DC-DC CMOS (HEX INVERTER) REVERSE BIAS?

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
Can you guys show me a picture? Because in my head, idk if what im thinking is what you guys are saying. In my head, I see a turret (looks like a standoff?) soldered to the pcb where the wire would normally go. And then the capsule center termination wire soldered to that. Is that right? And what is the "in the air" thing look like? Thanks guys.
 
Yes the capsule wires are directly soldered to the pcb, where it says to solder them. Not sure where else I could solder them?

Rogs - The PCB came with the 4584D already soldered in place, should I replace it with a 74c14 or 40106 instead? I can always order one or two from mouser, they aren't pricey at all.

I was also wandering, instead of the OPA1642...could I use a OPA1612? I have a few of them lying around here in my parts bin. Just wandering.

Thanks guys!
Probably not worth replacing the 4584 IC if it's already in place..... The lower hysterisis level will simply result in a slightly higher oscillator frequency, which shouldn't cause any problems.

The OPA1612 is not really suitable in this application. Although the noise and distortion figures are execellent, it draws about twice the current of the OPA164*series.
More importantly, it doesn't have a JFET input, so it doesn't really have a high enough input impedance for the impedance converting function required here.

You might improve the signal to noise ratio marginally by making the audio output single sided. To try that out simply remove the two 2k2 resistors that supply the second amplifier, and then join pins 6 and 7 together.
It will reduce the output level by 6dB, but should reduce the noise level by about 9dB.
This sample waveform shows the kind of improvement you might expect....
https://tinyurl.com/opic-noiseThe first reference tone is followed by the residual noise output... This is from a single sided output.
A second similar tone follows, and a second noise only output......This is from a differential output.
The level of the 1st tone has been increased by 6dB in the DAW software.
Both noise figures are really 'down in the weeds', to quote the Alice OPA designer Jules Ryckebusch :) - but the single sided output is about 3dB better, if very low noise is an important parameter for your requirements.
 
Sometimes a single cap is not enough. Try inserting a 100-200r resistor instead of the link that's circled in red, and maybe increase C10.
There is another, very distant possibility of the CD4584 being unstable. Did you try another?
I added 200r resistor at the part you mentioned, and changed C10 to 47uF. The noise actually got worse sadly.
 
Probably not worth replacing the 4584 IC if it's already in place..... The lower hysterisis level will simply result in a slightly higher oscillator frequency, which shouldn't cause any problems.

The OPA1612 is not really suitable in this application. Although the noise and distortion figures are execellent, it draws about twice the current of the OPA164*series.
More importantly, it doesn't have a JFET input, so it doesn't really have a high enough input impedance for the impedance converting function required here.

You might improve the signal to noise ratio marginally by making the audio output single sided. To try that out simply remove the two 2k2 resistors that supply the second amplifier, and then join pins 6 and 7 together.
It will reduce the output level by 6dB, but should reduce the noise level by about 9dB.
This sample waveform shows the kind of improvement you might expect....
https://tinyurl.com/opic-noiseThe first reference tone is followed by the residual noise output... This is from a single sided output.
A second similar tone follows, and a second noise only output......This is from a differential output.
The level of the 1st tone has been increased by 6dB in the DAW software.
Both noise figures are really 'down in the weeds', to quote the Alice OPA designer Jules Ryckebusch :) - but the single sided output is about 3dB better, if very low noise is an important parameter for your requirements.
I am about to try this out now! Thank you
 
That is weird. I can't remotely find an explanation.
Yeah, when I take out the dc hex power board, and replace it with the power board from the MP-V57 circuit kit from Microphone Parts.com, the noise is almost non existent.

Same alice op amp board, sounds great. I am thinking my dc / dc board might just be faulty? Or maybe the bias voltage of 80v is too much. The V57 power board is biased at 60v.
 
Probably not worth replacing the 4584 IC if it's already in place..... The lower hysterisis level will simply result in a slightly higher oscillator frequency, which shouldn't cause any problems.

The OPA1612 is not really suitable in this application. Although the noise and distortion figures are execellent, it draws about twice the current of the OPA164*series.
More importantly, it doesn't have a JFET input, so it doesn't really have a high enough input impedance for the impedance converting function required here.

You might improve the signal to noise ratio marginally by making the audio output single sided. To try that out simply remove the two 2k2 resistors that supply the second amplifier, and then join pins 6 and 7 together.
It will reduce the output level by 6dB, but should reduce the noise level by about 9dB.
This sample waveform shows the kind of improvement you might expect....
https://tinyurl.com/opic-noiseThe first reference tone is followed by the residual noise output... This is from a single sided output.
A second similar tone follows, and a second noise only output......This is from a differential output.
The level of the 1st tone has been increased by 6dB in the DAW software.
Both noise figures are really 'down in the weeds', to quote the Alice OPA designer Jules Ryckebusch :) - but the single sided output is about 3dB better, if very low noise is an important parameter for your requirements.
It helped a little bit, but I didn't like the fact I had to gain more on the mic preamp, adding preamp noise. My whole goal with this build, was to find a super sensitive mic, that I can use with low gain, with low noise. I guess i'll just have to buy a TLM103 lol.
 
It helped a little bit, but I didn't like the fact I had to gain more on the mic preamp, adding preamp noise. My whole goal with this build, was to find a super sensitive mic, that I can use with low gain, with low noise. I guess i'll just have to buy a TLM103 lol.
As I mentioned, the second differential audio output will add about an extra 3dB of noise. In the demonstration audio sample, the additional gain required was added digitally, so no additional analogue noise added there.

If your preamp adds signaficant extra noise as well as gain, then of course the comparison will fail. The perceived noise floor will also depend on the sensitivity of the capsule. So a more sensitive capsule will require less overall gain for the same output signal.
That doesn't affect the absolute noise floor of course, but it can help practically.

The TLM 103 you described has a published noise figure of 7dB(A). That's about 1dB better than you might expect from a single sided OPA preamp, using a the OPA 1641.
I use my 4th Generation Rode NT1 as my noise reference. It has a published noise figure of 4.5dB(A). I recently fitted an OPIC preamp into the mic, during some testing, and observed a noise floor of c. 8dB(A) . So only about 1dB worse than the TLM103.

In addition, the TLM103 has a sensitivity about 3dB lower than the Rode. (-32dB v. -29dB ). So to get the same output level, the Neumann will require an extra 3dB of channel gain.
Now, whether that will be better or worse than an OPA preamp will depend on the sensitivity of the capsule being used?

I should add that the Neumann has an extra 6dB of headroom at the top of the input signal range, compared to the Rode, (138dB v. 132dB ) but that's not the end of the scale we're discussing here! :)
 
As I mentioned, the second differential audio output will add about an extra 3dB of noise. In the demonstration audio sample, the additional gain required was added digitally, so no additional analogue noise added there.

If your preamp adds signaficant extra noise as well as gain, then of course the comparison will fail. The perceived noise floor will also depend on the sensitivity of the capsule. So a more sensitive capsule will require less overall gain for the same output signal.
That doesn't affect the absolute noise floor of course, but it can help practically.

The TLM 103 you described has a published noise figure of 7dB(A). That's about 1dB better than you might expect from a single sided OPA preamp, using a the OPA 1641.
I use my 4th Generation Rode NT1 as my noise reference. It has a published noise figure of 4.5dB(A). I recently fitted an OPIC preamp into the mic, during some testing, and observed a noise floor of c. 8dB(A) . So only about 1dB worse than the TLM103.

In addition, the TLM103 has a sensitivity about 3dB lower than the Rode. (-32dB v. -29dB ). So to get the same output level, the Neumann will require an extra 3dB of channel gain.
Now, whether that will be better or worse than an OPA preamp will depend on the sensitivity of the capsule being used?

I should add that the Neumann has an extra 6dB of headroom at the top of the input signal range, compared to the Rode, (138dB v. 132dB ) but that's not the end of the scale we're discussing here! :)
Well then there is definitely something wrong with mine, because the TLM103 I had was way quieter and more sensitive than the build I just made. I haven't tried the NT1 but if you say it's even quieter...then I guess at the price, it's a better option for me right now, as I can't afford another $1100 microphone at the moment, after having to buy an oscilloscope and some other electronics stuff because this is a hobby that is quickly turning into an obsession.
 
Well then there is definitely something wrong with mine, because the TLM103 I had was way quieter and more sensitive than the build I just made. I haven't tried the NT1 but if you say it's even quieter...then I guess at the price, it's a better option for me right now, as I can't afford another $1100 microphone at the moment, after having to buy an oscilloscope and some other electronics stuff because this is a hobby that is quickly turning into an obsession.

As you say, there must be something wrong.

Here's a short stereo track with a simultaneous speech recording of an OPA1641 preamp (left channel) and a Rode NT1 (right channel)....

https://tinyurl.com/OPIC-left-NT1-rightt
No processing, except for an 80Hz Hi-pass to remove the worst of the LF 'room rumble' noise.

Not a very quiet background from a 'proper' test perspective, but quiet enough to illustrate that the noise level of the OPA preamp is only about 2 or 3dB worse than the Rode.
So as I mentioned earlier, about the same noise floor as the TLM103.
The OPA preamp is being fed from one of Ari's Flat K.47 caspules, but only polarised with about 58 volts. It's slightly less sensitive than the Rode at the moment......
 
Last edited:
As you say, there must be something wrong.

