Grounding scheme for dual mono supplies
Ok, I've though it through and done a little experimentation. I found I can get my phonoclone to hum, quite badly, if I disconnect the two TT ground wires (one runs from the arm, the other the motor/base assembly) from the phonoclone TT ground lug, and power on the turntable. With the turntable turned off, there is no hum even with the ground wires disconnected. I dont get any additional hum when the turntable motor is spinning compared to when it is stopped. Reconnecting the turtanble motor ground wire cures the problem completely. Recommening just the arm ground wire gets rid of 70% of the hum.
My turntable is direct drive, so all the above means is the hum is coming from the power transformer located underneath the platter. Properly shielding the offending source by grounding the surrounding metal case to the phonoclone chassis solves the problem.
When the ground wires are disconnected, the whole system is much more sentive to RFI and weirdness (I can kill the hum when I touch the case, for example, and clicking noises induced by the WiFi router sitting nearby is greatly enhanced.)
Conclusion: the arm and turntable should be fully shielded and grounded to the phonoclone case via the TT ground lug. It could well be that the phonoclone circuit is more susceptible to this interferance that other designs, due to the inverting, zero impedance input. I suspect this is so, but I cant quite grasp the whole situation well enough to explain it at this point. Those of you with DIY tables and arms seem to be having problems, while those of you with commercial decks seem to be OK ... limited statistics, but this is the trend so far. Commercial units tend to be properly shielded ... DIY units not always so well.
Lastly, there is only one way to properly connect the grounds when a single transformer is used. Its the way shown on the web page. It involves a compromise layout due to the fundamental decision to use one power supply for two separate circuits (ie. shared COM). Most of the time, however, its fine, and anyway there is no alternative. This is the "case-to-COM" scheme.
There are two ways to connect two power supplies. The first is just add a second unit in parallel, according to the standard "case-to-COM" scheme. All COM, both power supply and PCB, run to the case TT lug. If the system runs properly with one shared power supply in this configuration, adding a second will also work properly.
The second is "case-to-IN-" scheme, shown below and described in earlier posts. This is actually the textbook correct, no compromise way to do it, and a direct advantage of the full dual-mono design choice. (it will not work with a shared supply) Fran tried it and found it worked for him, I suggest Tube300 also try it soon. The easiest was to do it is remove the plastic spacer from the back of the RCA jack, but you can run wires from the RCA tab to the chassis TT lug if thats what you prefer.
Why should case-to-IN- be better? In principle connecting to the input return at the input jack should have more immunity from noise pickup, especially at RF. Its worth a try if you are having problems, but please note that its not addressing the real source of the problem, which is an improperly shielded and grounded turntable.
regards,
Richard
Ok, I've though it through and done a little experimentation. I found I can get my phonoclone to hum, quite badly, if I disconnect the two TT ground wires (one runs from the arm, the other the motor/base assembly) from the phonoclone TT ground lug, and power on the turntable. With the turntable turned off, there is no hum even with the ground wires disconnected. I dont get any additional hum when the turntable motor is spinning compared to when it is stopped. Reconnecting the turtanble motor ground wire cures the problem completely. Recommening just the arm ground wire gets rid of 70% of the hum.
My turntable is direct drive, so all the above means is the hum is coming from the power transformer located underneath the platter. Properly shielding the offending source by grounding the surrounding metal case to the phonoclone chassis solves the problem.
When the ground wires are disconnected, the whole system is much more sentive to RFI and weirdness (I can kill the hum when I touch the case, for example, and clicking noises induced by the WiFi router sitting nearby is greatly enhanced.)
Conclusion: the arm and turntable should be fully shielded and grounded to the phonoclone case via the TT ground lug. It could well be that the phonoclone circuit is more susceptible to this interferance that other designs, due to the inverting, zero impedance input. I suspect this is so, but I cant quite grasp the whole situation well enough to explain it at this point. Those of you with DIY tables and arms seem to be having problems, while those of you with commercial decks seem to be OK ... limited statistics, but this is the trend so far. Commercial units tend to be properly shielded ... DIY units not always so well.
Lastly, there is only one way to properly connect the grounds when a single transformer is used. Its the way shown on the web page. It involves a compromise layout due to the fundamental decision to use one power supply for two separate circuits (ie. shared COM). Most of the time, however, its fine, and anyway there is no alternative. This is the "case-to-COM" scheme.
