preventative measure. The new caps were Sprague versions sourced from Mouser. It amazing how a relatively cheap item in America transforms itself to an expensive item when it arrives at your doorstep.
The amps immediately sounded different and I presumed that the caps simply needed to "play-in". The 50.2 that was my favourite amp lost it's magic and suddenly sounded rather average.
I then started playing with a KSA150 and put the KSA100, the 100 clone and 50.2 on the backburner.
This week I fired up the 100.2 after I had added some 6 yF by-pass caps on the PSU. The results can only be described as exceptional - especially when the cost of the caps are taken into consideration. In fact I have never heard my system sound so good. My gut feeling is that the KSA100 now outperforms the 150 which now sounds "aggressive".
The implication of the above "upgrade" suggest to me that:
a) The KSA100.2 is an incredible amp. Possibly grossly under-performing due to the limitations of it's components quality. If people just knew how easily they can transform this amp !!
b) My KSA50 clone project needs revisiting.
c) I need to experiment with different brands of by-pass caps. I have been told that the Mundorf Gold and Auricaps are well suited to use as a PSU by-pass caps.
d) The PCB by-pass enhancements Pierre has done on the 100 clone PCB suggest to me that the 100 clone should sound better than the original.
e) One possibly needs to revisit the design of the PSU of the clone project and be more careful in terms of the quality of caps you use.
f) There could be a difference in the caps Mouser sell and their counterparts that Krell sell ?
Regards
Jozua
For solid-state PSU purposes the M-Cap Silver is likely to work better than the Gold, although a standard M-Cap Supreme may work just as well. For valve gear PSU's it may be different though.
The distance to the PCB, and how it's grounded, is critical - the advantage of very low ESR and ESL and tandelta quickly diminishes if there's 5" of cable between it and the PCB so mounting them in parallel on top of the electrolytics will not yield optimal performance.
The only reason a rebranded Krell Sprague may sound different than a standard one of the same series would be if the Krell one is a few years older, which is likely since Krell doesn't use them in their newest stuff.
The distance to the PCB, and how it's grounded, is critical - the advantage of very low ESR and ESL and tandelta quickly diminishes if there's 5" of cable between it and the PCB so mounting them in parallel on top of the electrolytics will not yield optimal performance.
The only reason a rebranded Krell Sprague may sound different than a standard one of the same series would be if the Krell one is a few years older, which is likely since Krell doesn't use them in their newest stuff.