Hi Padamiecki,
I don't think it really matters since there are no ground loops. As long as you don't have them you're OK. But this thing about large currents with half waves flowing around the frontend inducing large even harmonics, that's something to think of.
As I recall, Mr. Pavel's PCB also had the same issue. I wonder what are his experiences.
Best regards,
Nick
I don't think it really matters since there are no ground loops. As long as you don't have them you're OK. But this thing about large currents with half waves flowing around the frontend inducing large even harmonics, that's something to think of.
As I recall, Mr. Pavel's PCB also had the same issue. I wonder what are his experiences.
Best regards,
Nick
Nick, looks okay ! What was wrong with my layout ?
Dytln, changing the input cap to 10uF is perfectly okay, maybe you can even use 22uF elyt ?
Mike
Dytln, changing the input cap to 10uF is perfectly okay, maybe you can even use 22uF elyt ?
Mike
hi Quattor,
With your desgin PCB you may be able to rotate the 1000uF filter caps 90 degrees to get shorter and straighter tracks. Watch out these caps don't get in the way of the mounting bolts for the output transistors.
The vbias transistor is not inline with the other transistors. This may make mounting to the heatsink difficult. I guess you are actually going to mount this transistor independent of the PCB, on the heatsink closer to one of the output transistors?
The feedback takeoff point worries me because it in not symmetrical between the two Re resistors, but its probably just a case of me reading too many books.
hi Michael,
There's probably nothing "wrong" with your layout as it is a well proven successful design. I guess Quattor feels the need to "personalise" your design from his previous audio experiences. I can understand that. 😀 I wouldn't get too worried.
regards
With your desgin PCB you may be able to rotate the 1000uF filter caps 90 degrees to get shorter and straighter tracks. Watch out these caps don't get in the way of the mounting bolts for the output transistors.
The vbias transistor is not inline with the other transistors. This may make mounting to the heatsink difficult. I guess you are actually going to mount this transistor independent of the PCB, on the heatsink closer to one of the output transistors?
The feedback takeoff point worries me because it in not symmetrical between the two Re resistors, but its probably just a case of me reading too many books.
hi Michael,
There's probably nothing "wrong" with your layout as it is a well proven successful design. I guess Quattor feels the need to "personalise" your design from his previous audio experiences. I can understand that. 😀 I wouldn't get too worried.
regards
Hi everyone,
Mike, there’s nothing wrong with your layout. On the contrary.
Fist of all, few elements available to me were not fitting within your design (smaller lead distance on the PCB than with actual elements, capacitors before everything else). Secondly, I really don’t like to place resistors and capacitors standing on the PCB. Furthermore, jumpers are something that I don’t like at all. Unless they are really, really necessary. Last, but not least, as Mr. Erskine said, I needed to "personalize" your design from my own previous audio experiences. And it’s too much fun for me.
I acknowledge your work in full, and I’m doing this for my personal pleasure only. Hope you don’t take this against me.
Anyway, that BD139 is not in-line with other trannies simply because it’s in TO126 case which has smaller thickness then TO220 and TOP-3 mounted there. That’s why I presumed its back side will be just nicely lined up with other “thicker” trannies. Tomorrow I’ll check that in real life, but I think I’m OK with this setup.
Mr. Erskine, maybe we were reading the same books. I’m worried as well that the feedback takeoff point is not symmetrical between the two Re resistors. In my previous constructions, emitter resistors were directly soldered to output trannies, and their “meeting” point had two leads – one to the speaker and one was the feedback lead. What’s more, resistors were actually made of resistive wire, pure flat, no inductance at all. Trying to compromise a bit, I’ve took one lead on the PCB for the feedback and one for the RC combination. Placing some good and thick jumper will ease my mind a bit regarding this issue. Anyway, too many commercial designs a far, far, away even from this solution, and yet, play such a nice music. I’ll settle with this for now.
About those 1000uF/50V electrolytic capacitors, I’ll see to it tomorrow. I’ve been thinking of this already trying to refine this design a bit more.
Best regards to ya all,
Nick
Mike, there’s nothing wrong with your layout. On the contrary.
Fist of all, few elements available to me were not fitting within your design (smaller lead distance on the PCB than with actual elements, capacitors before everything else). Secondly, I really don’t like to place resistors and capacitors standing on the PCB. Furthermore, jumpers are something that I don’t like at all. Unless they are really, really necessary. Last, but not least, as Mr. Erskine said, I needed to "personalize" your design from my own previous audio experiences. And it’s too much fun for me.
I acknowledge your work in full, and I’m doing this for my personal pleasure only. Hope you don’t take this against me.
Anyway, that BD139 is not in-line with other trannies simply because it’s in TO126 case which has smaller thickness then TO220 and TOP-3 mounted there. That’s why I presumed its back side will be just nicely lined up with other “thicker” trannies. Tomorrow I’ll check that in real life, but I think I’m OK with this setup.
Mr. Erskine, maybe we were reading the same books. I’m worried as well that the feedback takeoff point is not symmetrical between the two Re resistors. In my previous constructions, emitter resistors were directly soldered to output trannies, and their “meeting” point had two leads – one to the speaker and one was the feedback lead. What’s more, resistors were actually made of resistive wire, pure flat, no inductance at all. Trying to compromise a bit, I’ve took one lead on the PCB for the feedback and one for the RC combination. Placing some good and thick jumper will ease my mind a bit regarding this issue. Anyway, too many commercial designs a far, far, away even from this solution, and yet, play such a nice music. I’ll settle with this for now.
About those 1000uF/50V electrolytic capacitors, I’ll see to it tomorrow. I’ve been thinking of this already trying to refine this design a bit more.
Best regards to ya all,
Nick
PMA said:
Can you explain what will happen if I change Input caps value with bigger value ?
Can this sym headamps use as pre-amps ??
There is not much difference in sound (in fact no difference) if you change low frequency -3dB corner from 1.53 Hz to 0.72 Hz.
Perceiving of bass depends on drivers (headphones) and also on CD player used, though it does not sound reasonable. Of course CD player freq response is flat. But perceiving of bass strongly depends on jitter, HF contamination etc. Impossible to answer simply.
Perceiving of bass depends on drivers (headphones) and also on CD player used, though it does not sound reasonable. Of course CD player freq response is flat. But perceiving of bass strongly depends on jitter, HF contamination etc. Impossible to answer simply.
dytln_02 said:
Not really, symasym is a feedback monster, the large feedback is mainly needed for the outputstage.
As Pavel already said, lowering the gain of this amp is very difficult, it will need lot of changes to be suitable. Then there are other proven designs, more suitable to be used as preamp.
Pavel, i know that increasing the input cap to 10uF shouldn't change the bass, but it does. With 10uF i had a bit more punch in the deep bass (<50hz), it might be due to some damping from the signal source ?
Or the phaseshift caused by the inputcap for low freqs ?But i thought of using a 10uf film cap as overkill, so decided to use the smaller 4.7uf, the 10uf costing ~4 times more. (10uf is the biggest film cap easily available from wima)
Mike
MikeB said:
Not really, symasym is a feedback monster, the large feedback is mainly needed for the outputstage.
As Pavel already said, lowering the gain of this amp is very difficult, it will need lot of changes to be suitable. Then there are other proven designs, more suitable to be used as preamp.
Pavel, i know that increasing the input cap to 10uF shouldn't change the bass, but it does. With 10uF i had a bit more punch in the deep bass (<50hz), it might be due to some damping from the signal source ?
But i thought of using a 10uf film cap as overkill, so decided to use the smaller 4.7uf, the 10uf costing ~4 times more. (10uf is the biggest film cap easily available from wima)
Mike
I have change Solen 4.7uF with Bennic 10uF.
Using Grado SR80 Headamps I hear More punch in the deep bass (<50hz). but it's loose punch bass at higher freq.
How get more punch in the deep bass (<50hz) without loosing punch bass at higher freq.
Quattor said:
Yes, and something to worry about - this I think is the only potential flaw in the PCB design. IIRC, there was a recomendation to connect the points under the OCB with a thick wire link. Whatthis does is effectively introduce part of the oposing half wave into the pcb tracks going round the input stage. The cancelation is not perfect, but it's better to have induction of something similar to the input signal, as oposed to it's half-wave rectified version. The cancelation depends on how low the resistance of the wire link is compared to the PCB tracks. It may be even possible that putting on the wire link, and cutting one of the tracks to the ground point reduces the problem further.
However, for a real solution one really needs to use jumpers (I know, I know...). Ideally, you could make some space between the 47k/220p components, and put two wire links with the ground point between them in the immagined line connecting the appropriate terminals of the 1000u/5-V caps. Preferably, you would keep the output point close to the ground point. This would be the local power ground and the place where you want to put a twisted pair of wires to the load. This way the input stage is spared from induction effects.\When I design PCBs, it's almost a matter of honor to not use jumpers 🙂 but sometimes they do solve a lot more problems than they introduce.
