Mark, msg me your email address and I can mail you the preliminary Gerbers. If somebody here cannot provide the exact details of the original boards we'll have to wait a week or so until I can get my hands on one. A few mm offset on the mounting holes for the boards to the heatsink and the whole attraction of the "compatible clone" is ruined. I'll try contacting Alex, he seems to have a few KSA's lying around.
I think standard FR-4, coloured boards will be fine, with gold plated pads, 1.55mm thickness and 70um (2oz) copper thickness. I'm personally fond of gloss black boards (looks so pretty with gold pads), but I suppose if people want the authentic look it should be matt green. Don't use G-10, it's inferior. I've never heard of Teflon boards.
As for transformers, I believe that anything with vacuum-impregnated cores, varnish-dipped windings and proper line and load regulation is all you need. A grounded shield between the windings and potting is nice to have but in my experience (admittedly class-AB) not really necessary. Many of the custom-wound transformers at local companies also tend to have better specs than some of the commercial ones, although I'm not sure where Avel and the rest fit in. Apparently the Signal transformers are excellent if you look past their rather high cost.
I think standard FR-4, coloured boards will be fine, with gold plated pads, 1.55mm thickness and 70um (2oz) copper thickness. I'm personally fond of gloss black boards (looks so pretty with gold pads), but I suppose if people want the authentic look it should be matt green. Don't use G-10, it's inferior. I've never heard of Teflon boards.
As for transformers, I believe that anything with vacuum-impregnated cores, varnish-dipped windings and proper line and load regulation is all you need. A grounded shield between the windings and potting is nice to have but in my experience (admittedly class-AB) not really necessary. Many of the custom-wound transformers at local companies also tend to have better specs than some of the commercial ones, although I'm not sure where Avel and the rest fit in. Apparently the Signal transformers are excellent if you look past their rather high cost.
Noisy transformers?
Those who does not know of buzzing and noisy transformers:
http://www.diyaudio.com/forums/showthread.php?s=&threadid=76961&highlight=
http://www.plitron.com/pages/lono.htm
http://www.diyaudio.com/forums/showthread.php?s=&threadid=76961&perpage=10&highlight=&pagenumber=2 (post #13)
http://www.psaudio.com/articles/hum.asp
and
http://www.plitron.com/PDF/AES.PDF#search="Transformer noise buzzing" This is a very interesting paper from plitron....read it,
Hope this helps to clear things out what is a buzzing transformer and why cheaper ones often buzzes.
Regards
Those who does not know of buzzing and noisy transformers:
http://www.diyaudio.com/forums/showthread.php?s=&threadid=76961&highlight=
http://www.plitron.com/pages/lono.htm
http://www.diyaudio.com/forums/showthread.php?s=&threadid=76961&perpage=10&highlight=&pagenumber=2 (post #13)
http://www.psaudio.com/articles/hum.asp
and
http://www.plitron.com/PDF/AES.PDF#search="Transformer noise buzzing" This is a very interesting paper from plitron....read it,
Hope this helps to clear things out what is a buzzing transformer and why cheaper ones often buzzes.
Regards
PWatts,
You need to turn your messaging on. Right now I can't send you one. I will check back in a couple of hours.
Flodstroem
I will go along with your transformer order as it will be one less thing I will have to do since I will be handling the heat sinks and probably the boards too. After all I do want to have at least 2 hours left in a week to be able to work on the thing. I may have to build my amp while I'm on the road since I've been away more than I've been home lately. Are these toroids at least 1kva units or larger and are they potted in a can or not? electrostatic shield and screen hopefully? I think you listed the specs but I went back and couldn't find them.
P.S. I will send you out another set of KSA-50 boards when I get back home... when ever that is. Assuming that you still have not received the first shipment. If so this makes a total of three shipments of KSA-50 boards to three different places that have been lost by the mails. I guess not so bad considering how many were shipped.
Mark
You need to turn your messaging on. Right now I can't send you one. I will check back in a couple of hours.
