Upupa Epops said:
U-Papa, I would have expected a comment like that from Jacco.😀
BTW, Stuart, this is good news, finally the seeds are germinating.🙂
Before you crank up the bias to class-A, can you please post the class-AB 8 ohm clipping with 100v unloaded rails? I want to pep myself up.... Oo Oo Oo more power.
NUTTTR said:
Aaron,
I'd take whatever goodwill your better half accords to you now and milk it for she will lose it anyday. 😉
work...
This is not my first work from the diy perspective, but I've never done any of this stuff professionally. I took a year or so of electronics back when I was at university, but that was all theory, nowadays I try and concentrate more on theory as it can be practised...
I try and make what I do look good, but I concentrate on functionality first...
Pavel, I guess I'm hoping your question is asked from a positive point of view, if you think there's something wrong, please let me know...
Stuart
This is not my first work from the diy perspective, but I've never done any of this stuff professionally. I took a year or so of electronics back when I was at university, but that was all theory, nowadays I try and concentrate more on theory as it can be practised...
I try and make what I do look good, but I concentrate on functionality first...
Pavel, I guess I'm hoping your question is asked from a positive point of view, if you think there's something wrong, please let me know...
Stuart
I have only positive wiews at DIY, Stuart, my friend 😎 . First directions by building of amp, is to make all conductive trace ( wire or layout ) shortest as possible. Long wire have higher inductance, which can cause unstability. If you have unstabil amp, you must make higher value of compensation, which often lead to higher distortion and bad frequency parametres. This problems rise, if you are using modern fast devices, which is cause of Krell 😉 . Second " parameter " is reliability, in this case mechanical reliability. All devices must be good fitted, threedimensional " on air " construction is highly unreliable, 'cos is easy " shakeable " ( I don't know correct word 
). Etc and etc. Try to think a bit about my words 😎 .
). Etc and etc. Try to think a bit about my words 😎 .
Upupa Epops said:
Pavel, i am a nut, dont you know ?
Just as some have a passion for developing the perfect economic AB amplifier, others find less energy-friendly topologies intrigueing.
I fell in love with class A amplifiers just after i constructed my first diy power amplifier.
I loved the Western 300B triode amplifiers, Jean Hiraga's class A and JH50 pushpull EL34 50 watt class A amplifiers, loved the Accuphase P102 class A model, Thresholds and Forte's.
You of all should know that it is addictive.
Actually i am a scrooge, i split a peanut in two to eat the other half tomorrow.
I do not need to rely on just one amplifier for listening, i'll promise you i'll run the KSA on night-power.
But wouldnt it be fun to know how an amplifier that never leaves class A sounds like ?
I often feel rich.
I love your sense of humor, it is very kind of you to offer so much advice on the threads.
Here is a picture of two of the heatsinks i found for the Krell.
They are really huge, placed a quartet of Sanken's on them to show the size.
Attachments
Holeeee! Jacco those sinks are rather perky... with a Papst (judging the picture) that baby could probably dissipate 2kW. Look at the thickness of the mounting area...!
By the way do you remember those Adcom 22000uF caps I had that we thought we not "that good", ... turns out they are Cornell Dubilier not Nippon Chemi. So... **sighs**
CD also bought out Mallory lytic division (Tant div got bought by Vishay) so they are not as ordinary as we thought. Also they have a surge rating of 125v so I am planning to use them in Stuart's mega powered 99-100vdc (unloaded) rails KSA-500... yeeeeeehaaaa! 😀
Pavel do you sell Amplifier kits etc or just help us hopeless cases. 😉
By the way do you remember those Adcom 22000uF caps I had that we thought we not "that good", ... turns out they are Cornell Dubilier not Nippon Chemi. So... **sighs**
CD also bought out Mallory lytic division (Tant div got bought by Vishay) so they are not as ordinary as we thought. Also they have a surge rating of 125v so I am planning to use them in Stuart's mega powered 99-100vdc (unloaded) rails KSA-500... yeeeeeehaaaa! 😀
Pavel do you sell Amplifier kits etc or just help us hopeless cases. 😉
The base is 1.25" , K-adcom.
Putting four 4.75" vents on the top will not cut it, probably will need to look for 5.5" models.
(and +200kuF/side of caps, sigh )
Those arent the ones you bought a doo-poo load of, A ?
The Adcoms you showed pictures of could never have been Chemicon's.
All Adcom capacitors i have seen so far on Ebay have the NorthAmerican look, i smell Japsters a mile away.
(i slice my own sushi with original Global knifes, btw)
Putting four 4.75" vents on the top will not cut it, probably will need to look for 5.5" models.
(and +200kuF/side of caps, sigh
)Those arent the ones you bought a doo-poo load of, A ?
The Adcoms you showed pictures of could never have been Chemicon's.
All Adcom capacitors i have seen so far on Ebay have the NorthAmerican look, i smell Japsters a mile away.
(i slice my own sushi with original Global knifes, btw)
shakiness...
Pavel,
I think you are right about doing point to point wiring under production conditions, but this is a prototype, I want access to everything while I am building and testing. Long term I doubt I'll need an amp of this size, and chances are I'll dismantle the thing once I am finished, unless someone makes me an offer I can't refuse...
Plus, my guess is that one channel of the completed amp will weigh >30kg, so lots of shaking is an unlikely scenario...movement of any sort won't be trivial...
Stuart
Pavel,
I think you are right about doing point to point wiring under production conditions, but this is a prototype, I want access to everything while I am building and testing. Long term I doubt I'll need an amp of this size, and chances are I'll dismantle the thing once I am finished, unless someone makes me an offer I can't refuse...
Plus, my guess is that one channel of the completed amp will weigh >30kg, so lots of shaking is an unlikely scenario...movement of any sort won't be trivial...