Regarding the NFB pick-off point, if you want to be sure about it, you can move the track down so it actually connects with the one going to the zobel network, just under the pad used for the output. This way you practically completely eliminate any half-wave rectification effect. The small penalty is that the FB tracks shares about a mm of current path for the zobel net, but compared to crossover effects due to pickof point, the zobel is a tame and linear load, that sees the full output signal, so you shouldn't really have a problem at all.
dytln_02 said:
I have change Solen 4.7uF with Bennic 10uF.
Using Grado SR80 Headamps I hear More punch in the deep bass (<50hz). but it's loose punch bass at higher freq.
How get more punch in the deep bass (<50hz) without loosing punch bass at higher freq.
Once more - it is not about low frequency corner of 1,4Hz or 0.7Hz, it makes absolutely no difference.
Bass perception is closely related to midrange clarity, jitter etc. Though Michael wrote about "punch" when changed input cap to 10u, I am very sure the right reason was somewhere else.
If you really want 10 uF, then mount the 4u7 unit as normal and connect another one directly underneath on the other side of the board. Problem solved!.
I can see there are many things to play with on this design!
-Chris
I can see there are many things to play with on this design!
-Chris
Yes, and it is very hard to understand what makes an amplifier sound as it sounds... 
Pavel, yes the low corner freq is not the problem, i once tried 100nf as input cap and was very suprised about the sound... The bass was weak, but the overall sound far from boring, just the deep bass missing.
Mike
Pavel, yes the low corner freq is not the problem, i once tried 100nf as input cap and was very suprised about the sound... The bass was weak, but the overall sound far from boring, just the deep bass missing.
Mike
Install a 220uf capacitor in your input and listen
You will see if has difference or not.
Difference can be something annoying, of course it can, but you will perceive.
Those things are easy to check..just install for some time and go listening for some minutes...them...remove it.
Also you can try two 470uf, negative conected to negative...use the positive sides only.
If you are not enougth surprised, now include a 220N capacitor in parallel with this input combination....and listen.
So easy to confirm those things...just do not be blocked because of calculations... move and check by yourself!
I found people that was sure that golden RCA connectors could produce better sound...and the guy go inside that belief for years long...when is easy to check...install in parallel other connector (not Golden unit) and you will see that those things are magic or not...i made the test....so...i know if this is magic or not.
The ones that did not tested...i am afraid that do not know...just can imagine.
Also speaker cables, another expensive myth...those monster heavy gauge ones...small capacitance cables (when zobel has much more capacitance).....small resistance cable (when normal speaker cable is so small that resistance is hard to measure and emitter resistors are much bigger to worry about the cable)...and low inductance (when you already have enormous inductance from your output filter)....compare it with simple wires (not parallell)..use A to B switching.... you will notice....maybe notice nothing...and this is a conclusion.
regards,
Carlos
You will see if has difference or not.
Difference can be something annoying, of course it can, but you will perceive.
Those things are easy to check..just install for some time and go listening for some minutes...them...remove it.
Also you can try two 470uf, negative conected to negative...use the positive sides only.
If you are not enougth surprised, now include a 220N capacitor in parallel with this input combination....and listen.
So easy to confirm those things...just do not be blocked because of calculations... move and check by yourself!
I found people that was sure that golden RCA connectors could produce better sound...and the guy go inside that belief for years long...when is easy to check...install in parallel other connector (not Golden unit) and you will see that those things are magic or not...i made the test....so...i know if this is magic or not.
The ones that did not tested...i am afraid that do not know...just can imagine.
Also speaker cables, another expensive myth...those monster heavy gauge ones...small capacitance cables (when zobel has much more capacitance).....small resistance cable (when normal speaker cable is so small that resistance is hard to measure and emitter resistors are much bigger to worry about the cable)...and low inductance (when you already have enormous inductance from your output filter)....compare it with simple wires (not parallell)..use A to B switching.... you will notice....maybe notice nothing...and this is a conclusion.
regards,
Carlos
destroyer X said:
Carlos, not a myth... Some time ago i replaced my standard 4mm Oehlbach cable with Kimber 4TC, (both 2 x 3meters) the difference was stunning !
On my Yamaha amp with mission speakers the sound improved heavily, much clearer trebles, better bass... (not subtile)
On the other hand, with my Musical Fidelity A1X connected to Epos ES12 (Hard to drive), the Kimber cable was a disaster, all trebles heavily distorted...
BTW, symasym is happy driving these Epos speakers...
Mike