Flodstroem
I will go along with your transformer order as it will be one less thing I will have to do since I will be handling the heat sinks and probably the boards too. After all I do want to have at least 2 hours left in a week to be able to work on the thing. I may have to build my amp while I'm on the road since I've been away more than I've been home lately. Are these toroids at least 1kva units or larger and are they potted in a can or not? electrostatic shield and screen hopefully? I think you listed the specs but I went back and couldn't find them.
P.S. I will send you out another set of KSA-50 boards when I get back home... when ever that is. Assuming that you still have not received the first shipment. If so this makes a total of three shipments of KSA-50 boards to three different places that have been lost by the mails. I guess not so bad considering how many were shipped.
Mark
Mark,
I really dont know what is the specs for a KSA 100mkII transformer because I dont have the transformers number (model number). Hopefully somebody in this forum that has access to an original Krell KSA 100 amp could check this model number out
Avel Lindberg supplied Krell company with 15 models of transformers but they do not know which transformer was used to which amplifier. Thats the reason for the need of some sort of model number of transformer used in this amp. When we have found it/received it we can check this data at Avel (total VA output, screenings, windings, potted in a metal can, etc...).
I really appreciate if you do that, thanks Mark.
Regards
I really dont know what is the specs for a KSA 100mkII transformer because I dont have the transformers number (model number). Hopefully somebody in this forum that has access to an original Krell KSA 100 amp could check this model number out
Avel Lindberg supplied Krell company with 15 models of transformers but they do not know which transformer was used to which amplifier. Thats the reason for the need of some sort of model number of transformer used in this amp. When we have found it/received it we can check this data at Avel (total VA output, screenings, windings, potted in a metal can, etc...).
I really appreciate if you do that, thanks Mark.
Regards
Mark,
When KRELL switched their supplier from Avel to Holden for the main transformers was in the end of 1987 (ref: A.Lindberg), but in the tech. inform. from Krell they say september 1988. But I think Krell had a stock Avel transformer up to that date (-88).
The schematics I posted in my last post was for the mk-I not the mk-II. I dont know if Krell made any upgrades/ changes from the mk-I to the mk-II transformer.
Only a transformers serial/model number can tall us. We had to wait to see if somebody could help us to find it.
Regards
When KRELL switched their supplier from Avel to Holden for the main transformers was in the end of 1987 (ref: A.Lindberg), but in the tech. inform. from Krell they say september 1988. But I think Krell had a stock Avel transformer up to that date (-88).
The schematics I posted in my last post was for the mk-I not the mk-II. I dont know if Krell made any upgrades/ changes from the mk-I to the mk-II transformer.
Only a transformers serial/model number can tall us. We had to wait to see if somebody could help us to find it.
Regards
Ok, here is what I just did. I found a guy selling a KSA-100 on Audiogon and I wrote to him to see if he could remove his cover and find any numbers at all on the transformers. Weather or not he is willing to do this I don't know but I will report back as soon as I hear from him. I also gave him a link to this thread so he can see that I'm not "pulling his leg" so to say.
Other than watching for amps for sale and seeing if those guys will take a look see inside for us I don't know any other way to find out.
Is Holden still in buisness? I did a google and came up empty.
Mark
Other than watching for amps for sale and seeing if those guys will take a look see inside for us I don't know any other way to find out.
Is Holden still in buisness? I did a google and came up empty.
Mark
OK, I've made a few small changes and one major one: thickened some traces, made a few cosmetic adaptions to routing and silkscreen, added the silkscreen legend to indicate the direction of the trimpot to increase bias, and added small SMD resistors and LED's to the power rails for those who want to use them. They're so small that they don't interfere with anything, so it comes at no cost to the board. Personally I like a little tiny SMD LED telling me all is well on both rails
(BTW what makes the most sense: clockwise for bias increase or decrease?)
