Can someone shed some light on the need for thick cabling here? Last time I measured current demand was almost constant and below 100mA.
If you would power large caps on a varying load I could imagine the necessity for low cabling resistance. The thicker the conductors the more capacitive they get
If you would power large caps on a varying load I could imagine the necessity for low cabling resistance. The thicker the conductors the more capacitive they get
Its not a need per se, its rather a perfectionism towards a flattest possible, non inductive output impedance in the mOhm range that such regs are capable of. So capacitive is not a bad thing here, especially when outbalancing the rising self inductance due to length. The remote sensing nodes are a practical measure towards the same goal if the regs have to be away enough. There, the error amp senses the extra impedance as part of the load, and the gauges can be relaxed to a point. Must not pick up hash or create oscillations though. Twisting the sense wires is beneficial. Of course if it does not make a subjective change, in some build, its a non issue. I could easily tell the difference between having the reg in the psu box VS in the phono box for instance, eliminating enough a cable run. For smaller differences when inside the phono box, it must be checked out if its under some perception threshold in certain systems. It certainly can be measured, but its a hassle, and not in the simple DIY experimental level. Takes some well equipped guy, with experience in measurements. Its not a power amp as you say, but it has to retain max resolution for sub mV input signal and push it up to line level with no additives, so it pays off being thorough with it I guess.
I'd only add that with longer cables the resistance/capacitance/inductance start to create the usual filters. Whether it makes a difference for this current variance, it's debatable. I'd say it doesn't, but I'd still make the wires short and thick, just to not have that question sign in my mind. I would think the capacitance would depend more on the proximity of the two wires and the insulator properties, than on the thickness of the cable. The formula for a parallel plate capacitor is
C = 0.224 * (k * s * (n-1)) / d
where
C = capacitance
k = dielectric constant
s = area of one plate in square inches
n = number of plates
d = distance between plates
Assuming only two cables (plates) you can see that if the dielectric constant or area is large, then the capacitance will be large. If the distance is large, the capacitance will be small.
The thickness of the plates does not play a factor. For two cables I guess you can argue that the thicker they are, the large the area between them... but not really.
The inductance also doesn't seem to depend on the thickness of the cable.
Resistance on the other hand, it does.
There is a longer discussion about the effect of a little bit of resistance and a little bit of inductance in the wires between the shunt and the load, in the other thread (low voltage reg). I ran some simulations, and the output impedance seems very much affected by these parasitics. Again, whether this actually affects the sound, I don't know.
Just my 2c.
C = 0.224 * (k * s * (n-1)) / d
where
C = capacitance
k = dielectric constant
s = area of one plate in square inches
n = number of plates
d = distance between plates
Assuming only two cables (plates) you can see that if the dielectric constant or area is large, then the capacitance will be large. If the distance is large, the capacitance will be small.
The thickness of the plates does not play a factor. For two cables I guess you can argue that the thicker they are, the large the area between them... but not really.
The inductance also doesn't seem to depend on the thickness of the cable.
Resistance on the other hand, it does.
There is a longer discussion about the effect of a little bit of resistance and a little bit of inductance in the wires between the shunt and the load, in the other thread (low voltage reg). I ran some simulations, and the output impedance seems very much affected by these parasitics. Again, whether this actually affects the sound, I don't know.
Just my 2c.
Hi Ricardo,
Moving the connection between mosfets at the top (behind mosfets) you can move R1 right on mosfet's pin. Then you will see that the rest of the layout can be much more compact and clear.
Thanks for asking
For the time being, work and children (schools opened..) suck all my available time. But the last 3 months were very productive (diy wise..). Ahh I'm breaking in Pavel's Dispre II. For the Phono I have all parts ready and most of the "matching" is done. I need to finish with Dispre evaluation first in order to proceed with RIAA.
For the time being, work and children (schools opened..) suck all my available time. But the last 3 months were very productive (diy wise..). Ahh I'm breaking in Pavel's Dispre II. For the Phono I have all parts ready and most of the "matching" is done. I need to finish with Dispre evaluation first in order to proceed with RIAA.