There are two ways to connect two power supplies. The first is just add a second unit in parallel, according to the standard "case-to-COM" scheme. All COM, both power supply and PCB, run to the case TT lug. If the system runs properly with one shared power supply in this configuration, adding a second will also work properly.
The second is "case-to-IN-" scheme, shown below and described in earlier posts. This is actually the textbook correct, no compromise way to do it, and a direct advantage of the full dual-mono design choice. (it will not work with a shared supply) Fran tried it and found it worked for him, I suggest Tube300 also try it soon. The easiest was to do it is remove the plastic spacer from the back of the RCA jack, but you can run wires from the RCA tab to the chassis TT lug if thats what you prefer.
Why should case-to-IN- be better? In principle connecting to the input return at the input jack should have more immunity from noise pickup, especially at RF. Its worth a try if you are having problems, but please note that its not addressing the real source of the problem, which is an improperly shielded and grounded turntable.
regards,
Richard
Attachments
Smallest VSPS?
Hi there,
I've also build a VSPS. As I wanted it to plug in my converted Quad 33 I've build it very small. Just point to point on a piece of strip board. OPA 2134 as op-amps. Capacitors polyester and polypropylene. Gain adjustable by jumpers (to the best Quad tradition).
The sound is actually better than my previous try, a Rod Elliot project 6 (but comparison is not fair, I've used that one with an asymmetric power supply).
For details on the Quad 33 see this thread: http://www.diyaudio.com/forums/showthread.php?s=&threadid=79794
It might be interesting to know that the board is fed by a two stage power supply. First stage regulated to about 24V with LM317/337. Then a second stage to 16V again using LM317/337. On the board you see two caps two have some reservoir near the op-amps.
Greetz, MArco
Hi there,
I've also build a VSPS. As I wanted it to plug in my converted Quad 33 I've build it very small. Just point to point on a piece of strip board. OPA 2134 as op-amps. Capacitors polyester and polypropylene. Gain adjustable by jumpers (to the best Quad tradition).
The sound is actually better than my previous try, a Rod Elliot project 6 (but comparison is not fair, I've used that one with an asymmetric power supply).
For details on the Quad 33 see this thread: http://www.diyaudio.com/forums/showthread.php?s=&threadid=79794
It might be interesting to know that the board is fed by a two stage power supply. First stage regulated to about 24V with LM317/337. Then a second stage to 16V again using LM317/337. On the board you see two caps two have some reservoir near the op-amps.
Greetz, MArco
Attachments
I just assembled my kit #8 of the RC PhonoClone, and already I am impressed with the matching and careful packing 
I have been wondering what the minimum voltage supply is, I know you recommend a 12 Volt supply, but I have been considering batteries and would the ideal then be the equavalent to the rectified 17 volts that you would get from a regular supply. I have a 12V lead battery at hand, but would that be going too low?
I have been wondering what the minimum voltage supply is, I know you recommend a 12 Volt supply, but I have been considering batteries and would the ideal then be the equavalent to the rectified 17 volts that you would get from a regular supply. I have a 12V lead battery at hand, but would that be going too low?
The voltage regulators on the board are 10 volt types. The input voltage required to maintain regulation is a little over 2 V, so given that the battery output voltage can drop over the discharge cycle I dont think it would be a good idea to use 12 V batteries.
You can of course remove the regulators and short over the input-output pads, but we've been through all this before to conclude that a good linear power supply sounds better than batteries anyway.
You can of course remove the regulators and short over the input-output pads, but we've been through all this before to conclude that a good linear power supply sounds better than batteries anyway.
I'm going to built the VSPS. I was reading and saw that too large of caps make it sound worse?? I was going to use some 4700uF caps before the regulators to smooth out any ripple. How do larger caps affect sound quality? Wouldn't big caps on such a small load basically act like a battery?
Also, my amp has a 5K passive attenuator input. Would the standard VSPS design give me enough output to drive that? I'm thinking of adding a second stage with a 20K/1K resistor configuration for a little more output.
Also, my amp has a 5K passive attenuator input. Would the standard VSPS design give me enough output to drive that? I'm thinking of adding a second stage with a 20K/1K resistor configuration for a little more output.
You could try 2200uf -10R-2200uf... for improved ripple reduction, and also if you remember ohms law, when you use two parallel capacitors their resistive component becomes less...
I just remembered I never posted pictures of my VSPS... it is likely the largest one... (63V rail caps) will take some tommorrow.
