NewYorkDave
Well-known member
Thanks for all the input, guys. This is as much effort as we should expend, because the circuit is just a freebie for a friend. I'll post the final result.
Low-noise opamp for battery-powered preamp.
- Thread starter NewYorkDave
- Start date
Help Support GroupDIY Audio Forum:
Opamps run on 2.5 volts and microamps nowdays.
But how much power do they consume when pumping out 40 db?
The quiesent might not be a factor as soon as you record.
9 volts means field work to me, which means build it rugged.
Your camping, you leave it outside the tent, it rains, maybe a couple of lightning stikes here and there, people spilling bear on it, so a fet opamp might not be a good idea.
The transistor thing would certainly be rugged.
Whats that site with all the bootleggers? Dat Heads or something?
All they do is talk about 9 volt bootleg machines, stereo mics in the hat, the whole 9 yds.
Steelers, and I'll give you 5. Finesse offense fails in the big ones.
But how much power do they consume when pumping out 40 db?
The quiesent might not be a factor as soon as you record.
9 volts means field work to me, which means build it rugged.
Your camping, you leave it outside the tent, it rains, maybe a couple of lightning stikes here and there, people spilling bear on it, so a fet opamp might not be a good idea.
The transistor thing would certainly be rugged.
Whats that site with all the bootleggers? Dat Heads or something?
All they do is talk about 9 volt bootleg machines, stereo mics in the hat, the whole 9 yds.
Steelers, and I'll give you 5. Finesse offense fails in the big ones.
NewYorkDave
Well-known member
Right, like the Shure FP11--which is an interesting circuit to study if you're into battery-powered preamps. It's old and the opamp it used isn't even available anymore, it's since been superseded by a lower-noise version, and it was never treasured for its sound quality. But it was a workhorse.
Good one!
I won't edit.
Not in the spirit of learning, but if this is going to be used for recording, don't overlook buying a portable casette player or something that already has a mic input.
Pick your price point on this one.
I won't edit.
Not in the spirit of learning, but if this is going to be used for recording, don't overlook buying a portable casette player or something that already has a mic input.
Pick your price point on this one.
skipwave
Well-known member
[quote author="Gus"]Why not a 2 or 3 transistor stage, fuzz face, neve, RCA hobby transistor manual type. I think ESP has a simple two transistor one.[/quote]
Smartass. :razz:
Smartass. :razz:
bcarso
Well-known member
[quote author="CJ"]Opamps run on 2.5 volts and microamps nowdays.[/quote]
But watch out for crossover distortion with very low quiescent current op amps. Also many of the low power amps are severely limited in voltage---often 5.5V abs max and so forth. For reasonable runtime with a typical 9V alkaline a few mA can surely be spared---add a super efficient LED with a resistor and zener in series running at a 100uA for a low batt indication perhaps. It will also remind you to turn the power off when not in use.
As far as high levels signals pulling excess current, the only place you particularly need low impedance is in the first section, for low noise---and then the signals are small. After that first 20dB you can run at higher Z's and still be dominated by the first section. So, not a big issue.
But watch out for crossover distortion with very low quiescent current op amps. Also many of the low power amps are severely limited in voltage---often 5.5V abs max and so forth. For reasonable runtime with a typical 9V alkaline a few mA can surely be spared---add a super efficient LED with a resistor and zener in series running at a 100uA for a low batt indication perhaps. It will also remind you to turn the power off when not in use.
As far as high levels signals pulling excess current, the only place you particularly need low impedance is in the first section, for low noise---and then the signals are small. After that first 20dB you can run at higher Z's and still be dominated by the first section. So, not a big issue.
NewYorkDave
Well-known member
I had an unexpected half-hour of free time this morning, so I sketched this for my friend:
http://groupdiy.twin-x.com/albums/userpics/10031/BattPowerMicPreamp.png
It's not an "optimized" design by any means, but I think it'll do what he needs it to do.
http://groupdiy.twin-x.com/albums/userpics/10031/BattPowerMicPreamp.png
It's not an "optimized" design by any means, but I think it'll do what he needs it to do.
