A Low Noise Balanced Input Moving Coil Preamp Using the ZTX851
- mediatechnology
- Posts: 5466
- Joined: Sat Aug 11, 2007 2:34 pm
- Location: Oak Cliff, Texas
- Contact:
Re: A Low Noise Balanced Input Moving Coil Preamp Using the ZTX851
It is set to 0.48 Fs.
They show response plots out to 80 kHz for Fs=192kHz.
I run mine at 96 kHz and have verified it goes out to 40 kHz.
The HPF is the one that moves higher as Fs is increased. The rolloff is about 0.3 dB at 20 Hz at Fs=192kHz.
They show response plots out to 80 kHz for Fs=192kHz.
I run mine at 96 kHz and have verified it goes out to 40 kHz.
The HPF is the one that moves higher as Fs is increased. The rolloff is about 0.3 dB at 20 Hz at Fs=192kHz.
Re: A Low Noise Balanced Input Moving Coil Preamp Using the ZTX851
D'oh. I just realised that I hadn't read the spec sheet properly! The input SD modulator has only 2 clock options. I'm an eejit.
- mediatechnology
- Posts: 5466
- Joined: Sat Aug 11, 2007 2:34 pm
- Location: Oak Cliff, Texas
- Contact:
Re: A Low Noise Balanced Input Moving Coil Preamp Using the ZTX851
I edited my post to correct the HPF rolloff.
BTW I'm looking at going with a more conventional front-end for the EVM board.
At a minimum I'm going to install coupling caps so I don't have to sink current from its inputs floating at Vcm.
The insertion loss for measurement purposes is the real Debbie-downer.
It loads a 49Ω output by about 0.7dB.
But its quiet. I get the -124 dBFS "A" weighted DR.
Unweighted is almost as good.
Unbalanced isn't as good.
BTW I'm looking at going with a more conventional front-end for the EVM board.
At a minimum I'm going to install coupling caps so I don't have to sink current from its inputs floating at Vcm.
The insertion loss for measurement purposes is the real Debbie-downer.
It loads a 49Ω output by about 0.7dB.
But its quiet. I get the -124 dBFS "A" weighted DR.
Unweighted is almost as good.
Unbalanced isn't as good.
Re: A Low Noise Balanced Input Moving Coil Preamp Using the ZTX851
Great to see you have a final proto with final measurements.
What about the noise from the servo ?
At the op amp output
At C6
The effectiveness of noise removal by the 100 nF cap.
What about the noise from the servo ?
At the op amp output
At C6
The effectiveness of noise removal by the 100 nF cap.
- mediatechnology
- Posts: 5466
- Joined: Sat Aug 11, 2007 2:34 pm
- Location: Oak Cliff, Texas
- Contact:
Re: A Low Noise Balanced Input Moving Coil Preamp Using the ZTX851
I don't know how final they are but I'm comfortable in starting a layout.
There's one more thing I want to try first.
C6 could probably be eliminated since its in-band effect is minimal but I think I would like a spot on the board for it to lower the differential impedance at that node for RF/EMI.
The entire bridge could be scaled lower by a factor of 5-10.
At high MC gains the front-end noise dominates.
At low gains, for mic preamp use, the servo might be a noise contributor but the 124X is likely to be as much or more.
I would do the cross-coupled with resistors in the 2KΩ range.
At that point servo noise could matter.
My focus right now is the MC preamp.
There's probably quite a few things I would do differently at mic impedance and gain ranges.
There's one more thing I want to try first.
C6 could probably be eliminated since its in-band effect is minimal but I think I would like a spot on the board for it to lower the differential impedance at that node for RF/EMI.
The entire bridge could be scaled lower by a factor of 5-10.
At high MC gains the front-end noise dominates.
At low gains, for mic preamp use, the servo might be a noise contributor but the 124X is likely to be as much or more.
I would do the cross-coupled with resistors in the 2KΩ range.
At that point servo noise could matter.
My focus right now is the MC preamp.
There's probably quite a few things I would do differently at mic impedance and gain ranges.
