Hi Folks,
I've breadboarded a JFet, class A mic preamp. (https://www.tangible-technology.com/ipr/AE230/html/wk_6/hamptone_fet_scan.pdf).
It requires 2 voltages:
I'm currentlying regulating a 28V Mean Well wall-wart down to 24V, as shown in schematic.
For 48V, I'm using a 48V Mean Well wall-wart.
Even without the 48V (using a dynamic mic), the noise is high, "hiss" (not 60Hz).
For the PCB version, which I'll design soon, I could use advice on how to build the dual-voltage power supply. (i.e. voltage doubler for 48V? dual-secondary transformer? all linear, remove the regulator?)
Requirements:
Thanks for any ideas!
BW
I've breadboarded a JFet, class A mic preamp. (https://www.tangible-technology.com/ipr/AE230/html/wk_6/hamptone_fet_scan.pdf).
It requires 2 voltages:
- 24V for Vcc
- 48V for phantom powered mics
I'm currentlying regulating a 28V Mean Well wall-wart down to 24V, as shown in schematic.
For 48V, I'm using a 48V Mean Well wall-wart.
Even without the 48V (using a dynamic mic), the noise is high, "hiss" (not 60Hz).
For the PCB version, which I'll design soon, I could use advice on how to build the dual-voltage power supply. (i.e. voltage doubler for 48V? dual-secondary transformer? all linear, remove the regulator?)
Requirements:
- 24V for Vcc
- 48V for phantom powered mics
- budget / affordable
- low noise
- small-ish component count (I can design PCBs, but if it gets too complex, I will not complete it)
Thanks for any ideas!
BW
Using a voltage doubler with a single transistor cap multiplier for regulation may be the easiest solution for the 48V supply - it's not like you need more than about 4 mA there. You'd need a transformer of about 5 VA (maybe 7.5 or 10) with something like a 22-25 V secondary, so nothing super huge either way.
I would expect the whole amplifier with 1:10 input transformer and two JFP modules to have a gain of approximately 20 dB + 45 dB + 45 dB = 110 dB when fully cranked up, which for a mic preamp is way excessive, so no surprise if it seems noisy. You should be comparing it to a known preamp with the same acoustic signal level, microphone and location and observe signal to noise ratio instead.
The circuit does have some quirks - it does require the 33R/220µ RC filtering as shown since inherent PSRR is low, and output stage current is poorly defined, with various transistor models for the current source yielding anywhere from as low as 27 mA to as high as 80 mA in simulation depending on each type's beta. Transistor beta is not the kind of thing you want to be relying on, given that it has a habit of varying between units and with temperature. At the very least, I would suggest replacing the 47k by a 10k while adding a 3k3 from NPN base to ground... that should keep output stage current within a few mA of 40 mA over a fairly wide range of transistors.
I would expect the whole amplifier with 1:10 input transformer and two JFP modules to have a gain of approximately 20 dB + 45 dB + 45 dB = 110 dB when fully cranked up, which for a mic preamp is way excessive, so no surprise if it seems noisy. You should be comparing it to a known preamp with the same acoustic signal level, microphone and location and observe signal to noise ratio instead.
The circuit does have some quirks - it does require the 33R/220µ RC filtering as shown since inherent PSRR is low, and output stage current is poorly defined, with various transistor models for the current source yielding anywhere from as low as 27 mA to as high as 80 mA in simulation depending on each type's beta. Transistor beta is not the kind of thing you want to be relying on, given that it has a habit of varying between units and with temperature. At the very least, I would suggest replacing the 47k by a 10k while adding a 3k3 from NPN base to ground... that should keep output stage current within a few mA of 40 mA over a fairly wide range of transistors.
Thanks for the analysis & ideas sgrossklass. I like the voltage doubler approach. I'll plan on using a single transformer with 1 secondary to supply ~22VAC to both the doubler & the 24VDC power supply (a bridge rectifier +caps, which will feed a LM7824 regulator for a final 24VDC). Most power transfomers offer center-tap; just leave it floating / unconnected, correct?
You're right, 110dB is a lot of gain. I was comparing it to my Soundcraft Signature Mk2 preamps, which have 60dB of gain. Neither of my test mics (CAD condensor, AT dynamic) need that much gain. Should I consider reducing the gain? (The input transformer, a Neutrik NTE10/3, has 1:3:10 secondary taps, so perhaps I should put a switch in, allowing 3, 7, & 10 voltage step-up options). If so, would I need a make-before-break switch, or can a microphone output survive a switching of the secondary with a break-before-make switch?
I'll try the current-source resistor changes you suggested on the breadboard. The 47K is causing the top of the test sine-wave to clip, so at the very least, I'd like to bias the output stage closer to the center of the 0V-24V range. Why is it important to have well defined output stage current? Is it so that, regardless of Hfe, the gain of the preamp can be specified? Or that, with wildly varying Hfe, each build could accidentally get more gain than I want, & therefore more noise?
You're right, 110dB is a lot of gain. I was comparing it to my Soundcraft Signature Mk2 preamps, which have 60dB of gain. Neither of my test mics (CAD condensor, AT dynamic) need that much gain. Should I consider reducing the gain? (The input transformer, a Neutrik NTE10/3, has 1:3:10 secondary taps, so perhaps I should put a switch in, allowing 3, 7, & 10 voltage step-up options). If so, would I need a make-before-break switch, or can a microphone output survive a switching of the secondary with a break-before-make switch?
I'll try the current-source resistor changes you suggested on the breadboard. The 47K is causing the top of the test sine-wave to clip, so at the very least, I'd like to bias the output stage closer to the center of the 0V-24V range. Why is it important to have well defined output stage current? Is it so that, regardless of Hfe, the gain of the preamp can be specified? Or that, with wildly varying Hfe, each build could accidentally get more gain than I want, & therefore more noise?