Hi Mike,
I've been waiting for the MJL21195's to come in and test the lot. I've also been fighting with @%#&*!!@@^%$ Windows XP. I fear a reload clean, then days of reloading programs and restoring data. Watch me lose something important.
Microsoft support, yeah right!
No matter what the problem is, reload the OS. But first you have to prove it's not a driver!! Then you have to format and reload.
I'll keep you posted on my findings Mike.
-Chris
I've been waiting for the MJL21195's to come in and test the lot. I've also been fighting with @%#&*!!@@^%$ Windows XP. I fear a reload clean, then days of reloading programs and restoring data. Watch me lose something important.
Microsoft support, yeah right!
No matter what the problem is, reload the OS. But first you have to prove it's not a driver!! Then you have to format and reload.I'll keep you posted on my findings Mike.
-Chris
I'm interested in this amp now 😉
MikeB: Which SPICE models did you use to model this amp? I noticed that the OnSemi model for BD139 doesn't seem to respond correctly - the only way i could bias the circuit to 40mA with that model was to remove the 500R resistor. With the Fairchild model, it works OK.
I'd also want to increase the rails a bit as I have a 600VA 35-0-35 transformer sitting about. I spotted a "high power" version but noticed the LTP used BC546, which would fail at 50v rails...
anatech: whats up with your computer ? I've fixed more windows f***ups than I care to count, and have managed to rescue systems that other "experts" have said can only be fixed by reinstall.
MikeB: Which SPICE models did you use to model this amp? I noticed that the OnSemi model for BD139 doesn't seem to respond correctly - the only way i could bias the circuit to 40mA with that model was to remove the 500R resistor. With the Fairchild model, it works OK.
I'd also want to increase the rails a bit as I have a 600VA 35-0-35 transformer sitting about. I spotted a "high power" version but noticed the LTP used BC546, which would fail at 50v rails...
anatech: whats up with your computer ? I've fixed more windows f***ups than I care to count, and have managed to rescue systems that other "experts" have said can only be fixed by reinstall.
borys said:
Yes, it is much better now.
What is your averaging and number of FFT samples?
It would be a soundcard noise now, I agree.
jaycee you are right using higher traffo your going to stress
this design causing major changes in tis parameters
john
this design causing major changes in tis parameters
john
Hi Borys, yes looks much better now ! Did you use the onboard sound ? At which powerlevel did you make the measurement ?
Chris, the MJL21196/5 definitely needs matching, and needs in symasym bias > 100ma, or you get audible distortions. (The typical Solid State - sssingerssss)
Have you tried simply installing the latest DirectX ? (9.0c)
Jaycee, i used the supplied bd135 model, it works fine. The OnSemi bd139 model is useless, it has way off vbe's.
I can not recommend using symasym with that high voltage, it shouldn't be >40v. The best would be getting a 2*25v toroid. 35v-AC is not only a bit higher, 48v DC will blow the mpsa18 and heat up some the to92. The bc546 has 65v rating, but has to manage 3ma. With 36v it's already very warm.
This high power version is not finished and does not work...
Mike
Chris, the MJL21196/5 definitely needs matching, and needs in symasym bias > 100ma, or you get audible distortions. (The typical Solid State - sssingerssss)
Have you tried simply installing the latest DirectX ? (9.0c)
Jaycee, i used the supplied bd135 model, it works fine. The OnSemi bd139 model is useless, it has way off vbe's.
I can not recommend using symasym with that high voltage, it shouldn't be >40v. The best would be getting a 2*25v toroid. 35v-AC is not only a bit higher, 48v DC will blow the mpsa18 and heat up some the to92. The bc546 has 65v rating, but has to manage 3ma. With 36v it's already very warm.
This high power version is not finished and does not work...
Mike
Ok... yeah I did think symasym wouldn't work too good on the higher voltage... was just looking for something with twin outputs to use with that transformer. The Leach Amp seems a bit over the top.
For anyone else simulating symasym, use the BD135_137_139 model from Fairchild!
For anyone else simulating symasym, use the BD135_137_139 model from Fairchild!
PMA
My sample rate is 9600, fft sample 32k and averaging is 1 (linear).
I have problem with signal wireing, I must change the wires to much better. IMHO measurments could improve much more - I must work on this.