I just remembered I never posted pictures of my VSPS... it is likely the largest one... (63V rail caps)
will take some tommorrow.
PCBs finally arrived... and the components?
Hello all,
I have to announce that I have officially received the PCBs! I have located all the components as well.
Question, I was not able to find the 4,7 uF cap for the coupling, instead can I use 10uF? It shouldn't make a difference, or I am wrong again?
I hope to start soon. I hope it snows so I can send my wife to the mountains for skiing so I can do this porject in peace!!!
Greetings to all!
Hello all,
I have to announce that I have officially received the PCBs! I have located all the components as well.
Question, I was not able to find the 4,7 uF cap for the coupling, instead can I use 10uF? It shouldn't make a difference, or I am wrong again?
I hope to start soon. I hope it snows so I can send my wife to the mountains for skiing so I can do this porject in peace!!!
Greetings to all!
mavallarino,
The 4,7 uF can be purchased from anywhere you can get Black Gate caps, Partconnexion being the obvious choice. However, these items will soon be gone as they are out of production (apparently, really this time).
I suggest instead using a 1uF film capacitor if you cannot use the 4,7 uF Black gate.
boomy,
the regulator takes care of reducing the noise. adding enormous coupling caps just pushes up the conduction factor ... not healthy. The CRC suggested by Nordic is ok except that the regulator again takes gare of the noise, and is actually quite noisy itself... more noise is put back ... so I suggest a minimal amount of capacitance here and an CRC filter on the regulator output for final filtereing of noise... if you insist on this.
The 4,7 uF can be purchased from anywhere you can get Black Gate caps, Partconnexion being the obvious choice. However, these items will soon be gone as they are out of production (apparently, really this time).
I suggest instead using a 1uF film capacitor if you cannot use the 4,7 uF Black gate.
boomy,
the regulator takes care of reducing the noise. adding enormous coupling caps just pushes up the conduction factor ... not healthy. The CRC suggested by Nordic is ok except that the regulator again takes gare of the noise, and is actually quite noisy itself... more noise is put back ... so I suggest a minimal amount of capacitance here and an CRC filter on the regulator output for final filtereing of noise... if you insist on this.
Hello the Phonoclone is playing fine now .
I tamed the hum devil.
Its not totaly gone but its low enough
it only hums on levels that i dont use
It was the tonearm the ground shield was not connected to the metal inside.
My cable from cart to the phonodude is stil very sensitive for movements it needs carefully placement for lesser hum.
when i tap on the cable i hear that also amplified thru my speakers.
I tryed the grounlug also connected on the inputground then is the cable not sensitive anymore but i have more hum then.
I tamed the hum devil.
Its not totaly gone but its low enough
it only hums on levels that i dont use
It was the tonearm the ground shield was not connected to the metal inside.
My cable from cart to the phonodude is stil very sensitive for movements it needs carefully placement for lesser hum.
when i tap on the cable i hear that also amplified thru my speakers.
I tryed the grounlug also connected on the inputground then is the cable not sensitive anymore but i have more hum then.
fuses
I'm not really qualified to make recommendations on safety, so I kept silent on the issue. In my view, the fuse should be weak enough that it will definitely blow before you melt the transformer.
Lets say you have a 50VA transformer, and 100V AV line voltage, this means the max current the transformer primary should support is 500mA, but there is probably a certain amount of safety margin. Thus in that example I would set the fuse to 500 mA, slow blow type. It is not super critical, so in fact I would say 500mA would be ok for all 100V-ish lines, 250mA for all 200V-ish lines.
I'm not really qualified to make recommendations on safety, so I kept silent on the issue. In my view, the fuse should be weak enough that it will definitely blow before you melt the transformer.
Lets say you have a 50VA transformer, and 100V AV line voltage, this means the max current the transformer primary should support is 500mA, but there is probably a certain amount of safety margin. Thus in that example I would set the fuse to 500 mA, slow blow type. It is not super critical, so in fact I would say 500mA would be ok for all 100V-ish lines, 250mA for all 200V-ish lines.
When I first powered up my dual mono PS using a 250mA slow blow, it blew the fuse (nothing connected at the time, not even the phonoclone). I thought there was something wrong with my wiring which I then went back and checked over and over again.
Ended up putting in a 1A fast blow which is what has been there ever since.
Probably is a bit rich for it though.
Fran
Ended up putting in a 1A fast blow which is what has been there ever since.
Probably is a bit rich for it though.
Fran