> It's not an "optimized" design by any means
You may have to hard-ground the bottom of the volume pot, and AC-couple in and out.
As drawn, the maximum attenuation at 20Hz is "only" 40dB (10K and 100uFd), which may be fine, but it also dumps first stage output back to first stage input. I don't think it will be unstable, but somethng to watch for.
You may have to hard-ground the bottom of the volume pot, and AC-couple in and out.
As drawn, the maximum attenuation at 20Hz is "only" 40dB (10K and 100uFd), which may be fine, but it also dumps first stage output back to first stage input. I don't think it will be unstable, but somethng to watch for.
NewYorkDave
Well-known member
Ah! Excellent point. I'll change the V/2 decoupling to 1000uF. The can size should still be manageable in a 16V cap.
moamps
Well-known member
Hi,
IMHO, the bottom end of volume pot shouldn't be connected to V/2. The bias comes from the IC1A's output. The capacitor 1uF should be then changed to elco 10uF or higher and the pot to 100klog.
Regards,
Milan
IMHO, the bottom end of volume pot shouldn't be connected to V/2. The bias comes from the IC1A's output. The capacitor 1uF should be then changed to elco 10uF or higher and the pot to 100klog.
Regards,
Milan
NewYorkDave
Well-known member
The drawing has been updated. I re-deployed the two 1uF caps to better effect and, to avoid bass roll-off from the 1uF coupling, changed the pot to 100K-A.
http://groupdiy.twin-x.com/albums/userpics/10031/BattPowerMicPreamp.png
Thanks to all for the suggestions.
http://groupdiy.twin-x.com/albums/userpics/10031/BattPowerMicPreamp.png
Thanks to all for the suggestions.
bcarso
Well-known member
Looks good now. I might add a couple of teensey caps (10-20 pF) across the two 100k feedback R's to limit bandwidth and ensure the stuff hanging off the inverting input doesn't shift phase into the oscillatory realm (not typically a problem for a 3MHz GBW product, but if the layout is sloppy enough for the gain switch it's not out of the question).
> are you sure you want a 062?
Agree. The 072/082 take more supply current than 062, but I have never gone broke putting 9V batts in TL072 toys. I suppose it depends on your gig: a quickie or an all-day affair.
bcarso showed that even the 062 has enough slew, though is not over-kill for GBW. (I can't complain too much because I suggested a single amp.)
But if we could believe TI's datasheets, the 062 has twice the noise of the 072/082 parts. Makes sense. And by bcarso's numbers, the 072/082 is marginal, the 062 noise is much higher than reflected mike noise. Again, depends on the gig.
My instinct would be to use the 072, or at least until the battery bills piled up. Of course a socket leaves you the option to change chips as needed.
Agree. The 072/082 take more supply current than 062, but I have never gone broke putting 9V batts in TL072 toys. I suppose it depends on your gig: a quickie or an all-day affair.
bcarso showed that even the 062 has enough slew, though is not over-kill for GBW. (I can't complain too much because I suggested a single amp.)
But if we could believe TI's datasheets, the 062 has twice the noise of the 072/082 parts. Makes sense. And by bcarso's numbers, the 072/082 is marginal, the 062 noise is much higher than reflected mike noise. Again, depends on the gig.
My instinct would be to use the 072, or at least until the battery bills piled up. Of course a socket leaves you the option to change chips as needed.
bcarso
Well-known member
[quote author="eded"]can i ask a question about this application ?
i used a LT1012 once that seemed quite good on current draw.
very low. would this op amp from linear work in this application as a drop in thus saving battery power ??
just trying to learn.[/quote]
Did you look at the datasheet in light of the discussion so far?
http://www.linear.com/pc/downloadDocument.do?navId=H0,C1,C1154,C1009,C1100,P1221,D3013
Yes, current drain is low.
