Hi Richard, just spitballing here, if one was paralleling resistors in the riaa but keeping the standard cap values, what resistor values would give the best riaa match?
Seems like a waste of mental energy to even consider such a thing. With the list component values, the deviation is already less than 0.1 dB.
In practice the accuracy limit is finding capacitors that are within 33.0+/-0.3 nF and 3.30+/-0.03 nF. Provided one worries about such things, which I do not recommend. A few judicious excisions from the RIAA here and there is not a big big deal provided they are consistent across L and R channels.
In practice the accuracy limit is finding capacitors that are within 33.0+/-0.3 nF and 3.30+/-0.03 nF. Provided one worries about such things, which I do not recommend. A few judicious excisions from the RIAA here and there is not a big big deal provided they are consistent across L and R channels.
Attachments
I have a huge supply of 1% styrene caps and a 5 digit lcr meter so matching the caps is no issue.
The list values, assuming you can match all of them to <1%, will yield RIAA+/- 0.05dB.
There are no set of two parallel resistors that would give you a better result.
There are no set of two parallel resistors that would give you a better result.
I would be interested to know more and your schematic.
I have an RME soundcard that performs much better with balanced inputs.
Thanks Richard, I'm building AMRs boards with the smt riaa and loading placements and he's put extra pads on the riaa for trimming in series.
I'll stick with the standard values
I'll stick with the standard values
Please note the phonoclone.com website will be a little messed up for a while. I'm upgrading to SSL but right now the images aren't displaying.
I can confirm that your suggestion to run the Sapphire as a line stage to boost the output of the Emerald works amazingly well.
This gives enough output to the input of the M-Audio Audiophile-192 to achieve a useful recording level.
Meanwhile, I acquired, for very little outlay, a used RME HDSPe AIO card which has both Consumer and Pro input settings.
For this card I do not need to use the Sapphire as a line stage... the output of the Emerald is sufficient.
My Mundorf Supreme caps arrived this week so tomorrow I can install and start burn-in.... so looking forward to LP ripping with this set-up.!
This gives enough output to the input of the M-Audio Audiophile-192 to achieve a useful recording level.
Meanwhile, I acquired, for very little outlay, a used RME HDSPe AIO card which has both Consumer and Pro input settings.
For this card I do not need to use the Sapphire as a line stage... the output of the Emerald is sufficient.
My Mundorf Supreme caps arrived this week so tomorrow I can install and start burn-in.... so looking forward to LP ripping with this set-up.!
In general the Sapphire makes a very nice line stage preamp. You wouldn't normally dial in as much gain as you did [by my calculation you need an extra 12 dB to convert consumer line (-10 dB) to pro input (+4 dBu)] but the Sapphire works well there too since it's designed as a headphone amp. Just note that it doesn't have colossal amounts of voltage headroom, which by my calculation is 13 dB for +4 dBu. [The maximum output signal before clipping is about 5 V rms, less than half that of a standard audio op amp operating from +18V/-18V rails.]
Richard
Is a -52 dB 50Hz hum to be fixed or accepted ?
Not sure if I am trying to track down something that matters not or matters...🙂
My Emerald is connected to an RME sound-card for recording.
When I test the sound-card with nothing connected the noise level is lower than I can measure.
When I test it with the Emerald connected and the Emerald inputs shorted the noise level is lower than I can measure.
When I test it with the Emerald connected and the Emerald inputs open there is a 50Hz spike @ -52dB and harmonics at 150 and 250 Hz.
When I test it with the Emerald connected and the Emerald inputs connected to my MM phono cartridge there is a 50Hz spike @ -52dB and harmonics at 150 and 250 Hz.
Screenshot attached.
I have tried all I can think of to reduce the 50Hz but perhaps I am chasing insignificance ? 😱
I see from the Audio-wise Trio (Emerald build) that the s/n is quoted...90dB for MM
Not sure if I am trying to track down something that matters not or matters...🙂
My Emerald is connected to an RME sound-card for recording.
When I test the sound-card with nothing connected the noise level is lower than I can measure.
When I test it with the Emerald connected and the Emerald inputs shorted the noise level is lower than I can measure.
When I test it with the Emerald connected and the Emerald inputs open there is a 50Hz spike @ -52dB and harmonics at 150 and 250 Hz.
When I test it with the Emerald connected and the Emerald inputs connected to my MM phono cartridge there is a 50Hz spike @ -52dB and harmonics at 150 and 250 Hz.
Screenshot attached.
I have tried all I can think of to reduce the 50Hz but perhaps I am chasing insignificance ? 😱
I see from the Audio-wise Trio (Emerald build) that the s/n is quoted...90dB for MM
Attachments
I'd normally measure the noise of a circuit with the inputs shorted, or with a dummy load. This can be considered the true output noise generated by the circuit.
With the inputs open circuit it's not too surprising to get some hum, as the input circuit has an impedance of 47 kohms so easily picks up electrostatic interference from the power supply.
