Thank you Joachim, you answered my question, BEFORE I could even put it up.
Therefore, for simplicity, I recommend adding a PPS coupling cap by removing R13, and substituting an appropriate valued cap in that location, PPS or better, only, NO CERAMICS. Following the cap, we should put a 1 Meg bleeder resistor to ground.
What should the value of that cap be?
Therefore, for simplicity, I recommend adding a PPS coupling cap by removing R13, and substituting an appropriate valued cap in that location, PPS or better, only, NO CERAMICS. Following the cap, we should put a 1 Meg bleeder resistor to ground.
What should the value of that cap be?
No, CG, a Dennon is an exotic cartridge, but it would work in a pinch, with this simple design.
Choices to handle DC offset in high gain phono stage:
1. "Servo" DC auto-null.
2. DC blocking cap in feedback.
3. Same as one, but instead of an opamp based auto-null circuit, use a pot that is set by the user.
4. DC blocking cap at output of phono section, and live with the asymmetrical clipping points.
1. "Servo" DC auto-null.
2. DC blocking cap in feedback.
3. Same as one, but instead of an opamp based auto-null circuit, use a pot that is set by the user.
4. DC blocking cap at output of phono section, and live with the asymmetrical clipping points.
Good thinking CG, I did consider offering a servo, but it is really 'overkill' for this design, why?
Comming back to the coupling cap it will see the 1MOhm bleeder plus the 50KOhm Potmeter and the 1MOhm input resistor. I whould opt for a 1uF, then we have a little rolloff under 20Hz. Kind of a shallow subsonic filter that could help a budget table. I whould not use anything that has to be adjusted during manufacture out of various reasons.
A 1uF in good quality ( like a WIMA MKS2 or RIFA or Pannasonic PPS ) is not that expensive i think.
A 1uF in good quality ( like a WIMA MKS2 or RIFA or Pannasonic PPS ) is not that expensive i think.
One uF might be OK. Certainly not more. Also, we would expect about 100mV offset, even less, and 300mV worst case. Big deal! ;-)
CG, if you heard this difference, I recommend a different volume control. Perhaps this is something we can talk about. Some pots are VERY SENSITIVE to loading or current though their wiper, others are not. Why?
Line stage gain
In my opinion, the gain in the line stage is too high for real world (digital) sources. A switch in series with R6 sets gain to zero dB. Depending on the actual PCB layout, a small cap across R4 may be needed.
I guess more EMC minded engineers would like to see some RF protection on the input and the output.
Also, a PSU with decoupling and some way of muting the output while power-up or -down are needed. Or am I adding a wheel now 🙂
In my opinion, the gain in the line stage is too high for real world (digital) sources. A switch in series with R6 sets gain to zero dB. Depending on the actual PCB layout, a small cap across R4 may be needed.
I guess more EMC minded engineers would like to see some RF protection on the input and the output.
Also, a PSU with decoupling and some way of muting the output while power-up or -down are needed. Or am I adding a wheel now 🙂
No, the input pot will provide lots of protection. Perhaps a 100pf polypropylene input cap to ground could be useful, on all positive inputs.
Of course for really minimum parts you could change the output stage to inverting, put the pot on the output and use a low pass filter on the second opamp output for dc feedback to the input opamp. You can then get rid of the output resistors from both opamps.
The second input could be summed in and the selector switch would AC short to ground the unused input.
But being too cheap is probably not a good idea.
In the current design if you use a linear pot for the VC then it could be loaded at 10% of it's value to simulate a log control. The advantage is that it is easier to get good channel tracking in a linear pot than an "audio" taper unit.
The second input could be summed in and the selector switch would AC short to ground the unused input.
But being too cheap is probably not a good idea.
In the current design if you use a linear pot for the VC then it could be loaded at 10% of it's value to simulate a log control. The advantage is that it is easier to get good channel tracking in a linear pot than an "audio" taper unit.
No, keep the output resistors. They serve a purpose: stability into any load. And they do no harm.
I don't like that idea. First, the Zin will vary from 50k to less than 5k depending on potmeter position. Second, heavily loading the wiper may introduce distortion. Wipers do not like current 🙂
I've found that loading affects the performance of both Alps "Blue Velvet" as well as P&G pots.
Apparently I'm not alone:
Application Note 25 - Plastic potentiometers
No such problem with a switched TKD...
Back to the preamp.
One possible consideration might be to add some means to minimize "thumps" when the power is applied or removed.
Another consideration is the source switch.
Yet another is the entire power supply system.
2Quad: Less gain for the line stage in this case will actually sound worse than more gain. It will likely measure worse, too. Please read page 9 of the opamp's data sheet.
Apparently I'm not alone:
Application Note 25 - Plastic potentiometers
No such problem with a switched TKD...
Back to the preamp.
One possible consideration might be to add some means to minimize "thumps" when the power is applied or removed.
Another consideration is the source switch.
Yet another is the entire power supply system.
2Quad: Less gain for the line stage in this case will actually sound worse than more gain. It will likely measure worse, too. Please read page 9 of the opamp's data sheet.
The phono stage output ONLY sees the pot. The pot is a BIG resistor, therefore it is OK to remove R13. Please think it through. Second, why not leave it on, all the time? It only dissipates less than 1W total. Don't tell anyone.
We already have a 20 to 1 load to pot value ratio, that should be enough for this project. 10Meg pots are OK, so are 150 mph rated tires on a VW, but is it worth the extra cost?
- Status
- Not open for further replies.