MikeB
I have 3 PC's in home. Now I have used Audigy 24bit (oem). Load was 8,24R and output was 10V - aprox 12W.
My sample rate is 9600, fft sample 32k and averaging is 1 (linear).
I have problem with signal wireing, I must change the wires to much better. IMHO measurments could improve much more - I must work on this.
MikeB
I have 3 PC's in home. Now I have used Audigy 24bit (oem). Load was 8,24R and output was 10V - aprox 12W.
Hi jaycee,
Man, I'd love a hand with this! If you can help it will save me a ton of work.
I am running Direct X 9c (4.09.0000.0904) on Windows XP Home 5.1, build 2600. The system had been perfectly stable for a while until the June 16 updates were done. Since that time I've had no sound from either of my two sound cards. One is the built in Intel (I think) running on the Soundmax driver, the other was a Creative Live! 24 bit. I could get sound directly (not any more) in diagnostics, but the Microsoft Synthesizer failed (no driver loaded).
Tech support consisted of trying to prove it was a third party driver problem. Once we proved it was a kernel problem they threw up their hands and told me to reload the system. I was directed to do a repair install. They sent me a new disc and I repeated the process. This resulted in no sound from diagnostics and a number of kernel drivers that can't load (as there is a driver loaded already!). Mostly it's Microsoft Kernel Acoustic Echo Canceller.
I have since installed two different sound cards in place of the Live! 24, currently I have an X-Fi ExtremeMusic installed. There are no device conflicts showing up. The system is a P4 running at 1.8 GHz with 384 MB of ram. It does take longer to load (not surprisingly). I have removed all soundcards and drivers. I have re installed them one at a time a few times with no success. They are indicated in system devices as working properly.
You can contact me directly at bhome at sympatico dot ca . I am about to attempt to reload the system on a smaller drive to see if it's successful.
Thanks for any help!
-Chris
Man, I'd love a hand with this! If you can help it will save me a ton of work.
I am running Direct X 9c (4.09.0000.0904) on Windows XP Home 5.1, build 2600. The system had been perfectly stable for a while until the June 16 updates were done. Since that time I've had no sound from either of my two sound cards. One is the built in Intel (I think) running on the Soundmax driver, the other was a Creative Live! 24 bit. I could get sound directly (not any more) in diagnostics, but the Microsoft Synthesizer failed (no driver loaded).
Tech support consisted of trying to prove it was a third party driver problem. Once we proved it was a kernel problem they threw up their hands and told me to reload the system. I was directed to do a repair install. They sent me a new disc and I repeated the process. This resulted in no sound from diagnostics and a number of kernel drivers that can't load (as there is a driver loaded already!). Mostly it's Microsoft Kernel Acoustic Echo Canceller.
I have since installed two different sound cards in place of the Live! 24, currently I have an X-Fi ExtremeMusic installed. There are no device conflicts showing up. The system is a P4 running at 1.8 GHz with 384 MB of ram. It does take longer to load (not surprisingly). I have removed all soundcards and drivers. I have re installed them one at a time a few times with no success. They are indicated in system devices as working properly.
You can contact me directly at bhome at sympatico dot ca . I am about to attempt to reload the system on a smaller drive to see if it's successful.
Thanks for any help!
-Chris
Hi Mike,
-Chris
I always match everything. Unless I am specifically testing a circuit for the effects of mismatched devices. I feel that is a valuable test.
-Chris
jaycee said:
Carlos, who championed the Explendid amp also likes this one alot - not much more complicated to build than Mike's amp. He show's plans for a 50V rail version.
http://www.zen22142.zen.co.uk/Circuits/Audio/gem100.htm
Sheldon
Sheldon said:
I liked his idea of not using frequency compensation as so it state in the text yet the schematic showes the typical 22pF phase Lag capacitor so often used, whats up with that?
Re: Install a 220uf capacitor in your input and listen
Hello, Carlos:
I know you written this mean us to test my own action,
and I do believe this is very good to ourself because of this touchable actions, and this will impression us deeply.
But not all of us have the condition you mentioned ,
so would you please give out the result ?
There's hardly any different occurs: we can still test this whenever we got this condution.
Thank you very much !
P.S. : I was very appreciate your dedicate to the Amps, really, very very very !
destroyer X said:
Hello, Carlos:
I know you written this mean us to test my own action,
and I do believe this is very good to ourself because of this touchable actions, and this will impression us deeply.
