Just a couple of relatively minor points - its looking overall fine.
First one is that you'll end up with a few hundred mV of offset at the outputs of the first 5532s owing to the input bias current (0.2uA) into 10k plus the offset voltages multiplied by the DC gain (X100). Just something to watch out for when debugging the circuit, its not a big impact on its performance as it barely steals any headroom.
Second is that the top 5532 is going to be seeing a less than optimum load - the 1k resistor values are a bit small. They don't need to be this low. Generally its best to keep loads on opamps above 2k for the best distortion performance. Looking into the inverting side of a 4-resistor in-amp configuration, the top 5532 'sees' about 670R assuming a balanced input (a rather big assumption admittedly). That's then in parallel with its own fb resistor.
First one is that you'll end up with a few hundred mV of offset at the outputs of the first 5532s owing to the input bias current (0.2uA) into 10k plus the offset voltages multiplied by the DC gain (X100). Just something to watch out for when debugging the circuit, its not a big impact on its performance as it barely steals any headroom.
Second is that the top 5532 is going to be seeing a less than optimum load - the 1k resistor values are a bit small. They don't need to be this low. Generally its best to keep loads on opamps above 2k for the best distortion performance. Looking into the inverting side of a 4-resistor in-amp configuration, the top 5532 'sees' about 670R assuming a balanced input (a rather big assumption admittedly). That's then in parallel with its own fb resistor.
I think this is a good approach for reducing noise in the circuit, although you may want coupling caps between the stages or a servo arrangement. Here's a link to an excellent example that you might simplify for your situation :
DIY Mic Preamp Kits - Seventh Circle Audio
Solving the preamp noise is only part of the problem, however, the net result will still be disappointing because of the noise figure of the microphone selected. I experienced this directly when I obtained a MXL V63M mic, very good specs, except for the noise. When compared side by side with a Rode NT2A or AKG 3000B (both with similar specs but lower noise), through identical preamps, it always gets left in the mic locker.
DIY Mic Preamp Kits - Seventh Circle Audio
Solving the preamp noise is only part of the problem, however, the net result will still be disappointing because of the noise figure of the microphone selected. I experienced this directly when I obtained a MXL V63M mic, very good specs, except for the noise. When compared side by side with a Rode NT2A or AKG 3000B (both with similar specs but lower noise), through identical preamps, it always gets left in the mic locker.
Lee did not say what the mic was used for. In many applications the absolute noise-floor of the microphone is not important.
The last circuit he posted, using 15V (termed vb, quaint old for battery voltage), the mic was powered through 6k8 resistors. That value seems far to high for operating this mic from a 15v supply.
Also, I never see any input protection diodes on Lee's circuits. Unplugging the mic or turning on/off the phantom power could do loverly things to the IC.
E
The last circuit he posted, using 15V (termed vb, quaint old for battery voltage), the mic was powered through 6k8 resistors. That value seems far to high for operating this mic from a 15v supply.
Also, I never see any input protection diodes on Lee's circuits. Unplugging the mic or turning on/off the phantom power could do loverly things to the IC.
E
I think you should look at those chips I mentioned before. What you're doing is what is in those chips but they can optimize better than we can.
G²
Thanks for all the input. 
Good point with the input bias current. I try to buy some of the LM4562 that has only 72nA input bias current.
I would like to avoid DC blocking caps between stages if possible and rather buy a slightly more expensive opamp that anyway performs better.
After looking at the datasheet I agree the low load for the first opamps is indeed a problem. I better use higher value resistors. But on the other hand side there's a noise penality from high resistor values...
Yes, the 6k8 resistors are too high for such a low phantom power voltage. I should lower them. I think about something dangerous: Connecting pin 1 of the XLR to the negative rail but this might smoke some mics. No, better lower the resistors.
Okay, protection diodes added. I barely see protection diodes on DIY amps inputs. Not used properly they might degrade sound but without them ESD on the input will degrade the input stage. I will use lower values than drawn in schematic, just had none in library.
Stratus, I looked at the chips you mentioned. They might be a good idea to use but I can't source them easily.
By the way: Attached another schematic I copied from when I built the preamp many years ago. As you see I took the worst parts from a lot of circuits to build the worst mic preamp ever built.
Good point with the input bias current. I try to buy some of the LM4562 that has only 72nA input bias current.
I would like to avoid DC blocking caps between stages if possible and rather buy a slightly more expensive opamp that anyway performs better.
After looking at the datasheet I agree the low load for the first opamps is indeed a problem. I better use higher value resistors. But on the other hand side there's a noise penality from high resistor values...
Yes, the 6k8 resistors are too high for such a low phantom power voltage. I should lower them. I think about something dangerous: Connecting pin 1 of the XLR to the negative rail but this might smoke some mics.
No, better lower the resistors. Okay, protection diodes added.
I barely see protection diodes on DIY amps inputs. Not used properly they might degrade sound but without them ESD on the input will degrade the input stage. I will use lower values than drawn in schematic, just had none in library. Stratus, I looked at the chips you mentioned. They might be a good idea to use but I can't source them easily.
By the way: Attached another schematic I copied from when I built the preamp many years ago. As you see I took the worst parts from a lot of circuits to build the worst mic preamp ever built.
Attachments
LM4562 is a nice part but so far I haven't got it to give a good sound without some tweaking. It seems that its input stage is much more RF sensitive than the humble old 5532. In a pair of active speakers I bought recently, the sound to my ears was too shrill, I fixed that by downgrading to NE5532 from LM4562.
Bear in mind that the noise penalty from resistors goes as the squareroot of their value - so after a gain of 100 as you have here, you'd need resistor values of 100,000X to get into the realm of compromising noise. That's assuming of course that the opamp's current noise isn't going to dominate - which in practice it would. You used 10R in effect as the lower leg but that's way lower than the opamp's voltage noise. Just going up to 22R would allow a feedback resistor which didn't compromise distortion at the same gain and barely increase noise. Ah, I see from looking at your new schematic that's exactly what you've done
Yes, the 6k8 resistors are too high for such a low phantom power voltage. I should lower them. I think about something dangerous: Connecting pin 1 of the XLR to the negative rail but this might smoke some mics.
Why would putting pin1 at the negative rail potentially smoke some mics? I must admit here I'm no expert in mic preamps from the practical side
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.