Thanks a lot jan,
There are some differences with the last schematics:
R125 =?
R125 must be 820R
R126 must be 5K 10-turn pot
and i asked earlyer but no answer came from enyone, is a quiestcurrent from 2,5A good? must this current be measured over all off the four emitterresistors or per rail half off that for each resistant (so 1,25A over each off the four resistors)
There are some differences with the last schematics:
R125 =?
R125 must be 820R
R126 must be 5K 10-turn pot
and i asked earlyer but no answer came from enyone, is a quiestcurrent from 2,5A good? must this current be measured over all off the four emitterresistors or per rail half off that for each resistant (so 1,25A over each off the four resistors)
"2sc3955" is something that i can't find from regular suppliers in australia... Is it possible there is something else we can use?
Another question - with most of the components, is there a list of "IDEAL" substitutions we can do without affecting the circuits greatly?
Also, with the mounting of the bias transistor - i am not sure about locating this... Can we quite literally run leads between it and the mainboard as all the output's will be on the heatsink not the mainboard?
Thanks
Aaron
[edit - also do we have a "design" for the output boards? I know you can simply wire up the leads of the outputs with the resistors, however i'd prefer a board as it will make things 50,000x easier for me to work out!
Is there a list for the output devices? Thanks ]
Another question - with most of the components, is there a list of "IDEAL" substitutions we can do without affecting the circuits greatly?
Also, with the mounting of the bias transistor - i am not sure about locating this... Can we quite literally run leads between it and the mainboard as all the output's will be on the heatsink not the mainboard?
Thanks
Aaron
[edit - also do we have a "design" for the output boards? I know you can simply wire up the leads of the outputs with the resistors, however i'd prefer a board as it will make things 50,000x easier for me to work out!
Is there a list for the output devices? Thanks ]Parts
NUTTTR said:"2sc3955" is something that i can't find from regular suppliers in australia... Is it possible there is something else we can use?
Another question - with most of the components, is there a list of "IDEAL" substitutions we can do without affecting the circuits greatly?
Also, with the mounting of the bias transistor - i am not sure about locating this... Can we quite literally run leads between it and the mainboard as all the output's will be on the heatsink not the mainboard?
Thanks
Aaron
[edit - also do we have a "design" for the output boards? I know you can simply wire up the leads of the outputs with the resistors, however i'd prefer a board as it will make things 50,000x easier for me to work out!
I think the 2SC3955 can be purchased here in the US. I have some on order, let see if they show up.
If they can be had here, I can send you some to build your amp.
George
2sc3955
Hi Again,
I have about 4 different schematics for the ksa series amps, at least 1 of them (post 1983 ksa100ii) shows a mjl15030 there, so I would think it is a reasonable substitute. At least I hope so, since I plan to use it...
The mj15003 are the original outputs but ACD's schematic shows some very nice transistors (mjl21193/4) which are technically better, though marginally less robust. Both are 'preferred' on-semi devices, so obtaining them should be straightforward. The soar curves for the transistors would suggest that 3 of the mjl parts (3a @ 50v) should be roughly as 'tough' as 2 of the to3 (5a @ 50v). But if the voltage is lowered from 50 both have capacity to spare, so I guess I'd choose based on the power supply I ended up with.
With respect to the bias transistor, ideally it would be mounted immediately adjacent to the hottest transistor, and the leads run from there to the board. it should not be far away from the outputs no matter what. As fearless diyers, we could probably get everything working, let everything warm up, measure a few temperatures with it in a temporary position then put it where the temp is highest...
Stuart
Hi Again,
I have about 4 different schematics for the ksa series amps, at least 1 of them (post 1983 ksa100ii) shows a mjl15030 there, so I would think it is a reasonable substitute. At least I hope so, since I plan to use it...
The mj15003 are the original outputs but ACD's schematic shows some very nice transistors (mjl21193/4) which are technically better, though marginally less robust. Both are 'preferred' on-semi devices, so obtaining them should be straightforward. The soar curves for the transistors would suggest that 3 of the mjl parts (3a @ 50v) should be roughly as 'tough' as 2 of the to3 (5a @ 50v). But if the voltage is lowered from 50 both have capacity to spare, so I guess I'd choose based on the power supply I ended up with.