With a cartridge connected the impedance drops to 500 ohms or so, but on the other hand you have an electromagnetic transducer (the cartridge) and several feet of wire attached to the input, providing additional points for interference noise pickup.
To confirm that the VSPS itself is blameless, you can connect a 500 ohm dummy load to the inputs instead of just shorting them. This gives the input section the same opportunity for noise pickup as in actual use with a cartridge.
If the hum is gone, the problem exists in front of the VSPS inputs: the cartridge/phono cable is picking up noise. If the hum remains, there is strong interference getting into the input section, so it will be necessary to distance the circuit from the source of the interference, or shield it.
-52 dB measured at the output is not terrible, not great. Definitely more than I'd be happy with, but low all the same.
I'm going to guess it's an interference issue. Ground loops are usually omnipresent regardless of whether the inputs are shorted or not.
With the inputs open circuit it's not too surprising to get some hum, as the input circuit has an impedance of 47 kohms so easily picks up electrostatic interference from the power supply.
With a cartridge connected the impedance drops to 500 ohms or so, but on the other hand you have an electromagnetic transducer (the cartridge) and several feet of wire attached to the input, providing additional points for interference noise pickup.
To confirm that the VSPS itself is blameless, you can connect a 500 ohm dummy load to the inputs instead of just shorting them. This gives the input section the same opportunity for noise pickup as in actual use with a cartridge.
If the hum is gone, the problem exists in front of the VSPS inputs: the cartridge/phono cable is picking up noise. If the hum remains, there is strong interference getting into the input section, so it will be necessary to distance the circuit from the source of the interference, or shield it.
-52 dB measured at the output is not terrible, not great. Definitely more than I'd be happy with, but low all the same.
I'm going to guess it's an interference issue. Ground loops are usually omnipresent regardless of whether the inputs are shorted or not.
With a dummy load of 470 ohm the 50Hz peak remains but is reduced in amplitude to -63dB....
So..... it seems that both options need pursuing... ?
So..... it seems that both options need pursuing... ?
I have made considerable progress.
1)Shortened and braided the leads from the remote power supply...no change.
2) up-rated the on-board supply caps from 1000uF to 4700uF... slight improvement.
3)replaced the twisted pairs of input and output signal cables with twin and shield, fixing the shield to the ground on the board...much worse!
but that was a big hint ! and after checking the grounding of the boards I found an error in my assembly and corrected it.
Now, with 470 ohm shorting plugs on the inputs the 50Hz peak is at -80.7 dB which would appear to be acceptable.
With the cartridge connected instead the peak rises to -64.3 dB so I have some more work to be done on cables between the cartridge and Emerald... (though Grados are known to be 'hummy' so I may try another cartridge just to check)
Thank you Richard, for pointing me in the right direction, and the useful comments as to expected behaviours and sources to investigate.
1)Shortened and braided the leads from the remote power supply...no change.
2) up-rated the on-board supply caps from 1000uF to 4700uF... slight improvement.
3)replaced the twisted pairs of input and output signal cables with twin and shield, fixing the shield to the ground on the board...much worse!
but that was a big hint ! and after checking the grounding of the boards I found an error in my assembly and corrected it.
Now, with 470 ohm shorting plugs on the inputs the 50Hz peak is at -80.7 dB which would appear to be acceptable.
With the cartridge connected instead the peak rises to -64.3 dB so I have some more work to be done on cables between the cartridge and Emerald... (though Grados are known to be 'hummy' so I may try another cartridge just to check)
Thank you Richard, for pointing me in the right direction, and the useful comments as to expected behaviours and sources to investigate.
That's great! -80 dB sounds about right, and yes, Grados are "hummy" in my experience, so nothing abnormal there.
What was the error that you had to fix?
What was the error that you had to fix?
😱 I had moved the boards to a new case recently and must have been in a hurry as I forgot to connect the Ground point (between the In/Outs on the board) back to the earth on the remote power supply case.
As it was under the board and the wires were hidden by the boards I didn't think to check they were connected !!
As it was under the board and the wires were hidden by the boards I didn't think to check they were connected !!
That's interesting. Normally the GND connects to the chassis of the VSPS, it doesn't route back to the power supply / earth, unless it's a one box build. I say that because in my experience making that connection increases hum, but here it's the opposite. Good to know both possibilities exist.
I have re-routed cables and placement of equipment.
Now I have around -70dB @50Hz peak at the start of an LP and -54dB at the end with the Grado cartridge.
This would appear to be the best that the Grado can achieve and in the coming week I will sub my Nagaoka which should be more immune to fields and more consistent across an LP.
Whether it sounds better is yet to be found..🙂
Now I have around -70dB @50Hz peak at the start of an LP and -54dB at the end with the Grado cartridge.
This would appear to be the best that the Grado can achieve and in the coming week I will sub my Nagaoka which should be more immune to fields and more consistent across an LP.
Whether it sounds better is yet to be found..🙂
Those numbers seem about right. Grados are simply more prone to hum than other brands, at least that was my experience way back before I started using the Denon.