I run some more measurement sets to double check the function of SSM2141 differentials devices at output.
Yes, in one respect they are slightly worsening things: SNR is increased but by just 0.5 db, from - 81.3 to -80.8 dbV. Yesterday I had 80.8 without them...so really not relevant) . A weighted N+D went from 88.5 to 85.5 dbV A, so 3 db worst.
However, now balanced signal is completely equal, + output is same RMS as - output which was not the case without them. In addition now I can use single ended amp without issue, small noise loss is mainly in LF where it cannot be heard IMO.
As so far I changed my mind and I will keep them. Used last jumper on second stage (Rg =3k3) that gives me 1029 VRMS output (differential) from 5mV input.
Only listening can , if it can, determine whats better here.
This is then final diagram that goes inside TT now:
Yes, in one respect they are slightly worsening things: SNR is increased but by just 0.5 db, from - 81.3 to -80.8 dbV. Yesterday I had 80.8 without them...so really not relevant) . A weighted N+D went from 88.5 to 85.5 dbV A, so 3 db worst.
However, now balanced signal is completely equal, + output is same RMS as - output which was not the case without them. In addition now I can use single ended amp without issue, small noise loss is mainly in LF where it cannot be heard IMO.
As so far I changed my mind and I will keep them. Used last jumper on second stage (Rg =3k3) that gives me 1029 VRMS output (differential) from 5mV input.
Only listening can , if it can, determine whats better here.
This is then final diagram that goes inside TT now:
The preamp installed in TT , shielded by copper and connected to MM cartridge.
Here is noise floor of different situations:
Red trace is from bench measurements with test signal for reference.
Dark green , blue and orange is with preamp powered in its place and cart connected, arm placed in middle of LP. It is with no power to TT, powered TT and motor running on TT respectively. There is no obvious signal pick up from motor. Here is obvious that cart + cable in the arm pick up more HF noise than test signal cable, Probably more important is that cart is not 800 ohm resistor, its impedance goes up with frequency, hence more noise as frequency raises.....
Finally , light green trace on top is cartridge playing silent groove just to illustrate where we are in real world.
Here is noise floor of different situations:
Red trace is from bench measurements with test signal for reference.
Dark green , blue and orange is with preamp powered in its place and cart connected, arm placed in middle of LP. It is with no power to TT, powered TT and motor running on TT respectively. There is no obvious signal pick up from motor. Here is obvious that cart + cable in the arm pick up more HF noise than test signal cable, Probably more important is that cart is not 800 ohm resistor, its impedance goes up with frequency, hence more noise as frequency raises.....
Finally , light green trace on top is cartridge playing silent groove just to illustrate where we are in real world.
That's interesting stuff. You have mechanical peaks at (roughly) 130Hz and 340Hz. They must have a cause, and my first guess would be the motor. What frequency does it receive? Do those peaks change in frequency if you switch between 33 and 45? If not, they could be tonearm resonances. What your graph does show is that electronic noise and interference is negligible <1kHz, but gets worse as frequency increases. You have bipolar inputs, implying current noise that develops voltage noise when it passes through the MM cartridge's source impedance. JFET input avoids that problem. I also note that you have not opted for synthesised loading resistance. Given that you have a balanced topology, it would be easy to synthesise loading resistance by adding a resistor between each input and the opposite polarity's output. With JFET input and synthesised loading resistance, it should be possible to get the blue and orange traces closer to the red.
Hi , sorry might be I picked wrong silent grove with end of it where needle jumps back. I just wanted to show surface noise relative to amp noise. The mechanics of this table are well analyzed in Limestone thread, Here is picture of few silent passages, one has that bump,
Probably I will experiment with JFET input later, have OPA2828 in mind for long. Wondering if it will be possible to hear difference.... Can you sketch or link this synthesized loading?
And one video more for fun, motor is after all visible, but only at start up as this is set up that 3 phase motor starts with higher amplitude first few seconds, than amplitude is reduced to smaller one needed to keep platter spinning.
At start up you see ramp up of motor as signal 1, then it reaches 180Hz (3 phases running at 60.61 Hz) and then when running lower amplitude is applied, its signal sinks in to noise floor.
At start up you see ramp up of motor as signal 1, then it reaches 180Hz (3 phases running at 60.61 Hz) and then when running lower amplitude is applied, its signal sinks in to noise floor.
The only reason to apply synthesized loading is to increase S/N.
But when weighted S/N is above 75dBA, improving this figure by a few dB’s serves no purpose and only complicates the circuit.
At 20kHz the surface noise is still 10dB above the preamp’s noise, so this preamp passes the test with flying colours.
Hans
But when weighted S/N is above 75dBA, improving this figure by a few dB’s serves no purpose and only complicates the circuit.
At 20kHz the surface noise is still 10dB above the preamp’s noise, so this preamp passes the test with flying colours.
Hans