Yet another DIY Alice build

[zanzeoo]

Hello

I just build a circuit board based on DJJULES but with some modifications (all smd parts and double sided board with the voltage multiplier included )

I have good results with cheap chineses condenser capsule found on Ebay
I have good results with the JLI-2555BXZ3-GP

I tried the JLI-3412 but sound is very less sensitive than the 2555 and very bassy, do you think this is normal ?

Here is the project : GitHub - zanzeoo/head-amp: Microphone capsule amplifier

 

[rogs]

Interesting to see another take on Jules Alice OPA concept..... My version went slightly in the other direction...
Rather than all SMD and double sided, I went for a simpler single sided 'stripboard' version. ( Some notes here: www.opic.jp137.com )
I decided to only regulate the voltage multiplier DC supply, rather than the supply to the op-amp as well.... That seemed to be the best way of optimising dynamic range.
I have found the results very satisfactory ..... I think Jules original decision to use the OPA164* op-amp for this task was an inspired choice.... and the adaption of Rory Holmes 1983 ETI article on using CMOS gates for the multiplier seems to be gaining more and more favour.
(The lack of the need for inductors probably helps with that choice.....)

 

[zanzeoo]

Yes I saw your project, very intresting !
For the Figure 8 pattern is made by substracting (Side 2- Side 1 ) the sides ?
Omni is by adding the 2 sides ?

 

[rogs]

Yes I saw your project, very intresting !
For the Figure 8 pattern is made by substracting (Side 2- Side 1 ) the sides ?
Omni is by adding the 2 sides ?

Yes -- Addition and subtraction are simple tasks for op-amps, so I used that as an easy way to derive the different patterns.
Also removes any problems that can occur if you use a pattern switching options that change capsule polarisation voltages.

In my other versions I simply use the op-amp as a non inverting unity gain buffer, to minimise the number of noise generating resistors in the signal path.
You might also note that I use a single sided audio output - although the line itself remains impedance balanced - and that helps to keep the noise figure slightly lower.

One other thing that I discovered that can make a difference with this circuitry is the choice of type of 40106 used in the multiplier...
I specify a CD40106 in my notes, and these will give the DC output results I posted.....
If however I substitute an HEF40106 for the CD40106, the difference in Schmitt thresholds provided by that type will give quite different results.
Using a 12V supply, the HEF40106 oscillator frequency is more likely to be around 180KHz, rather than the 125KHz from the CD401016.....
Add to that any differences those thresholds may make to the multiplier sections, and you can end up with a significantly different final output voltage.

My simplest version - THIS ONE - doesn't even use a multiplier..... It derives the capsule polarisation voltage from one side of the 48v phantom power supply.
So the sensitivity is reduced a little - but only by around 2dB from that you will get from using a 60V polarisation voltage.
Also the unbalancing of the current drawn from the phantom power makes it unsuitable for use with transformer coupled mic preamps.
But it works very well for such a simple circuit...

As I say, I think Jules' idea to use this op-amp instead of a more convention FET 'front end' was an inspired one.... especially for first time hobby mic builders

 

[doqmemory]

@rogs, have you tried using "audio grade" capacitors for the output capacitors (C5/C6 in the MP version, C2/C3 in the electret version), such as the Nicholson Nichicon "Fine Gold" (Mouser 647-UFG1H470MPM) or Elna Silmic II? Do they need to be 47µF, or would 33µF work as well?

 

[rogs]

@rogs, have you tried using "audio grade" capacitors for the output capacitors (C5/C6 in the MP version, C2/C3 in the electret version), such as the Nicholson "Fine Gold" (Mouser 647-UFG1H470MPM) or Elna Silmic II? Do they need to be 47µF, or would 33µF work as well?

No, I've not ventured into the world of 'audio grade' capacitors... I'm not quite sure how they are supposed to improve things?
The OPA164* series op-amps are pretty good - noise and distortion wise - and the comparative measurements I've been able to make haven't suggested that the Panasonic M series output caps - which I generally use - are introducing anything horrible into the signal path that I can hear - or measure?....

33uF caps have a reactance of c. 240R at 20Hz, against c.170R for 47uF, so I don't think there would be a lot of difference, working into impedances of 1 or 2K.
Add to that the often quite wide tolerance of electrolytic caps, and I don't believe you'd notice any real difference...

 

[doqmemory]

No, I've not ventured into the world of 'audio grade' capacitors... I'm not quite sure how they are supposed to improve things?
The OPA164* series op-amps are pretty good - noise and distortion wise - and the comparative measurements I've been able to make haven't suggested that the Panasonic M series output caps - which I generally use - are introducing anything horrible into the signal path that I can hear - or measure?....

33uF caps have a reactance of c. 240R at 20Hz, against c.170R for 47uF, so I don't think there would be a lot of difference, working into impedances of 1 or 2K.
Add to that the often quite wide tolerance of electrolytic caps, and I don't believe you'd notice any real difference...

Thanks, that is what I suspected. I have a couple of the Nicholson Nichicon "Fine Gold" caps that I will try on one of Mic Sharf's boards when they arrive. I'll report the results with all other components (and capsule) the same.