- mediatechnology
- Posts: 5466
- Joined: Sat Aug 11, 2007 2:34 pm
- Location: Oak Cliff, Texas
- Contact:
Re: A Low Noise Balanced Input Moving Coil Preamp Using the ZTX851
I made some noise measurements and the servo isn't a significant noise contributor.
With the preamp running at 62 dB gain I measured the servo's differential output noise.
It's below -104 dBu not too far above the measurement hum floor.
I also switched the preamp gain to +6 dB (Rgain open) and read the preamp's differential output noise both with and without the servo operational.
(I shorted C6 to turn it off.)
The output noise read about -102 dBu regardless of whether the servo was or on or off.
The voltage noise density at +6 dB gain is about 20 nV√Hz which is about the same as a THAT1512 at similar gain.
(THAT1512 is 34 nV√Hz at 0 dB gain.)
Using op-amp-based diff amps with lower Rfb would be appropriate in a mic preamp.
WRT lowering the values of the bridge resistors the TL431 shunt regulator isn't able to supply the current so a different scheme would be used.
As I mentioned previously, varying the reference voltage of the top bridge arms allows Ic to be "tuned" for the application without altering bridge gain.
In this application, a MC preamp, lowering bridge or dIff amp feedback values isn't really necessary since the front end noise, though low, dominates.
Measurement shows that.
With the preamp running at 62 dB gain I measured the servo's differential output noise.
It's below -104 dBu not too far above the measurement hum floor.
I also switched the preamp gain to +6 dB (Rgain open) and read the preamp's differential output noise both with and without the servo operational.
(I shorted C6 to turn it off.)
The output noise read about -102 dBu regardless of whether the servo was or on or off.
The voltage noise density at +6 dB gain is about 20 nV√Hz which is about the same as a THAT1512 at similar gain.
(THAT1512 is 34 nV√Hz at 0 dB gain.)
Using op-amp-based diff amps with lower Rfb would be appropriate in a mic preamp.
WRT lowering the values of the bridge resistors the TL431 shunt regulator isn't able to supply the current so a different scheme would be used.
As I mentioned previously, varying the reference voltage of the top bridge arms allows Ic to be "tuned" for the application without altering bridge gain.
In this application, a MC preamp, lowering bridge or dIff amp feedback values isn't really necessary since the front end noise, though low, dominates.
Measurement shows that.
- mediatechnology
- Posts: 5466
- Joined: Sat Aug 11, 2007 2:34 pm
- Location: Oak Cliff, Texas
- Contact:
Bias Current "Assistance"
Wanted to drop a bookmark back to this post: viewtopic.php?f=6&t=783&p=11351#p11351
An additional discussion regarding "bias current assistance" in the Audio Precision System One can be found here: https://proaudiodesignforum.com/forum/p ... 852#p12852
Updated schematic.
Flat Balanced Input Moving Coil Phono Preamp Using The ZTX851. Drawing updated March 15, 2019.
Change from previous schematic: Added R32 and R33.
+15 mV is developed across R33 to satisfy an estimate of the common mode bias current of Q1 and Q2.
Next up some THD measurements...
An additional discussion regarding "bias current assistance" in the Audio Precision System One can be found here: https://proaudiodesignforum.com/forum/p ... 852#p12852
Updated schematic.
Flat Balanced Input Moving Coil Phono Preamp Using The ZTX851. Drawing updated March 15, 2019.
Change from previous schematic: Added R32 and R33.
+15 mV is developed across R33 to satisfy an estimate of the common mode bias current of Q1 and Q2.
Next up some THD measurements...
Re: Bias Current "Assistance"
What good is removal of common mode biais current of Q1 and Q2 ?mediatechnology wrote: ↑Sat Mar 16, 2019 8:33 am Change from previous schematic: Added R32 and R33.
+15 mV is developed across R33 to satisfy an estimate of the common mode bias current of Q1 and Q2.
This + 15 mV lifts the inputs close to ground, removing DC at the source input, is this the only reason, or am I missing something ?
- mediatechnology
- Posts: 5466
- Joined: Sat Aug 11, 2007 2:34 pm
- Location: Oak Cliff, Texas
- Contact:
Re: A Low Noise Balanced Input Moving Coil Preamp Using the ZTX851
Just to move the inputs closer to ground, that's all.