Hi Ricardo,
Moving the connection between mosfets at the top (behind mosfets) you can move R1 right on mosfet's pin. Then you will see that the rest of the layout can be much more compact and clear.
Can you please post a picture or sketch ?
Ricardo
For the short distance (up to 10cm) we seem to agree on simplicity. Indeed does a long run of wire introduce drawbacks wrt impedance, sensitivity to external fields etc so -in that case- remote sensing would be beneficial.
Couple of years ago I bought this little book suggesting optimal copper wire diameter of 0.5 to 1.0mm for audio cables, because lager diameter introduces Time Smear in the low frequencies. As the regulator is interacting with the single ended RIAA-amp stages I considered the power wires as audio cables. Might be too enthusiastic there
Attachments
Last edited:
between a shunt reg and the circuit it powers. The closer and tighter the ties, the better. Shunt + Load = Love 
Lesson nº xyz
Remenber I had a "woosh" sound on the right channel ?
Well, after thinking about it, I connected the riaa right channel board to the left shunt, and left the right shunt disconnected.
Imediately the woosh disppeared and the sound became stable presenting a credible stereo image.
There is a problem with the right shunt....!
Now I lost a bit of dimension and power but gaining quitness I can now visualize the whole sound picture.
The low PSRR of the riaa benefits more from a low noise feed than sheer power.
The shunt plays a really important part on this design !
Ricardo
Remenber I had a "woosh" sound on the right channel ?
Well, after thinking about it, I connected the riaa right channel board to the left shunt, and left the right shunt disconnected.
Imediately the woosh disppeared and the sound became stable presenting a credible stereo image.
There is a problem with the right shunt....!
Now I lost a bit of dimension and power but gaining quitness I can now visualize the whole sound picture.
The low PSRR of the riaa benefits more from a low noise feed than sheer power.
The shunt plays a really important part on this design !
Ricardo
First suspects are the Zener and its cap, second set of suspects is the driving BC550 and its 2SK source. Unluckily Ricardo could not get the low noise red glass Zeners online yet. Those with the black plastic and silverish stripe that he has on the photo can be dodgy. Also some assymetry or grounding fault can plague one shunt when using 2 and a remote trafo/rectifier & main filter. Not necessarily by humming on one frequency, but boosting the noise floor well up as a whole. ''Woosh'' is low SNR. First must replace the Zener.
my pre is over
some photos
some photos
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
sorry I do not know how to properly post images on forum
http://img36.imageshack.us/img36/9218/spa1318.jpg
http://img29.imageshack.us/img29/1264/spa1317.jpg
http://img156.imageshack.us/img156/5611/spa1315.jpg
http://img33.imageshack.us/img33/1333/spa1319.jpg
http://img36.imageshack.us/img36/9218/spa1318.jpg
http://img29.imageshack.us/img29/1264/spa1317.jpg
http://img156.imageshack.us/img156/5611/spa1315.jpg
http://img33.imageshack.us/img33/1333/spa1319.jpg
Congratulations!
What exactly did you combine beyond the simplistic in there? Is it OK for hum, and did you have time to listen to it? How do you like it?
The offending zener was replaced and now there is no more "woosh"... both channels sound the same with a very low ripple at full volume...
It seems this component is critical for the "transparency" of the shunt.
With two shunts, I notice a much better transient response and wider soundstage (I believe this is due to the better separation provided by the dual mono setup)
I believe a glass low noise zener with a big Ruby ZLH would make a big difference.
Luca.... Splendid works !
Ricardo
It seems this component is critical for the "transparency" of the shunt.
With two shunts, I notice a much better transient response and wider soundstage (I believe this is due to the better separation provided by the dual mono setup)
I believe a glass low noise zener with a big Ruby ZLH would make a big difference.
Luca.... Splendid works !
Ricardo
Last edited: