Preamp upgrade

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

matthieu68

Active member
Joined
Jul 11, 2017
Messages
25
hello, I am embarking on the redesign of the cards of an old mixing console and I would have liked some suggestions concerning the Pre-amp. 

To improve it I was already thinking of changing this opamp by an op134 for example (I have some on hand) but the values ​​adjusting the gain could be revised to reduce the noise.

Do you have any suggestions on which values ​​to use?  The potentiometer is connected via a cable so it could very well end up elsewhere depending on the topology used.

  I think starting by lowering the value of the potentiometer and adding capacitors to filter the opamp is a good start but then?  suggestions ?  thank you
 

Attachments

  • IMG_20210304_113925.jpg
    IMG_20210304_113925.jpg
    379.1 KB · Views: 138
matthieu68 said:
To improve it I was already thinking of changing this opamp by an op134
Do you have any suggestions on which values ​​to use?  The potentiometer is connected via a cable so it could very well end up elsewhere depending on the topology used.
What is the original opamp? In terms ofnoise, you can hardly beat a 5534. If you choose it, you will need to add a 30pF cap between pins 1 & 8.
You wouldn't need to alter the value of the gain pot, since the 1k resistor is perfectly adequate.
Indeed OPA134 is not bad either, would probably be an improvement, but not as much as 5534.

I think starting by lowering the value of the potentiometer and adding capacitors to filter the opamp is a good start but then?
What do you mean "filtering the opamp"? If you mean reducing the BW, I don't think it has much value.
If you mean decoupling the power rails, I would strongly recommand it. Many older opamp-based circuit lack in this department. Adding a pair of ceramic 100nF caps never does any harm.

Now, after that, it may put in evidence lack of performance of the rest of heh circuitry...
 
it is a 4558, it is double, it is shared between the Pre amp and the Baxandall team.  Yes for the filter I am talking about filtering the power supply, it lacks the famous 100n capacitor.

Thank you
 
matthieu68 said:
it is a 4558, it is double, it is shared between the Pre amp and the Baxandall team.
OK, you can use 5532, NJM2068 or OPA2134. Some woud recommand LM4560.
 
Yes for the filter I am talking about filtering the power supply, it lacks the famous 100n capacitor.
I don't understand how the designer could let his product go out without them...
 
thank you for your advice.  Should resistance values ​​be changed?  For the 100n filters I could read here and there that the old opamp was much slower and did not necessarily require this filter?

  thank you. 

besides, as long as we are talking about these 100n filters, what would be the best type, ngo or film?
 
matthieu68 said:
thank you for your advice.  Should resistance values ​​be changed?  For the 100n filters I could read here and there that the old opamp was much slower and did not necessarily require this filter?

  thank you. 

besides, as long as we are talking about these 100n filters, what would be the best type, ngo or film?
Although they filter the supply, those are called "bypass" capacitors. They can be cheap ceramic. But you only really need them if the op amp demands them. In particular, fast op amps like OPA134 would need them whereas NE5534 would not. NE5534 would need a compensation cap but it's probably easy to tack that on from the bottom. NE5534 have a bit of an offset that will show up on the insert output but it's probably not something to worry about in this circuit. I would not get too crazy with this pre. It's simply not the best / lowest noise design so there is only so much you can do. But if you swap in an NE5534 in place of a 4558, you should get a decent improvement in noise and it should work well enough for practical things. If you insist on doing more, you could reduce the impedance of the op amps resistors VR1, R20, R10, R11 and R7 (not R19). Just divide all by 10. So use a 10KB pot, R11 becomes 100R and so on. Then make the 47uF cap 470uF (although as low as 100u would be ok too). That might give you a bit of an improvement in noise too. Hard to say.
 
hummm thank you, I will test on these points that you mentioned, thank you for all that.  I know the console is not at the top level.


My second solution to improve this console was to remove the Pre amp and use the two inputs as line input (line 1, line 2) I already have differential opamp cards for input the line inputs.  The console would be like a big summator with eq.  My idea on this point was to lower the level to -2dBu (for the internal nominal level), with an input potentiometer (after the Switch line 1/2) which would allow me to adjust the level by + 4dBu (level  arrival) to infinity through the nominal level -2dbu. 

Maybe use the opamp which was for the Pre-amp as voltage follower to keep the impedance correct.  What do you think, the second version seems viable?  or null and useless?
 
matthieu68 said:
What do you think, the second version seems viable?  or null and useless?
Again, I probably wouldn't get too crazy with this. Just do the op amp upgrade, replace critical electrolytics and play around with it for a while. If it's really old it could be that some switches are noisy and pots are scratchy. It's something to learn from and have fun with but it's never going to perform at a level that justifies gutting and replacing lots of stuff. You can buy a mixer on Ebay for $200 USD that will outperform (numbers-wise) anything made before 1980.
 
matthieu68 said:
thank you for your advice.  Should resistance values ​​be changed?  For the 100n filters I could read here and there that the old opamp was much slower and did not necessarily require this filter?
That is a common misbelief. Decoupling capacitors not only serve to maintain stability by neutralizing the effects of the traces inductance, they also keep transient currents in a short loop, as long as the caps are returned to a ground that's close to the stage. For output stages, that draw more current from the rails, these 100nF caps are generally not up to the task, and need to be helped with higher value electrolytics.
Note that up to 100uF in similar size are available.

besides, as long as we are talking about these 100n filters, what would be the best type, ngo or film?
Film caps have too high stray inductance. NG0/COG is not necessary, unless you worry about caps being microphonic. Basic X7R, used lavishly, are plenty good.
 
abbey road d enfer said:
R11 is the resistor that defines the impedance on the inverting input (without pad); that's what counts for noise.

I see. Thanks for clarifying the resistor to which you were referring.
Around that area I'm a bit confused by the Line/Mic Switch detail - It seems to show both Line and Mic (transformer) signals connected together and into the opamp circuit as shown and both going to 0V when switched the other way. I guess it's just a 'drawing thing'.
On the circuit itself - we don't know the transformer detail (+20dB ? etc) but yes - you wouldn't go far wrong  with a 5532 or similar even if it is a bit power hungry - if it's just a single preamp.
Personally I'd be tempted to re-arrange or dedicate to either Mic / Line and lose the attenuating voltage divider on the Line input.

Decoupling capacitors. I'm wondering if they might actually be in the circuit but shown elsewhere on the schematic. It's quite usual to group the voltage rails and decoupling caps on a schematic to aid clarity.
But yes - various ways to describe their function - stability against oscillation; a local low impedance supply and minimising current loop area. 'Noise Filtering' not a primary function esp since on their own they have little impedance to 'work against'. Although often seen with 10 or 22 Ohm resistors to form RC so giving some hf filtering but the point of those resistors is arguably just as much to act as fuses in case an op amp goes wrong and would otherwise take out a supply rail that feeds a number of other channels in a mixer. Hence use of non-flammable NFR resistors.
 
Newmarket said:
Around that area I'm a bit confused by the Line/Mic Switch detail - It seems to show both Line and Mic (transformer) signals connected together and into the opamp circuit as shown and both going to 0V when switched the other way. I guess it's just a 'drawing thing'.
Yes. It's typical of schemos made by draftsmen who are not technicians.
 
Actually it is a mistake in the schem. Only one of the poles for line or mic should be grounded. If you swap the branches of pole switching the transformer then I think it would be correct.
 
yes it is a schematic error, nothing very dramatic.  For the supply filter capacitors they are on the diagram (further) but there is only 2x general electrolytic 47uf for the whole board.  I will take your advice and try the improvements mentioned, but not to get too carried away.  Thank you
 

Latest posts

Back
Top