With respect to the bias transistor, ideally it would be mounted immediately adjacent to the hottest transistor, and the leads run from there to the board. it should not be far away from the outputs no matter what. As fearless diyers, we could probably get everything working, let everything warm up, measure a few temperatures with it in a temporary position then put it where the temp is highest...
Stuart
Too True Stuart!
I guess i'll have to wait till i build one up
When can we get prices for the GB? We're looking at getting heatsinks/casings made shortly So potentially we could have the case ready before the rest of the amp!! Wouldn't surprise me!
Should be an excellent unit to build, but will have a fair bit of power drain!!!
How many watts of power should each channel (mono module) drain from the mains with ~45vdc rail?
Thanks
Aaron
I guess i'll have to wait till i build one up
When can we get prices for the GB? We're looking at getting heatsinks/casings made shortly
So potentially we could have the case ready before the rest of the amp!! Wouldn't surprise me!Should be an excellent unit to build, but will have a fair bit of power drain!!!
How many watts of power should each channel (mono module) drain from the mains with ~45vdc rail?
Thanks
Aaron
Krell bias transistor location
It is interesting to note that not long after the first KSA series were introduced, Krell moved the bias transistor off of the main heatsink and onto the same heatsink as the drivers. Because the outputs are operated in pure class A mode, the temperature profile is more stable.
It is interesting to note that not long after the first KSA series were introduced, Krell moved the bias transistor off of the main heatsink and onto the same heatsink as the drivers. Because the outputs are operated in pure class A mode, the temperature profile is more stable.
Re: Krell bias transistor location
That is interesting....
Potentially the drivers might get hotter quicker than the outputs?! No idea why - perhaps (i do mean perhaps) they had no problems with the output temperature and felt very sure that there wouldn't be any problems whatsoever?
Is there any problem mounting the drivers/bias transistor/outputs all onto the one heatsink?
On a side note, i know previously it has been mentioned with our power supplys and voltages (30-0-30 @ ~10-13amp) that we'd be able to push about 100 (maybe more) wrms @ 8 ohm, more obviously at 4... I am "aiming" to make it 4 ohm stable... I'm guessing it will easily push ~200wrms into 4 ohm... What sort of heat will be generated in class a with that sort of output? I'd be expecting enough heat to burn up the room!!! That is part of the reason i will be looking at 3 pairs of TO-3's or maybe 4 - or 4 pairs of 93/94's???
What does everyone think? What rails will the majority be running on the amp?
I think it's to the point where we need a new post!
How many will be attempting to construct these excellent units?
Aaron
lmaclean said:
That is interesting....
Potentially the drivers might get hotter quicker than the outputs?! No idea why - perhaps (i do mean perhaps) they had no problems with the output temperature and felt very sure that there wouldn't be any problems whatsoever?
Is there any problem mounting the drivers/bias transistor/outputs all onto the one heatsink?
On a side note, i know previously it has been mentioned with our power supplys and voltages (30-0-30 @ ~10-13amp) that we'd be able to push about 100 (maybe more) wrms @ 8 ohm, more obviously at 4... I am "aiming" to make it 4 ohm stable... I'm guessing it will easily push ~200wrms into 4 ohm... What sort of heat will be generated in class a with that sort of output? I'd be expecting enough heat to burn up the room!!! That is part of the reason i will be looking at 3 pairs of TO-3's or maybe 4 - or 4 pairs of 93/94's???
What does everyone think? What rails will the majority be running on the amp?
I think it's to the point where we need a new post!
How many will be attempting to construct these excellent units?
Aaron
power
If you bias the amp at say 4amps, you are going to pull approx. 360 watts from the wall continuously per channel, but 4amps is not enough to give the full class A power ~200w) into 4 ohms, we get about 120rms. Not sure if that matters, but if the amp can put out +/-40v into 4 ohms you are looking at peak currents in the 10amp range, if you bias the amp to do this class A, the thermal output jumps to 450w per channel, with the actual output into 4 being 200w RMS, much more than the rated power.