But not all of us have the condition you mentioned ,
so would you please give out the result ?
There's hardly any different occurs: we can still test this whenever we got this condution.
Thank you very much !
P.S. : I was very appreciate your dedicate to the Amps, really, very very very !
FWIW, attached is a minor re-arrangement of the Sym 5 pcb layout of MikeB’s popular design.
Rccs is an alternative CCS feed. A school of thought suggests some benefit from feeding from a clean (signal) ground?? (Don’t subscribe to it myself, but both are available for comparison, which I haven’t done).
I assembled it to use as a “reference” against which to compare other pcb layouts I’ve built of a somewhat extended “MikeB-like” topology, so I didn’t want to deviate too far from the original.
Regards,
Brian.
Rccs is an alternative CCS feed. A school of thought suggests some benefit from feeding from a clean (signal) ground?? (Don’t subscribe to it myself, but both are available for comparison, which I haven’t done).
I assembled it to use as a “reference” against which to compare other pcb layouts I’ve built of a somewhat extended “MikeB-like” topology, so I didn’t want to deviate too far from the original.
Regards,
Brian.
thanks alot for that layout, I am doing a similer topology however it is for headphones. I have just got around to the layout aspect and yours is a great starting reference.Pingrs said:
Hi Brian, i'd really be interested in your scopeshots... 😉
About the PCB, nice work !
About the Rccs, that's a big nono. You would feed ~1.6ma into the signal ground, contaminating it with supply garbage and causing a DC-offset. Also it would kill the PSRR. If you don't want to connect Rccs to powergnd, you might connect it to the +Rail (and double the 22k). This has also the advantage of a more balanced load to the rails.
And, do you like the sound of symasym ?
Ppl, don't you think that is a bit overkill for a headphone amp ?
The way you have cascoded the vas, you should also cascode the input diffamp, that heavily increases quality, giving a nearly constant Vds to the jfets. (exploding PSRR for lower freqs and increasing dynamics)
BTW, i do love that topology...
Mike
About the PCB, nice work !
About the Rccs, that's a big nono. You would feed ~1.6ma into the signal ground, contaminating it with supply garbage and causing a DC-offset. Also it would kill the PSRR. If you don't want to connect Rccs to powergnd, you might connect it to the +Rail (and double the 22k). This has also the advantage of a more balanced load to the rails.
And, do you like the sound of symasym ?
Ppl, don't you think that is a bit overkill for a headphone amp ?
The way you have cascoded the vas, you should also cascode the input diffamp, that heavily increases quality, giving a nearly constant Vds to the jfets. (exploding PSRR for lower freqs and increasing dynamics)
BTW, i do love that topology...
Mike
I think it would make an awesome headphone amp if you can get the gain down low enough so you don't damage your hearing. 
This is one of the most detailed amps I have ever heard. It would be incredible through a good set of cans.
Blessings, Terry
This is one of the most detailed amps I have ever heard. It would be incredible through a good set of cans.
Blessings, Terry
MikeB said:Hi Brian, i'd really be interested in your scopeshots... 😉
About the PCB, nice work !
About the Rccs, that's a big nono. You would feed ~1.6ma into the signal ground, contaminating it with supply garbage and causing a DC-offset. Also it would kill the PSRR. If you don't want to connect Rccs to powergnd, you might connect it to the +Rail (and double the 22k). This has also the advantage of a more balanced load to the rails.
And, do you like the sound of symasym ?
Ppl, don't you think that is a bit overkill for a headphone amp ?
The way you have cascoded the vas, you should also cascode the input diffamp, that heavily increases quality, giving a nearly constant Vds to the jfets. (exploding PSRR for lower freqs and increasing dynamics)
BTW, i do love that topology...
Mike
hI
look more closly at the input stage it is a pair of JFETS cascoded to BJT's in addition to your noted attrabutes of a cascoded input stage i would add ultra low and constant input bias current that stays steady regardless of input swing and the reduced miller effect. Now is this overkill for a headphone Amp no i think headphone Amps are way to simple and not given the attention of High end Lodspeaker amps so my intent is to bring High end Loudspeaker Amp topolgy to Headphones and the result is fantastic. Thanks for your interest.