Help! Can't get a NE5532 stage to work.

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
It should be simple to get a basic NE5532 stage to work but I've been trying for days, built 2 different regulated supplies and 3 circuits and can't get anything to work. Loud buzzing and distorted sound on everything. Frustratingly, a chi-fi single 12v supply pcb fed with 15vDC from my bench supply works perfectly - silent and good sound. Pictured in red. No schematic available.

I have +-15v out of the regulated supplies, standard book circuit. The whole system is grounded through my tube amp via the interconnects. I also put 1uF in series on the input to the NE5532 stage in case there was DC offset from the DAC (output stage there is also a NE5532). Didn't help. I could try 1uF on the output of the NE5532.

I've built a lot of tube amps but I'm unfamiliar with chips like this and have no idea what's not working. I need help!
 

Attachments

  • MyNE5532-1.jpg
    MyNE5532-1.jpg
    115 KB · Views: 263
  • DAC+ System Layout.png
    DAC+ System Layout.png
    35.8 KB · Views: 258
  • NE5532 stage chifi-mod.png
    NE5532 stage chifi-mod.png
    33.6 KB · Views: 277
The circuit looks good.
So either a mistake in the wiring, a blown IC or you didn't connect zero volts on supply to zero volts in circuit.

Thanks for contributing! As the topology shows, I have 20-0-20v AC coming from one box. This goes into the rectification and voltage circuit board which is a standard circuit and outputs +-15v DC. The 0v is referenced to the screen of the interconnects to the DAC (not grounded) and to the Tube amp (grounded). This grounding system works with the single 12v supply pcb shown in red. Doesn't seem to work with the circuits I've been building.

What are you suggesting exactly regarding grounding? I could ground the centre point of the 20-0-20 toroidal transformer at the transformer itself - should I do that?
 
Last edited:
I connected the centre point of the toroidal AC output to earth and then measured what was coming out of the voltage regs. And I'm getting +15v and -4v. When I made the voltage reg circuit and measured it, it was regular +-15v. So something is affecting the negative rail - maybe something in the NE5532 circuit. I don't know if the opamp is frying the negative reg or visa versa.
 
The first step with troubleshooting is measung supply voltage (+-15V) and op-amp output dc-voltage (0V). Any significant deviations will point you to some fault.

Indeed. I went through one of my 2 voltage reg stages and corrected a dry joint, and I now have +-15v. I'll connect it up to the NE5532 stage and see if it stays stable. The other voltage reg stage is also faulty on the negative rail. Perplexing. When I corrected the first voltage reg stage I took out the Schottky diodes and put in IN4007 instead since that's what's usually in the schematics.
 
Last edited:
Administrator
Joined 2007
Paid Member
One point to note, the 5532 could (will) have a very significant DC offset when used in the circuit as drawn due to massively unbalanced input resistances.

Also the 78/79 regs can suffer unpredictable latch up under certain conditions which cause the voltage to be incorrect, easily fixed by adding inverse parallel diodes across the regulator and also the stabilised rails.

Edit... And the image from the web is wrong... D2 should be the other way around.
 

Attachments

  • Annotation 2020-05-10 180737.jpg
    Annotation 2020-05-10 180737.jpg
    23.4 KB · Views: 176
Thanks Mooly.
- Can you talk more about the "massively unbalanced input resistances" and how to correct them?

- Are you saying that in the image you have posted up in the last post, the D2 diode as shown in that image is incorrect? i.e. it should be the reverse of D1?

D2 is wrong in diagram.
Try pulling 5532 and see if power supply goes back up.
If it does the 5532 is fried.
 
Administrator
Joined 2007
Paid Member
Thanks Mooly.
- Can you talk more about the "massively unbalanced input resistances" and how to correct them?

- Are you saying that in the image you have posted up in the last post, the D2 diode as shown in that image is incorrect? i.e. it should be the reverse of D1?

Yes, D2 needs turning 180 degrees. It is drawn the wrong way around.

The 5532 has relatively high input bias currents. These are the tiny currents that flow out of the opamp input pins. Both inputs will be very well matched but the current could be as high as 800 nano amps (from the data sheet).

That doesn't sound much but that current flows out of the input pin and into the 39k and so that develops a voltage of 31 millivolts across that resistor.

You have the opamp configured as DC coupled and so that offset gets multiplied by the gain of the circuit... and you get a high offset at the output.

So to eliminate it (and you could just use AC coupling at the input and output) you should make the feedback resistor equal 39k (which keeps the two input offset voltages the same (31mv for each but zero between the two) and AC couple the 1k to keep the gain at DC equal to unity.

Now 39k may not suit in terms of noise and so on and so the alternatives are to use an opamp with lower input bias currents such as the LM4562 or even a FET opamp which has essentially zero input bias current.
 
Thanks Mooly.
- Can you talk more about the "massively unbalanced input resistances" and how to correct them?
It not a problem, one input is 39k to ground, and with 0.2µA of bias current that's a few millivolts of offset, worst case bias current is 0.8µA (30mV offset).
- Are you saying that in the image you have posted up in the last post, the D2 diode as shown in that image is incorrect? i.e. it should be the reverse of D1?
Yes, since the current flows the other way in the negative regulator.
 
Good news - it's all working. Did some fault finding, installed protection diodes, earthed the centre point of the 20-0-20 toroid. No hum, music coming out now. Now I can start optimising the setup.

I have a few different op-amps to play with -
NE5532
LM4562
OPA2134
LM6172
OPA2228
also NE5534, interesting and will be a future project.

I also have two setups - the new dual supply stage and my existing single supply chi-fi board which sounds pretty good in fact. All SMD except the socketed op-amp. Right now listening to a LM4562 in the single supply board. Better than the NE5532, particularly in the treble, and more detail. It's pretty nice. The idea of an initial op-amp stage is clearly realistic in my tube setup.

Thanks guys for all your help up to this point - meant a lot!! I'll post any more news about tweaking it.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.