If we stick with class A ratings of 50 into 8, 100 into 4, we can be more conservative with the bias, 2.5amps will do it, giving 225w per channel. This doesnt mean the amp can't be used for arc welding etc, just the output won't be fully class A for the max output into <4 ohm loads.
If you are having custom cases and heatsinks made, and you know the things can get rid of a couple of kilowatts of heat, add another output transistor and crank it...;^) ...It must get cold where you live...
Stuart
If you bias the amp at say 4amps, you are going to pull approx. 360 watts from the wall continuously per channel, but 4amps is not enough to give the full class A power ~200w) into 4 ohms, we get about 120rms. Not sure if that matters, but if the amp can put out +/-40v into 4 ohms you are looking at peak currents in the 10amp range, if you bias the amp to do this class A, the thermal output jumps to 450w per channel, with the actual output into 4 being 200w RMS, much more than the rated power.
If we stick with class A ratings of 50 into 8, 100 into 4, we can be more conservative with the bias, 2.5amps will do it, giving 225w per channel. This doesnt mean the amp can't be used for arc welding etc, just the output won't be fully class A for the max output into <4 ohm loads.
If you are having custom cases and heatsinks made, and you know the things can get rid of a couple of kilowatts of heat, add another output transistor and crank it...;^) ...It must get cold where you live...
Stuart
Full bias into class a
So for what you are saying ~4amps bias (ouch) current will mean that we'd have 120w class a then the remainder into class a/b? the power supplies i guess with the caps will be about +-40-45v... which is a large amount, and 10 amps bias is 450w of HEAT!!! Well, as much as i want this to work, that is a seriously large amount of heat and makes 1.6kw per amp unit of heat, i don't mind, my part of the house stays fairly cold (where the amp will live) but i will HAVE (read MUST) to use the bias adjuster for the hot periods, i guess that makes the amp hugely flexible!!!
I'm happy with 450w of heat, but, sorry to sound stupid, but if we are looking at using 800va toroids, this doesn't sound possible to me... is it?
Output transistors are plenty, i've got enough for 50 pairs per amp...
So, if we are REALISTICALLY looking at getting ~180-200w into 4 ohm per channel at full class A, approx 450w/heat per channel, what toroid would we need? I can do simple calculations but i feel there is something i'm missing in my calcs (and that is experience!) so any help would be appreciated... I don't want to have to have 1600va toroids Maximum is about 1kva prefer 800va as that would be easier (and cheaper)....
So what realistically is a good start for transformers?
Our cases are going to be fully custom and we can potentially get rid of stupid amounts of heat (heatsink is about 500mmx800mm 2mm fins about 50mm deep)... we are planning on cutting those down to smaller sizes, obviously, but the point is we will have stacks of heatsink area!
Thanks for the help once again Stuart!
Aaron
So for what you are saying ~4amps bias (ouch) current will mean that we'd have 120w class a then the remainder into class a/b? the power supplies i guess with the caps will be about +-40-45v... which is a large amount, and 10 amps bias is 450w of HEAT!!! Well, as much as i want this to work, that is a seriously large amount of heat and makes 1.6kw per amp unit of heat, i don't mind, my part of the house stays fairly cold (where the amp will live) but i will HAVE (read MUST) to use the bias adjuster for the hot periods, i guess that makes the amp hugely flexible!!!
I'm happy with 450w of heat, but, sorry to sound stupid, but if we are looking at using 800va toroids, this doesn't sound possible to me... is it?
Output transistors are plenty, i've got enough for 50 pairs per amp...
So, if we are REALISTICALLY looking at getting ~180-200w into 4 ohm per channel at full class A, approx 450w/heat per channel, what toroid would we need? I can do simple calculations but i feel there is something i'm missing in my calcs (and that is experience!) so any help would be appreciated... I don't want to have to have 1600va toroids
Maximum is about 1kva prefer 800va as that would be easier (and cheaper)....So what realistically is a good start for transformers?
Our cases are going to be fully custom and we can potentially get rid of stupid amounts of heat (heatsink is about 500mmx800mm 2mm fins about 50mm deep)... we are planning on cutting those down to smaller sizes, obviously, but the point is we will have stacks of heatsink area!
Thanks for the help once again Stuart!
Aaron
transistors...
If you add 2 pairs of the to3 transistors and of course beef up everything else appropriately, then you are actually making a KSA100. I havent compared the values of all the components, but the schematics are virtually identical.
Personally I think I'm gonna stick with a KSA50, with 100 class A watts into 4. I can live with class B for lower impedances. However I don't think I can live with an amp pumping out a kilowatt all the time, so I'm going to lower the rails a little to make the voltage & current top out at the same time, should be about +/-35v, with 2.5amp bias, 'only' 175w per channel.
I'm almost done on a set of Alephs with largely similar ratings, should be interesting to compare the results.
Stuart
If you add 2 pairs of the to3 transistors and of course beef up everything else appropriately, then you are actually making a KSA100. I havent compared the values of all the components, but the schematics are virtually identical.
Personally I think I'm gonna stick with a KSA50, with 100 class A watts into 4. I can live with class B for lower impedances. However I don't think I can live with an amp pumping out a kilowatt all the time, so I'm going to lower the rails a little to make the voltage & current top out at the same time, should be about +/-35v, with 2.5amp bias, 'only' 175w per channel.
I'm almost done on a set of Alephs with largely similar ratings, should be interesting to compare the results.
Stuart
Re: transistors...
Should we be worried about having the rails at almost +-45v? Just wondering as it does make me wonder, that is A LOT of power for a class a amp for "so little" power output.... I'd realistically want it 4 ohm stable, 2 ohm is not needed, 200wrms @ 4 ohm into class a i'd like that to be a possibility not a distant dream as this amp is being built to last a long time and also allow for plenty of future upgrades....
Is this anything i should be concearned about?
Stuart Easson said:
Should we be worried about having the rails at almost +-45v? Just wondering as it does make me wonder, that is A LOT of power for a class a amp for "so little" power output.... I'd realistically want it 4 ohm stable, 2 ohm is not needed, 200wrms @ 4 ohm into class a i'd like that to be a possibility not a distant dream as this amp is being built to last a long time and also allow for plenty of future upgrades....
Is this anything i should be concearned about?
KSA100 vs KSA50
I don't mind building a "ksa100" but i DEFINATELY want to retain what the KSA-50 was renound for!
The resistors through the amp are different, but that's pretty much it, looks to me potentially that was due to different rail voltages... Thats the only reason i can see for different values, otherwise they are identical!
Aaron
I don't mind building a "ksa100" but i DEFINATELY want to retain what the KSA-50 was renound for!
The resistors through the amp are different, but that's pretty much it, looks to me potentially that was due to different rail voltages... Thats the only reason i can see for different values, otherwise they are identical!
Aaron
worried?
Hmm, there are to many unknowns to really have a perfect answer to this. I, and I assume we, don't know what the actual rail voltage and bias were on a KSA50. I don't believe based on pictures I've seen it can have been class A into 4 ohms, with 45v rails, it was a beefy amp, but not that beefy...So my gut feeling is that the rails were lower than 45, though perhaps not as low as I'm planning to go.
Should you be worried? tough call, not knowing your thermal solution etc it's hard to know if you are building something that will be fragile...if you get toroids and the voltage is too high, unwrap the tape and unwind a couple of turns, ( you may have to unwind one entire secondary depending on how the secondaries are wound ) measure and repeat until you are happy. If you need 5v extra for the driver stage, wind a few extra turns around the core, measure, repeat etc... put them in series with the secondary and rectify and smooth the result. Toroids are a diy'ers dream component, not that E cores are bad, just not as easily modified...
Krell had very good results with the KSA100, which must have had 45v minimum, and more than 2.5amp bias, so if you ended up with something equivalent to that, how bad would that be...
Don't skimp on transistors, heatsinks, transformers, capacitors etc and all will be well.
The KSA100 was just as tough as the 50, it was used as a reference amp for many magazines for quite a while.
Stuart
Hmm, there are to many unknowns to really have a perfect answer to this. I, and I assume we, don't know what the actual rail voltage and bias were on a KSA50. I don't believe based on pictures I've seen it can have been class A into 4 ohms, with 45v rails, it was a beefy amp, but not that beefy...So my gut feeling is that the rails were lower than 45, though perhaps not as low as I'm planning to go.
