The core to signal, the shield to ground (negative).
The unshielded wires you have right now, are basically dangling antennas.
Picking up interference from their surroundings.
The tiny hum I had with my build, was due to the proximity of my twisted wires to the AC lines and trafo.
I was ready to go for RG58, but moving them away an inch or so, was enough to solve the problem.
I thought shields should be connected only at one end ?
Looking back at your build, it seems we have a similar distance between signal wires and the IEC socket + transformer. And I have a supposedly quieter and shielded torroidal.
The RG wires you mentioned seem hard to get for me, would something like this work ? TinyURL.com - shorten that long URL into a tiny URL
I'm also gonna try the preamp on my main (earthed) system later this week to see if the test setup is not my main issue.
Thanks,
You can use these as well:
https://www.amazon.ca/Beyondoor-RG178-Coaxial-Length-Impedance/dp/B07SK1QVC2/ref=sr_1_3?keywords=rg178+cable&qid=1578479454&sr=8-3
I am no expert (still learning) but when you're using coaxial/shielded cables for the in/out lines, you're connecting the inner conductor to the signal(+) and the outer shield to ground(-) at both ends.
What happens as a result is that the outer braided wire, apart from establishing a return to ground(-) for the signal(+) it acts as a shield for the signal as well.
In some situations you can get away with using plain unshielded wires but you risk more interference. The twisted wires I use are single copper conductor ones from a cat. 7 network cable. They do resist interference a bit due to being a twisted pair, but they will never be like coaxial.
Another way to go, would be to use a twisted pair as usual PLUS dress them up in a braided copper earth wire.
That one you can simply connect to one side(earth ground) as you mentioned.
You can never exactly predict how interference will affect your build.
That is why many guys just go by the book and employ shielded cables and a star ground layout to be sure.
After all, I suppose the sound difference between a proper coaxial and a single core twisted pair is probably zero or very tiny, so many go for the shielded way to be sure.
Last edited:
The problem of induced hum is almost certainly within the case, so any shielding you do there (eg from the RCAs to the board) will probably help some. But not likely totally. My limited understanding about shielding is that if you connect both ends of it (board and RCA jack) you may just continue or worsen the ground loop problems you are having. Connecting on one end (the input usually) is supposed to (in my reading) act as an antenna for stray fields. Grounding only one end sends the noise to the ground as far from the board as possible with no possibility of a ground loop. Now that's the idea as I understand it. I don't know quite what I think of that yet...
Last edited:
So I tested tonight on my main system (ClearAudio Cart,Rega Planar 3, Stereo VSPS, and a NAD C316 from 2019) and sure enough the hum is there from 12h on the volume control. If Im playing music at 12oclock I will probably not stand close enough to the speakers to ear the hum anyway.
Nevertheless... I noticed the NAD doesnt have an earth ground either, and neither does most of the stuff in my media cabinet. I decided to run the phono with a 2 prong cable and to my surprise it changes nothing in the hum on that system (while it did on my test one).
So yeah, I have a couple options, I think I'm gonna enjoy it like it is for a bit. Then maybe move to a 2-case approach and/or use shielded signal cables.
Oh and maybe install a switch cause right now it will be on 24/7, will that shorten it's life?
Thanks a lot to everyone involved!
Edit: I never tested pulling the power supply parts out of the case, thats still on the todo list.
Nevertheless... I noticed the NAD doesnt have an earth ground either, and neither does most of the stuff in my media cabinet. I decided to run the phono with a 2 prong cable and to my surprise it changes nothing in the hum on that system (while it did on my test one).
So yeah, I have a couple options, I think I'm gonna enjoy it like it is for a bit. Then maybe move to a 2-case approach and/or use shielded signal cables.
Oh and maybe install a switch cause right now it will be on 24/7, will that shorten it's life?
Thanks a lot to everyone involved!
Edit: I never tested pulling the power supply parts out of the case, thats still on the todo list.
Last edited:
Hi cok666n,
Install an MOV rated at 150 VAC directly across the primary of the transformer. That will clamp the voltage the switch contacts see when you are switching off. You should find the life of your power switch greatly extended.
Please don't run audio equipment 24/7. Hydro plays at night for one. For two, the percentage of time you use your system compared to the total run time would be very small. So you would have failures sooner than if you turned it on and off. Besides, audio systems are not designed to run all the time (no matter what Bryston says *). IF something goes wrong, you would be further ahead if you were able to turn stuff off before the damage gets too bad..
-Chris
*Just an example of a company that claimed their product was designed to run 24/7. It wasn't.
Install an MOV rated at 150 VAC directly across the primary of the transformer. That will clamp the voltage the switch contacts see when you are switching off. You should find the life of your power switch greatly extended.
Please don't run audio equipment 24/7. Hydro plays at night for one. For two, the percentage of time you use your system compared to the total run time would be very small. So you would have failures sooner than if you turned it on and off. Besides, audio systems are not designed to run all the time (no matter what Bryston says *). IF something goes wrong, you would be further ahead if you were able to turn stuff off before the damage gets too bad..
-Chris
*Just an example of a company that claimed their product was designed to run 24/7. It wasn't.