Here's a short stereo track with a simultaneous speech recording of an OPA1641 preamp (left channel) and a Rode NT1 (right channel)....

https://tinyurl.com/OPIC-left-NT1-rightt
No processing, except for an 80Hz Hi-pass to remove the worst of the LF 'room rumble' noise.

Not a very quiet background from a 'proper' test perspective, but quiet enough to illustrate that the noise level of the OPA preamp is only about 2 or 3dB worse than the Rode.
So as I mentioned earlier, about the same noise floor as the TLM103.
The OPA preamp is being fed from one of Ari's Flat K.47 caspules, but only polarised with about 58 volts. It's slightly less sensitive than the Rode at the moment......
Way quieter than what mine sounds like. I will post a sample of what mine sounds like tonight.
 
As you say, there must be something wrong.

Here's a short stereo track with a simultaneous speech recording of an OPA1641 preamp (left channel) and a Rode NT1 (right channel)....

https://tinyurl.com/OPIC-left-NT1-rightt
No processing, except for an 80Hz Hi-pass to remove the worst of the LF 'room rumble' noise.

Not a very quiet background from a 'proper' test perspective, but quiet enough to illustrate that the noise level of the OPA preamp is only about 2 or 3dB worse than the Rode.
So as I mentioned earlier, about the same noise floor as the TLM103.
The OPA preamp is being fed from one of Ari's Flat K.47 caspules, but only polarised with about 58 volts. It's slightly less sensitive than the Rode at the moment......
Sorry for the lateness, been working on re-taping the drywall in my garage, but here is a short sample of the mic.

Just incase the attachment here doesnt work, here is a link to the audio
 

Attachments

  • mic test.wav
    3 MB
That does seem to have a measurably higher noise level than I have experienced using the OPA 164* series op amps?
The problem with a direct comparison is that there are several variables.... capsule sensitivity, preamp gain, ambient noise levels etc...

I offer the following sample as a refernece which endeavours to exclude some of those variables.

This audio sample : OPIC.noise.level.wav is a recording of an OPIC preamp, made using this circuit: OPIC LDC

• The first 10 seconds are of the unterminated input of the recording device (a Tascam DR100Mk3) wih the gain set to 0dB. The noise level recorded is around -90dB.
• A small 'glitch' sound indicates the connection of the OPIC circuit. There is no capsule connected, although the 1G input resistor remains in circuit.
• The remaining part of the sample indicates the extra noise added by the OPIC ciruitry (between 3 and 6dB of noise).
• There is no HPF introduced -- the recorded noise is full bandwidth.

Looking at the Alice OPA schematic on page 2 of this document: https://www.jlielectronics.com/content/OPABoardsREV_1.pdf would suggest 2 possible sources of extra noise:

• The 2 x 2k2 resistors shown controlling the second (differential) audio output op amp will allow the audio output to be increased by 6dB, but will also introduce extra noise into the signal path. This should only be in the region of some 3dB - possibly slightly less.

• The Alice circuit shows the power to the opamp regulated by a zener diode. Zeners are notoriously noisy, although one would expect the LPF introduced by the 200R resistor and the 47uF capacitor to control that noise.
Some designers will include a second 100nF capacitor across the 47uF electrolytic, to help with any unattenuated HF noise . Others say this is unnecessary with modern capacitors? ... One of those things that can be confirmed experimentally quite easily! :)

( The OPIC circuit does not regulate the supply to the opamp. This helps maximise amplifer headroom. A zener is only fitted to regulate the supply to the voltage multiplier, used for capsule polarisation volts.)

Testing your ciruit like this should help to confirm the noise level of the OPA preamp itself, independent of anything introduced by the capsule - or the associated polarisation circuitry.
 
That does seem to have a measurably higher noise level than I have experienced using the OPA 164* series op amps?
The problem with a direct comparison is that there are several variables.... capsule sensitivity, preamp gain, ambient noise levels etc...

I offer the following sample as a refernece which endeavours to exclude some of those variables.

This audio sample : OPIC.noise.level.wav is a recording of an OPIC preamp, made using this circuit: OPIC LDC

• The first 10 seconds are of the unterminated input of the recording device (a Tascam DR100Mk3) wih the gain set to 0dB. The noise level recorded is around -90dB.
• A small 'glitch' sound indicates the connection of the OPIC circuit. There is no capsule connected, although the 1G input resistor remains in circuit.
• The remaining part of the sample indicates the extra noise added by the OPIC ciruitry (between 3 and 6dB of noise).
• There is no HPF introduced -- the recorded noise is full bandwidth.

Looking at the Alice OPA schematic on page 2 of this document: https://www.jlielectronics.com/content/OPABoardsREV_1.pdf would suggest 2 possible sources of extra noise:

• The 2 x 2k2 resistors shown controlling the second (differential) audio output op amp will allow the audio output to be increased by 6dB, but will also introduce extra noise into the signal path. This should only be in the region of some 3dB - possibly slightly less.