Should you be worried? tough call, not knowing your thermal solution etc it's hard to know if you are building something that will be fragile...if you get toroids and the voltage is too high, unwrap the tape and unwind a couple of turns, ( you may have to unwind one entire secondary depending on how the secondaries are wound ) measure and repeat until you are happy. If you need 5v extra for the driver stage, wind a few extra turns around the core, measure, repeat etc... put them in series with the secondary and rectify and smooth the result. Toroids are a diy'ers dream component, not that E cores are bad, just not as easily modified...
Krell had very good results with the KSA100, which must have had 45v minimum, and more than 2.5amp bias, so if you ended up with something equivalent to that, how bad would that be...
Don't skimp on transistors, heatsinks, transformers, capacitors etc and all will be well.
The KSA100 was just as tough as the 50, it was used as a reference amp for many magazines for quite a while.
Stuart
Hi Jan-
On your website you mentioned the following:
"For those interested in arranging Group Buys on this PCB,
I can inform you some price ideas:
At 50 pcs. app. €8,00 or $10.00 per PCB
At 100 pcs. app. €4,00 or $5.00 per PCB"
Where are you sourcing the PCB's from and if I handle the shipping and payment details can we use your source for the production?
On your website you mentioned the following:
"For those interested in arranging Group Buys on this PCB,
I can inform you some price ideas:
At 50 pcs. app. €8,00 or $10.00 per PCB
At 100 pcs. app. €4,00 or $5.00 per PCB"
Where are you sourcing the PCB's from and if I handle the shipping and payment details can we use your source for the production?
rabstg said:
If you look on Jan's website, it mentions that all boards are sourced from the www.elprint.com site. I would guess the pricing he listed is an estimate from them.
The shipping on 100 boards should not be too bad either, usually when I order a box of Jensen caps it is about 25.00 - 30.00. The box of caps is a lot larger and heavier than even 200 boards. So shipping from Europe will only add add a quarter or so to the individual cost per board.
George
Panelhead
Yes..... That right
The prices are from elprint.com, and calculated using MACAOS
Prices are ex. VAT and freight!!!
Please note, that I have many other projects going on, so if some one else would be so kind to handle the group buy, I would be happy....
Please note that ALL can buy from elprint
Yes..... That right
The prices are from elprint.com, and calculated using MACAOS
Prices are ex. VAT and freight!!!
Please note, that I have many other projects going on, so if some one else would be so kind to handle the group buy, I would be happy....
Please note that ALL can buy from elprint
Hello,
I know it said somewhere in this thread that this was not meant to be a project for 'non experts' but it looks intriguing and I have some pieces already gathered that should be OK.
I have a pair of used Signal Transformer power transformers--(Model 68-8 (68 volts at 8 amps or 544 watt-- Two 34 volt secondaries. Input for 105, 115 and 125 volts. Will do plus and minus 44-52 volts depending on input tap used. 17lbs. each). I also have an 11" x 11" R-theta heatsink that is rated at 0.5oC/w/3". I don't have experience with Class A and how hot this clone amp would run. Do I have enough heatsink for 1 channel---or even for 2 channels? Is this a project that a careful non expert could tackle or would it be better to stick to something like a P101 or leach amp?
thanks,
Steve.
I know it said somewhere in this thread that this was not meant to be a project for 'non experts' but it looks intriguing and I have some pieces already gathered that should be OK.
I have a pair of used Signal Transformer power transformers--(Model 68-8 (68 volts at 8 amps or 544 watt-- Two 34 volt secondaries. Input for 105, 115 and 125 volts. Will do plus and minus 44-52 volts depending on input tap used. 17lbs. each). I also have an 11" x 11" R-theta heatsink that is rated at 0.5oC/w/3". I don't have experience with Class A and how hot this clone amp would run. Do I have enough heatsink for 1 channel---or even for 2 channels? Is this a project that a careful non expert could tackle or would it be better to stick to something like a P101 or leach amp?
thanks,
Steve.