Look at voltage noise, current noise, slew rate, and any data they may provide on how things work at low rail voltages.
i used a LT1012 once that seemed quite good on current draw.
very low. would this op amp from linear work in this application as a drop in thus saving battery power ??
just trying to learn.[/quote]
Did you look at the datasheet in light of the discussion so far?
http://www.linear.com/pc/downloadDocument.do?navId=H0,C1,C1154,C1009,C1100,P1221,D3013
Yes, current drain is low.
Look at voltage noise, current noise, slew rate, and any data they may provide on how things work at low rail voltages.
AMZ-FX
Well-known member
Dave,
My handy-dandy noise calculator says that in your applications, these opamps have noise at 1kHz as follows:
TL072 23.2nV
MC33178 16.5nV
RC4560 16.9nV
5532 17.6nV
The bipolar opamps do a bit better here than the jfet 072, as I would expect with your source.
Just grab a common (and cheap) 4560, which is used in many semi-pro mixers, and go with it. It will probably draw a bit less current from the battery than the 5532.
regards, Jack
My handy-dandy noise calculator says that in your applications, these opamps have noise at 1kHz as follows:
TL072 23.2nV
MC33178 16.5nV
RC4560 16.9nV
5532 17.6nV
The bipolar opamps do a bit better here than the jfet 072, as I would expect with your source.
Just grab a common (and cheap) 4560, which is used in many semi-pro mixers, and go with it. It will probably draw a bit less current from the battery than the 5532.
regards, Jack
bcarso
Well-known member
[quote author="AMZ-FX"]Dave,
My handy-dandy noise calculator says that in your applications, these opamps have noise at 1kHz as follows:
TL072 23.2nV
MC33178 16.5nV
RC4560 16.9nV
5532 17.6nV
The bipolar opamps do a bit better here than the jfet 072, as I would expect with your source.
Just grab a common (and cheap) 4560, which is used in many semi-pro mixers, and go with it. It will probably draw a bit less current from the battery than the 5532.
regards, Jack[/quote]
What is your handy-dandy noise calculator, and what does it use as input data? For the 4560 in particular at least the tabulated values are maddeningly obtuse (for example TI parrots the other mfg.'s "1.2uV RIAA" [!] with a 30kHz filter). JRC manages to give us a curve for e sub n out to 1kHz, which shows about 13nV/root Hz there. However some measurements I did on some of the Rohm parts gave me a considerably lower number.
My handy-dandy noise calculator says that in your applications, these opamps have noise at 1kHz as follows:
TL072 23.2nV
MC33178 16.5nV
RC4560 16.9nV
5532 17.6nV
The bipolar opamps do a bit better here than the jfet 072, as I would expect with your source.
Just grab a common (and cheap) 4560, which is used in many semi-pro mixers, and go with it. It will probably draw a bit less current from the battery than the 5532.
regards, Jack[/quote]
What is your handy-dandy noise calculator, and what does it use as input data? For the 4560 in particular at least the tabulated values are maddeningly obtuse (for example TI parrots the other mfg.'s "1.2uV RIAA" [!] with a 30kHz filter). JRC manages to give us a curve for e sub n out to 1kHz, which shows about 13nV/root Hz there. However some measurements I did on some of the Rohm parts gave me a considerably lower number.
AMZ-FX
Well-known member
I wrote the noise calc for my own personal use... it needs the Noise voltage in nV/√Hz and the Noise current in pA/√Hz, along with the resistor values.
There is a similar one online at: http://tangentsoft.net/audio/calc.html
I got the noise specs from a datasheet on the RC4560... I ran the calculations before I had Dave's schematic for reference so I just picked some values to use... the relative performance of the opamps should be about the same even without the schematic values.
regards, Jack
There is a similar one online at: http://tangentsoft.net/audio/calc.html
I got the noise specs from a datasheet on the RC4560... I ran the calculations before I had Dave's schematic for reference so I just picked some values to use... the relative performance of the opamps should be about the same even without the schematic values.
regards, Jack