• The Alice circuit shows the power to the opamp regulated by a zener diode. Zeners are notoriously noisy, although one would expect the LPF introduced by the 200R resistor and the 47uF capacitor to control that noise.
Some designers will include a second 100nF capacitor across the 47uF electrolytic, to help with any unattenuated HF noise . Others say this is unnecessary with modern capacitors? ... One of those things that can be confirmed experimentally quite easily! :)

( The OPIC circuit does not regulate the supply to the opamp. This helps maximise amplifer headroom. A zener is only fitted to regulate the supply to the voltage multiplier, used for capsule polarisation volts.)

Testing your ciruit like this should help to confirm the noise level of the OPA preamp itself, independent of anything introduced by the capsule - or the associated polarisation circuitry.

So basically, my testing should be as follows:

1. Desolder capsule and uninstall capsule from the mic.

2. Record 10 seconds unterminated Apollo X6 with gain at 0db.

3. While still recording, connect the capsule-less mic into Apollo X6.

4. Let it continue recording for another 10 seconds to record noise level of circuit.

Is that correct?


And I shall post my results here.
 
So basically, my testing should be as follows:

1. Desolder capsule and uninstall capsule from the mic.

2. Record 10 seconds unterminated Apollo X6 with gain at 0db.

3. While still recording, connect the capsule-less mic into Apollo X6.

4. Let it continue recording for another 10 seconds to record noise level of circuit.

Is that correct?


And I shall post my results here.
Yes - although once the capusle is diconnected, there should be no need to actually remove it...... All I did when recording the sample above was to disconnect the capsule connection to the preamp. I even left the polarisaion volts connected to the capsule...

You will of course need to close the mic casing -- the open input is very high impedance, and highly susecptible to picking up external noise....
 
Yes - although once the capusle is diconnected, there should be no need to actually remove it...... All I did when recording the sample above was to disconnect the capsule connection to the preamp. I even left the polarisaion volts connected to the capsule...

You will of course need to close the mic casing -- the open input is very high impedance, and highly susecptible to picking up external noise....

I just don't want the wires to dangle around in there. I don't mind going the extra step to remove the capsule completely. I feel it's safer, im a noob lol, and im scared of shorting something. Even with tape on the end, I just can't get the thought of a loose wire dangling in there out of my head. OCD maybe?

I also suspect it's my mic preamp on the Apollo X6. I disconnected everything from the Mic input, and turned the gain up without 48v Phantom, and the hiss is still there...even without a mic or xlr cable connected to it.
 
As I mentioned earlier, it's best to try and remove the variables to get an idea of how much noise is beng introduced by the mic circuitry.
You will hear hiss from any preamp, if the gain is turned up...
I have no idea how good or bad your Apollo preamp actually is? - I believe it's a high quality device, so I would expect it to have a decent dynamic range and a EIN figure in excess of 120dB.... As my Tascam preamp does.
What we are trying to establish at the moment is just how much noise the mic circuitry itself is adding to the overall signal path. So it's really question of how much worse the noise figure gets, when the mic circuitry is added.
Not a measure of an overall hiss level.
In my test, the Tascam mic preamp gain was set to minimum, to get a low noise reference point to work from...
 
As I mentioned earlier, it's best to try and remove the variables to get an idea of how much noise is beng introduced by the mic circuitry.
You will hear hiss from any preamp, if the gain is turned up...
I have no idea how good or bad your Apollo preamp actually is? - I believe it's a high quality device, so I would expect it to have a decent dynamic range and a EIN figure in excess of 120dB.... As my Tascam preamp does.
What we are trying to establish at the moment is just how much noise the mic circuitry itself is adding to the overall signal path. So it's really question of how much worse the noise figure gets, when the mic circuitry is added.
Not a measure of an overall hiss level.
In my test, the Tascam mic preamp gain was set to minimum, to get a low noise reference point to work from...
Ah ok, that makes sense.

In your vocal sample, how high did you set your mic preamp?
 
AFAIK, mic preamp noise specs involve a 150 ohm termination between pins 2-3.
Certainly the preamp noise specs can normally only be accurately determined with an appropriate dummy load (normally 150R as you say).

In this case I was trying to determine the amount of noise being added by the mic preamp circuitry. On the Tascam - as I indicated in my earlier post - I applied 0dB of gain. This is achieved on that device with the application of the mic pad switch. With that switch 'on' there is no measurable noise difference between a loaded and unload input.
So I simply plugged and unplugged the mic, to determine how much extra noise the circuitry added. The phantom power was left on.

Wihout using the pad, there may well be a difference in the measured noise? .... I should have specified the requirement for a 150R load resistor to be added for that condition. My bad. Apologies.
 
Back
Top