TL072 as mike input preamp

[MarcelvdG]

At the moment, the matter is really of academic interest, in case the information becomes useful later. There's no practical reason to use a TL072 instead of other lower-noise opamps like the NE5532, RC4558, etc. which are also dirt cheap. I've designed the PCB to accept just about any dual opamp will standard pin-outs.

Anyway, here's the basic schematic:

With an NE5532 or an NE5532A, R4 would be the main noise source.

 

[indianajo]

The question is, what can you get in your country that is not counterfeit? That $5 IC may be worth stealing at the border and replacing with a $.38 part relabled with a silk screen. ST33078 & NJM2068 are $.38 and $.80 parts already. 5532 is also common & cheap, although it requires a little more potent power supply if you're using a number of them. NJM4580 is $.71, quiet, and is nice if you're driving cables over 2 m but costs $10 freight for me to get here since my regular supplier doesn't have it. As far as system cost, my disco mixer with the 4558's cost $15 at the flea market. + about $20 in parts & freight to upgrade it to good sound.
Some border people don't even substitute a part. In many countries, it is best to have an authorized distributor like farnell or RS get their real parts across the border in large lots with a courier or handling service. I've got a bit of negative experience with a country south of here, where a $2000 scope disappeared on the way back from a warrenty shipment.

 

[Pimpom]

With an NE5532 or an NE5532A, R4 would be the main noise source.

Agreed. This is a hypothetical design with a TL072 for which the equivalent noise resistance is about 20 kΩ. If low-noise opamps were to be used, the value of R4 and the others could be modified without altering the PCB design. Ultra low noise opamps will require even more careful consideration.

 

[Markw4]

Although many dynamic mics are low impedance, they often sound much better very lightly loaded. For an SM-57, 2k-10k preamp input Z is the most useful range in terms of sound quality.

Also, TL072 can be a useful device for some things, but distortion goes up into a low Z load. Best to keep the load at least a few K.

 

[davidsrsb]

With a bipolar opamp, capacitor C1 is too small as the current noise will appear as voltage across it at low frequencies.

The TL072s inability to drive low resistance makes the choice of R4 tricky, with a NE5532 I would go down to about 330R

 

[MarcelvdG]

The effect of C1 is not so bad, I think. With a 200 ohm microphone and 330 ohm series resistor (presumably meant to protect the op-amp against excessive input signals), the reactance of the 10 uF C1 only comes into play below about 30 Hz. That's only 10 Hz out of 19980 Hz when you are interested in the unweighted noise from 20 Hz to 20 kHz. With A-weighting or ITU-R 468 weighting the effect of C1's reactance will be even smaller.

 

[Pimpom]

Thanks for your continued interest, guys. But please remember that this is not about how to make a high-quality microphone preamplifier, but about how a TL072 is likely to do in this function. How bad it sounds, if you want to put it that way.

 

[DPH]

Pimpom, have you gone over to micbuilders yahoo site and looked at some of their preamps? I think it's really going to depend on the cartridge you're planning to use. The low input current noise is really going to help you for electrets, even if the E_n is pretty high.

Also think about (even with some low power bipolars) putting a class-A buffer inside the TL072's feedback network as it doesn't like being loaded. Jim Williams's Linear Technology's app notes are a gold mine for info. Easy things to play around with if you've got the TL072's already and want to see what you can do.