The major change is in the 825R feedback resistor in series with the DC blocking cap. Its signal path is no much shorter and simpler than before, and the grounds leading back to the star point is now also considerably simplified. For all practical purposes the single ground line that runs up is only for the current regulators, the other two components connected to the same ground point is unlikely to conduct any current and is therefore not an issue to be placed on the same ground return. Another advantage is that the jumper to bypass the DC block cap is now on the top side where it's a bit more useful.
The only penalty of this arrangement is another resistor on the bottom layer and some careful soldering.
Up until I open up a Mk2 or somebody gives me the info of the original board, the layout has been halted though.
(BTW what makes the most sense: clockwise for bias increase or decrease?)
The major change is in the 825R feedback resistor in series with the DC blocking cap. Its signal path is no much shorter and simpler than before, and the grounds leading back to the star point is now also considerably simplified. For all practical purposes the single ground line that runs up is only for the current regulators, the other two components connected to the same ground point is unlikely to conduct any current and is therefore not an issue to be placed on the same ground return. Another advantage is that the jumper to bypass the DC block cap is now on the top side where it's a bit more useful.
The only penalty of this arrangement is another resistor on the bottom layer and some careful soldering.
Up until I open up a Mk2 or somebody gives me the info of the original board, the layout has been halted though.
Sharp of you Paul; I realized it soon after I posted it. It's corrected now, somehow I cannot edit the post so here's the corrected version. The reason you cannot see the Ref Des is because it's on the bottom layer and since it's so simple I decided to make the boards cheaper and leave out the bottom silkscreen.
I also just emailed Mark the preliminary Gerbers.
I also just emailed Mark the preliminary Gerbers.
AndrewT,
41 x 1.414 = 58V-1.4V (rectifier)=56.6V x 0.95(5% regulation)=53.77V DC rail
Yes you are right Andrew.
If this unit is equipped with small transformers (one for each channel) the "regulation" will be poor if compared to a stereo unit equipped with only one big transformer, having 5% regulation.
The transformers must have a regulation down to ca 8% indicating 300-500VA maximum!!??
Make this any sense ?
Regards
41 x 1.414 = 58V-1.4V (rectifier)=56.6V x 0.95(5% regulation)=53.77V DC rail
Yes you are right Andrew.
If this unit is equipped with small transformers (one for each channel) the "regulation" will be poor if compared to a stereo unit equipped with only one big transformer, having 5% regulation.
The transformers must have a regulation down to ca 8% indicating 300-500VA maximum!!??
Make this any sense ?
Regards
Mark A. Gulbrandsen said:
basic electronics
if we presume that xformer have good enough regulation (ie big enough) than everything is in cap bank ;
if caps are big enough , they can fill gap between highs and lows
in heavy A class biasing,proper mult factor is often near 1.2
yes-you'll always have losses in xformer and diode bridges (with increasing current losses increase too
),but that is usually negligible comparing to cap bank inability to fill holes between two sinusoidalsHi,
I find the root 2 rule VERY accurate.
The big error is mains incoming voltage. It changes from minute to minute and hour to hour. The swings are quite large.
The other big discrepancy is regulation, particularly for smaller transformers sometimes approaching 30% (but not in the Krell).
Finally the VA rating needs to be reduced to about 70% for a capacitor input PSU. When combined with the root 2 voltage multiplication after the rectifier, then the effective DC output current is almost exactly half the AC current. A nice easy rule to remember
i.e. 500VA 50Vac =10Arms
For a DC PSU this equates to about 70Vdc and 5A =350W
I find the root 2 rule VERY accurate.
The big error is mains incoming voltage. It changes from minute to minute and hour to hour. The swings are quite large.
The other big discrepancy is regulation, particularly for smaller transformers sometimes approaching 30% (but not in the Krell).
Finally the VA rating needs to be reduced to about 70% for a capacitor input PSU. When combined with the root 2 voltage multiplication after the rectifier, then the effective DC output current is almost exactly half the AC current. A nice easy rule to remember
i.e. 500VA 50Vac =10Arms
For a DC PSU this equates to about 70Vdc and 5A =350W