 

[Khron]

I love that autocorrect, by the way "Nicholson"...

Results, you say? One single mic, different takes?

 

[doqmemory]

I love that autocorrect, by the way "Nicholson"...

Results, you say? One single mic, different takes?

Not that scientific -- it will be two different microphones with the same components, capsule, & body, except for the output caps. If there seems to be any real difference between them, then I will try to examine in more detail. The cost is just about $1 more per mic, anyway.

Funny, I didn't even see the autocorrect. It must have thought that I was referring to Jack.

 

[Khron]

Not that scientific -- it will be two different microphones with the same components, capsule, & body, except for the output caps. If there seems to be any real difference between them, then I will try to examine in more detail. The cost is just about $1 more per mic, anyway.

Quite the contrary, that IS indeed the more scientific way (recording the same take with the "reference" mic and the "device under test" simultaneously, to thus eliminate the variability of the take itself accounting for at least some of the perceived differences).

 

[rogs]

Forgive my ignorance -- what exactly are these 'audio grade' capacitors expected to do? --- reduce distortion? - improve frequency response?
My inclination is to call 'snake oil'...... but I'm willing to be proved wrong.....

 

[Khron]

At the very least, they're designed to more efficiently slim down your account, and add that all-essential feel-good-juju to the audio...

 

[doqmemory]

Forgive my ignorance -- what exactly are these 'audio grade' capacitors expected to do? --- reduce distortion? - improve frequency response?
My inclination is to call 'snake oil'...... but I'm willing to be proved wrong.....

I bought a few along time back, for tube cathode bypass, but never determined if they were any "better" for that than other types.

I will see how they sound as the output caps in this circuit.

 

[Nick Nack]

Forgive my ignorance -- what exactly are these 'audio grade' capacitors expected to do? --- reduce distortion? - improve frequency response?
My inclination is to call 'snake oil'...... but I'm willing to be proved wrong.....

They impart an improvement - night and day, sometimes - that's clearly audible to the hobbyists at diyaudio.com but can't be measured or heard by the EEs and audio professionals at groupdiy.com.

 

[zanzeoo]

 

[Khron]

How's the noise?

 

[zanzeoo]

I was my next question

What is the protocol to mesure the noise and performance ? , I have all kind of mesurement devices at my disposal

I was thinking to remove the capsule and put a pure sinus at the input but is it the right way ?

 

[Khron]

How will you determine the noise level, if you're already applying a signal to the input?

My question stemmed more from the "no attention given to the high-impedance area" thay i noticed...

 

[rogs]

Looking at your schematic here: head-amp/documents.PDF at main · zanzeoo/head-amp I see you have chosen to add a second opamp to provide a differntial audio output. This will add a little extra noise to the circuit (around 2dB), but you should still expect to measure a noise figure of around 10dB(A).

I see you have also elected to regulate the DC supply to the opamps at c.12V ? --
While it is a good idea to regulate the supply to the voltage mulitplier, the OPA164x series of opamps are capable of operating with a DC supply of up to 36V (see HERE ) so regulating the supply to the opamp is not really necessary with these devices. It will merely serve to limit the available headroom in this case.

The simplest way to get a reasonable estimate of the noise figure generated by the pre-amp circuitry would be to disconnect the capsule and connect the device to a low noise mic pre-amp - ideally set to minimum gain...
Make a test recording, and take a note of the recorded noise level.
Disconnect the device, and fit a 150R metal film resistor across the XLR input terminals ( see HERE for details of why).
With the same pre-amp gain setting, make a second recording, and note the noise level recorded.
The difference between those 2 noise levels will show just how much noise the mic circuitry is adding.
As I say, with the differential audio output you've chosen, you are likely to find around 9 or 10dB of noise has been added.
( See HERE for more information on mic noise levels).

 

[zanzeoo]

Looking at your schematic here: head-amp/documents.PDF at main · zanzeoo/head-amp I see you have chosen to add a second opamp to provide a differntial audio output. This will add a little extra noise to the circuit (around 2dB), but you should still expect to measure a noise figure of around 10dB(A).

I see you have also elected to regulate the DC supply to the opamps at c.12V ? --
While it is a good idea to regulate the supply to the voltage mulitplier, the OPA164x series of opamps are capable of operating with a DC supply of up to 36V (see HERE ) so regulating the supply to the opamp is not really necessary with these devices. It will merely serve to limit the available headroom in this case.

The simplest way to get a reasonable estimate of the noise figure generated by the pre-amp circuitry would be to disconnect the capsule and connect the device to a low noise mic pre-amp - ideally set to minimum gain...
Make a test recording, and take a note of the recorded noise level.
Disconnect the device, and fit a 150R metal film resistor across the XLR input terminals ( see HERE for details of why).
With the same pre-amp gain setting, make a second recording, and note the noise level recorded.
The difference between those 2 noise levels will show just how much noise the mic circuitry is adding.
As I say, with the differential audio output you've chosen, you are likely to find around 9 or 10dB of noise has been added.
( See HERE for more information on mic noise levels).

Wow !Thx you for that

Do you think a preamp circuit designed with fully differential amplifier would be better ( less noise ?)

https://www.ti.com/lit/pdf/sloa054