@cok666n
The circuit draws minimal power. About half a watt. 4 kWh or <1$ a year. You can turn it off, or not, nothing in the VSPS "wears out" with use.
The more you describe the hum problem, the more I realise it is very low level, easily explained by capacitive or inductive coupling between the power supply and the VSPS input wires.
Capacitive coupling can be easily guarded against with a grounded metal shield of some kind between the emitting potential and the receiving conductor. A metal divider in the case between the AC input and the RCA jacks, for example. As long as it is electrically in contact with the case and fully encloses the noisy components it will do the trick. Inductive coupling is harder to deal with, since the generated magnatic fields will pass through thin metal. The range is short however. The easiest solution, to be honest, is a larger case or better yet, separate ones for the amplifier and power supply.
If you like you can record a .wav file of the VSPS line output and post it here. I can have a look at it in Audacity or RMAA and can give you some pointers about what the characteristics of the noise is.
The circuit draws minimal power. About half a watt. 4 kWh or <1$ a year. You can turn it off, or not, nothing in the VSPS "wears out" with use.
The more you describe the hum problem, the more I realise it is very low level, easily explained by capacitive or inductive coupling between the power supply and the VSPS input wires.
Capacitive coupling can be easily guarded against with a grounded metal shield of some kind between the emitting potential and the receiving conductor. A metal divider in the case between the AC input and the RCA jacks, for example. As long as it is electrically in contact with the case and fully encloses the noisy components it will do the trick. Inductive coupling is harder to deal with, since the generated magnatic fields will pass through thin metal. The range is short however. The easiest solution, to be honest, is a larger case or better yet, separate ones for the amplifier and power supply.
If you like you can record a .wav file of the VSPS line output and post it here. I can have a look at it in Audacity or RMAA and can give you some pointers about what the characteristics of the noise is.
Last edited:
Hi rjm,
I'm just going to make one point, then leave it.
You will also find that the failure rate doubles for every 10 °C temperature rise, and that is for every component. The MTBF rating for most capacitors is on the order of 2,000 hours to 5,000 hours. That's time to failure, the performance degrades steadily over time, leading to impaired performance at some point in time.
So that is my point, supported by the manufacturers of the parts you use. Therefore, turning the unit off when not in use matters to your customers. Please keep this in mind when making recommendations.
Best, Chris
I'm just going to make one point, then leave it.
That is not correct. All components have a mean time before failure specification. The electrolytic capacitors you are using are rated for "x" amount of hours for example. Look at the datasheets for the components you use.
You will also find that the failure rate doubles for every 10 °C temperature rise, and that is for every component. The MTBF rating for most capacitors is on the order of 2,000 hours to 5,000 hours. That's time to failure, the performance degrades steadily over time, leading to impaired performance at some point in time.
So that is my point, supported by the manufacturers of the parts you use. Therefore, turning the unit off when not in use matters to your customers. Please keep this in mind when making recommendations.
Best, Chris
The Nichikon UKW we use are rated at 2000 Hrs @ 85°C, I have no idea what that would mean at room temp but that does seem quite low (something like 6 months 24/7 ?)
Well, I'm going to install a switch...
Hi cok666n,
Well, every 10°C drop in temperature halves the failure rate. However, most internal temperatures are a bit higher than room temperature. Amplifiers that had high failure rates typically ran at 50°C.
The other issue is that some attention is being given to all the non-obvious power usage from low power devices in the home, and how it can add up to higher costs than you might expect.
Good decision for the switch.
-Chris
Well, every 10°C drop in temperature halves the failure rate. However, most internal temperatures are a bit higher than room temperature. Amplifiers that had high failure rates typically ran at 50°C.
The other issue is that some attention is being given to all the non-obvious power usage from low power devices in the home, and how it can add up to higher costs than you might expect.
Good decision for the switch.
-Chris
Hi rjm,
Tell that to my customers who have experienced failures.
Look, I'm not saying the equipment will surely fail in the near future. I don't even know what temperature it gets to inside your chassis (but you should with an ambient air temperature of 25 °C). But, by the same token, it isn't wise to leave your equipment on all the time. This is for many reasons. If you can turn something off when you aren't using it, then you should.
-Chris
Tell that to my customers who have experienced failures.
Look, I'm not saying the equipment will surely fail in the near future. I don't even know what temperature it gets to inside your chassis (but you should with an ambient air temperature of 25 °C). But, by the same token, it isn't wise to leave your equipment on all the time. This is for many reasons. If you can turn something off when you aren't using it, then you should.
-Chris
Hello,
I'm thinking of using a couple of single OPA627 on adapter, in place of the dual OPA2134.
Is there anyone that has tried such a thing already?
Would it be a compatible opamp with this particular circuit?
I have read that the OPA627 can be rather capricious regarding implementation.
Thanks in advance,
Nikos, Athens.
I'm thinking of using a couple of single OPA627 on adapter, in place of the dual OPA2134.
Is there anyone that has tried such a thing already?
Would it be a compatible opamp with this particular circuit?
I have read that the OPA627 can be rather capricious regarding implementation.
Thanks in advance,
Nikos, Athens.
This is correct. By all accounts (I have not tried myself) it is difficult to use, with little reward.