Stuart
jacco vermeulen said:The base is 1.25" , K-adcom.
Putting four 4.75" vents on the top will not cut it, probably will need to look for 5.5" models.
(and +200kuF/side of caps, sigh
Those arent the ones you bought a doo-poo load of, A ?
The Adcoms you showed pictures of could never have been Chemicon's.
All Adcom capacitors i have seen so far on Ebay have the NorthAmerican look, i smell Japsters a mile away.
(i slice my own sushi with original Global knifes, btw)
Yes, I got like 20 of then. Actually the Nippon Chemicon looks VERY much like them... same top / bottom/ material/ heatshrink color etc. A large distributor (capacitor King or cap12 on ebay) said they use the same factory or something...
Arif, whats the plan for the bias on that KSA ?
Please post it, then Pavel can bug you instead of me.
I suspect Upopo meant "erectonic" but made a typo, i would never ask such questions.
What kind of emitter resistors Stuart is using i do wonder, the 0.68's are cement types ?
Please post it, then Pavel can bug you instead of me.
I suspect Upopo meant "erectonic" but made a typo, i would never ask such questions.
What kind of emitter resistors Stuart is using i do wonder, the 0.68's are cement types ?
Trusty friend,
The plan for the bias depends on the heatsink capability. I do hope I can do at least 2 watts Pure class-A, but realistically with 100vdc rails (and I am on the rd and do not have access to my trusty class-A calculator) the idle dissipation will not be funny.
If Pavel bugs me about the class-A limits of my KSA-500, I will refer him to NP's new venture... ;-) ... You know First twit.
As for Stuart's Erectter resistors, I hear Cement makes em hard and consistent... also dissipates some serious heat.😉 But personally I am not too much into erotic components, I mean exotic components..... I am happy with the gobs of 0.68's our friend sent us.
The plan for the bias depends on the heatsink capability. I do hope I can do at least 2 watts Pure class-A, but realistically with 100vdc rails (and I am on the rd and do not have access to my trusty class-A calculator) the idle dissipation will not be funny.
If Pavel bugs me about the class-A limits of my KSA-500, I will refer him to NP's new venture... ;-) ... You know First twit.
As for Stuart's Erectter resistors, I hear Cement makes em hard and consistent... also dissipates some serious heat.😉 But personally I am not too much into erotic components, I mean exotic components..... I am happy with the gobs of 0.68's our friend sent us.
Upupa Epops said:
I've never understood the obsession with having ultra fast output transistors and then limit the frequency at INPUT to couple hundred kHz or so.
Yes, MJL21193/4 are faster and more linear than 15003/4 IN SOME APPLICATIONS. Used in output stage as common collector with plenty of local feedback via Re, and plenty of GLOBAL feedback if one wants (none in my case, I leave output stage completely outside the feedback loop and it works great), I doubt anyone would hear any difference.
Heck, on my amp 0.5MHz sine wave looks just fine, no triangulation of any sort.
And NO hf oscillations either 🙂
Bratislav
resistors...
The only things I know about the resistors are as follows:
10w, 0r68, 5%, Made by panasonic, shipped in boxes of 4 bags of 25 resistors.
I bought a couple of thousand on ebay a while back, basically because it was cheaper than buying the 36 I needed from mouser.
When I was visiting with him in england, pinkmouse commented they may be stacked film, certainly they are not like any wirewound resistor I've seen before. As far as I can tell they are not significantly inductive, my LC meter can resolve down to 1uH and doesn't read anything at all.
If, ultimately, people really can't stand them I'll give everyone back their money...
Stuart
The only things I know about the resistors are as follows:
10w, 0r68, 5%, Made by panasonic, shipped in boxes of 4 bags of 25 resistors.
I bought a couple of thousand on ebay a while back, basically because it was cheaper than buying the 36 I needed from mouser.
When I was visiting with him in england, pinkmouse commented they may be stacked film, certainly they are not like any wirewound resistor I've seen before. As far as I can tell they are not significantly inductive, my LC meter can resolve down to 1uH and doesn't read anything at all.
If, ultimately, people really can't stand them I'll give everyone back their money...
Stuart
Stuart,
I have not cracked one open yet, but from the dimensions, if I were to guess, they seem like metal film type (which are non inductive IIRC).
Good luck on your KSA-500 testing.... Let me know if it hits 500W otherwise it is no fun calling it a KSA-497....
I have not cracked one open yet, but from the dimensions, if I were to guess, they seem like metal film type (which are non inductive IIRC).
Good luck on your KSA-500 testing.... Let me know if it hits 500W otherwise it is no fun calling it a KSA-497....
emitter resistors Stuart is using
When i have back the HP impedance measurement gear i will do some measurements and show it on DIY.
Loek
When i have back the HP impedance measurement gear i will do some measurements and show it on DIY.
Loek
Bratislav said:
I never understood why cars get bigger engines and then a speed limiter is added.
Untill i hit the gas at the cross lights.
I used to be married to someone originating from Indonesia, even during summer she slept on an electric blanket.
The only she appreciated of me were my class A amplifiers, not the sight of a guy listening to music only dressed in shorts, i guess, bummer !
Bratislav
If you look back about ten pages in this thread you will see that when I put the scope on a 10K square wave with 15003/4s there was significant rounding at the corners. That's why I'm now recommending 1302/3281s for outputs, and I will try the 21193/4s as well when I get the time.
If you look back about ten pages in this thread you will see that when I put the scope on a 10K square wave with 15003/4s there was significant rounding at the corners. That's why I'm now recommending 1302/3281s for outputs, and I will try the 21193/4s as well when I get the time.
Al, rounding corners are not caused by 15003/4, but by too high compensation capacity and by input filter. 😉