Boost with NE5532

[Buffalo Tom]

Hello. Im building a stand alone mixer for my CE-1 Chorus pedal using the Suhr minimix schematic but modified with a boost on the return section. So here is the schematic Im using for the boost and it works fine as long as Im using the LF442 op-amp. But when I swap to NE5532 which is the op-amp I wanna use there is an oscillation type of sound when rotating the potentiometer. So any ideas why this circuit works perfect with LF442 but not so good with NE5532? Any parts I can add to make I stable? The reason why I wanna use NE5532 is that there is one left over in the Suhr minimix circuit and they will be on the same pcb so it would be nice to use that op-amp for the boost part.

[patrick398]

Looks like this is set up for bipolar supply, are you using a virtual ref at the non inverting input instead?

[patrick398]

Also if it's high pitch oscillation you could try a small cap across pins 1 and 2

[Buffalo Tom]

Im using a dc/dc converter to get bipolar supply +/-15V

[Buffalo Tom]

Also if it's high pitch oscillation you could try a small cap across pins 1 and 2

Thanks. I did tried before with 22pf cap in parallel with R3 but I realize now that that's not exactly between pin1 and 2. Cant try right now but will test tomorrow.

[patrick398]

22pf seems very small, I think you can afford to go larger without affecting audio frequencies

[antonis]

Start with 220pF going upwards ..

Also, reduce R2 to 100k..
(I know it dominates imput impedance 10 times but you deal with bipolar op-amp input bias..)

[Buffalo Tom]

Start with 220pF going upwards ..

Also, reduce R2 to 100k..
(I know it dominates imput impedance 10 times but you deal with bipolar op-amp input bias..)

Thanks antonis. Reduce the input Z to 100k will probably be fine.. Its the return section in a parallel mixer so its the output of the CE-1 pedal that will see this input impedance.. Im not going in direct with the guitar here. The return jack on my Skrydstrup BF2M has a impedance of 50k.. So Im better than that 

[amptramp]

What do you have for power supply bypass capacitors?  An NE5532 is a high-frequency device.  It is easy for it to oscillate if you don't use bypass capacitors to ground.  A value of 10 nF to 100 nF should be OK.

A feedback lead capacitor should not go across R3, it should go from the output pin to the inverting input so R5 does not get in the way.  R5 would turn it into a feedback lag which would promote oscillation.

[Buffalo Tom]

I have 10uF electrolytics and 100nF ceramics as power supply bypass capacitors on + and -. I have changed the schematic quite a bit and getting closer to a really good sounding clean boost with lots of headroom. The only problem I got with this one is that the pot gives a crackling background noice that is present when turning to pot. I believe this is DC? How can I remove this noice? 

[antonis]

Lower R3 value down to 100k (or less)..
(pot crackles due to DC difference between its lugs (output & inverting input))

P.S.
>I have 10uF electrolytics and 100nF ceramics as power supply bypass capacitors on + and -<

Sometimes is prefered to have a single pair of electro & ceramic between + & - rails..
(ground doesn't need to be garbage bin..)
Also, bypass caps should be physicaly mounted as close as possible to IC supply inputs..

[Buffalo Tom]

Thanks Antonis. Changed R3 to 20K.. Still the same. Moved power supply bypass capacitors so they are located direct to IC pins. Also changed the pot to a new 100K pot.. But nothing helps. Still "DC difference" and crackles.. Its not super loud but frustrating and it would be nice to get this right and learn what Im doing wrong.. 

Here is my power supply


And here is the breadboard

[anotherjim]

I wouldn't expect a feedback pot to crackle unless its a bad/dirty pot or the cap to 0v (C3) is leaky.  It may be crackling because it's being told to! An opamp with high input bias current can make the problem seem worse, and the 5532 wants bias higher than some opamps.
In a simple dual supply circuit, the decoupling cap C3 isn't actually needed unless you want a low cut EQ when the gain will be dependant on the impedance of that cap, which in a clean boost you probably don't.

[antonis]

In a simple dual supply circuit, the decoupling cap C3 isn't actually needed..

You're talking about "ideal" zero bias, Jim..
(or else, any DC offset from zero volts is multiplied with NFB loop gain and is present on output..)

I should suggest OP to measure DC voltage between output and GND and, if any, between output and inverting input..
(I also suspect leaky/short C3..)

[antonis]

Tom, does your breadboard exhibit continuity between sets of five holes on + & - rows..??
e.g. does back green aligator clip make contact with left edge of yellow jumper wire..??

[Buffalo Tom]

Tom, does your breadboard exhibit continuity between sets of five holes on + & - rows..??
e.g. does back green aligator clip make contact with left edge of yellow jumper wire..??

I have the potentiometer mounted in a enclosure and the green alligator clip is just chassis ground for that enclosure. Yellow wire connects the two negative rows. All segments it the rows exhibit continuity. So everything is grounded. Will do some DC measurements from the chip to ground tomorrow. And about C3. I have tried to remove that one and go direct with R6 to ground but when I did that the crackling noice got ten times worse. Does that make sense?   

[antonis]

More sense than you might think.. 

By shorting C3 you raise DC gain (hence crakling) by 10 times [1 + (100k//100k)/5k1]

C3 is DC gain "isolator".. It prevents it from passing through R6 so, with feedback resistor (R6) "open"  IC acts as DC unity gain amp..


>I have the potentiometer mounted in a enclosure and the green alligator clip is just chassis ground for that enclosure<

Hmmm... Try to lift up that chassis ground..
(maybe place it on Vin- of NMA0515SC..)

In case it doesn't help either, place a 22μF cap in series with pot one of outer lugs wire (after R5 & C2 joints..)
You can also try it with a smaller cap, just to verify crakle stop..

[merlinb]

5532 is a dual opamp. What are you doing with the other opamp?

[anotherjim]

I notice half the 5532 is not terminated. Link pins 1 & 2 together and link pin3 to 0v. This will stop it from behaving ramdomly and injecting noise into the supply of the used half. You can usually get away with an unterminated opamp, but you never know so it's a good habit to do it all the time.
Sometimes a crackly pot is a crackly pot -  have you tried another? Pots used in experiments are easily damaged. I prefer to use cheap skeleton preset potentiometer for breadboarding fitted direct into the breadboard.

[Buffalo Tom]

Thanks everyone for your help. So here is what Im ended up with. No more crackle!! Changed to 50K alpha pot mounted direct on the breadboard. Instead of the 100K cts pot mounted in the "remote" box. The circuit is more or less a modified Mr black tiger boost. I also put a trimmer to control exact how many dB the gain pot should give me. For example 5216 ohm gave me +20dB boost and 10226 ohm gave me +15dB boost. But the most important thing no more crackle, just a smooth volume control. Im very happy.



But I have a question about my power supply. Should DC input jack negative and OV from the DC/DC converter be connected together?? At the moment they are not and Im using OV as my ground and the incoming DC negative is only used to power the DC/DC converter.