Opa 1642 Alice LF cut and pad?

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

tedsorvino

Well-known member
Joined
Aug 1, 2018
Messages
74
Location
Athens, Greece
Hi, I would like to build a couple of op amp based (OPA 1642) Alice mics.
One will be based on an electret capsule and the other on a dual condenser one.
The voltage on the condenser will be provided by a Hex inverter cmos circuit (maybe two fo those, one for +bias and the other for -bias fir figure 8 mode). Both mics powered by 48V phantom power.
Could somebody provide any help for adding a LF cut filter and a -10db pad on both mics if it's possible?
I attach the schematic I follow in order to be more specific.

Thanks in advance
 

Attachments

  • opa alice mic schm.png
    opa alice mic schm.png
    153 KB · Views: 68
There isn't any gain in the circuit, so I suppose the only way to add a 'pad' is a capacitor in parallel with the microphone capsule, or lowering the polarisation voltage on the capsule.
 
Not related to the question but the OPA1642 seems to be an unobtainable chip at the moment...

As you say, mainstream sources seem to have really extended lead times at the moment...
You can find them on Aliexpress, but there's always the risk of fakes from there of course?
I took a chance on THESE and they seem fine.
 
As you say, mainstream sources seem to have really extended lead times at the moment...
You can find them on Aliexpress, but there's always the risk of fakes from there of course?
I took a chance on THESE and they seem fine.
Might take that chance, too. I've got few OPA1678s, those are CMOS (vs. JFET) opamps but the specs look similar. The Cin should be 6pf differential (low enough), though OPA1642 is supposed to have an extremely stable input capacitance vs. common mode voltage which might be good in this application (maybe MOSFET's are not so sensitive regarding this).
 
Last edited:
To pad the OPA Alice, you would put a 150pf cap across the element, depending on the element used, for about 10dB attenuation, to as high as 220pf for a 34mm element. Voltage depends on if you are using the inverter power supply 100v or so, if using the 2555, 35v or so will work.
 
I just realized that you must use an electret capsule with this amp if you don't plan using a DC/DC converter as the quiescent current per opamp is 1.8mA which will cause an 11.6V voltage drop over the 6k8 resistors in the phantom supply. About 35V can though bias some condenser capsules but not all properly. You could though use a 5-pin XLR and use a second mic channel to supply the full 48V just for biasing the capsule (using some y-splitter cable adapter).

(Would though rather use a 3-pin female XLR, so that you cannot mix with 48V phantom system, in mic with an unbalanced output and 60V (some current) input and convert the unbalanced to balanced in the PSU.)
 
I just realized that you must use an electret capsule with this amp if you don't plan using a DC/DC converter as the quiescent current per opamp is 1.8mA which will cause an 11.6V voltage drop over the 6k8 resistors in the phantom supply. About 35V can though bias some condenser capsules but not all properly. You could though use a 5-pin XLR and use a second mic channel to supply the full 48V just for biasing the capsule (using some y-splitter cable adapter).

(Would though rather use a 3-pin female XLR, so that you cannot mix with 48V phantom system, in mic with an unbalanced output and 60V (some current) input and convert the unbalanced to balanced in the PSU.)
I found it easier to just create a simple DC/DC converter for non-electret LDC capsules..
The one I used for this OPA project: OPIC LDC only cost a couple of pounds.
It gives the option of selecting the required DC polarisation voltage easily, and as it only draws around 1mA of current, it allows for around 24v DC to supply the op-amp itself. That allows for quite a good amount of headroom...
 
Yeah, the DC/DC converter is the best solution though if you want to keep it simple and still use an opamp it's not what you want (actually I have some PCBs available for the converter but not the hex inverters yet). Well, using the OPA1641 vs. OPA1642 halves the current and will leave almost 42V for the bias (6800/2*1.8mA = 6.12V). Also there are opamps like MCP6286 which has only about 500uA quiescent current but Vdd-Vss max of only 7V. Noise and input impedances are similar but input capacitance is a little bit greater but you have to use a zener (not so with OPAs). There might exist other opamps, too, in that category (low power, low noise, high input impedance, low input capacitance). MAX9945 might be good though has noise level of 15 nV/sqrt(Hz), it's thogh less than TL071 (18).
 
Last edited:
Noob question here: does the green triangle in the above schematic refer to a different kind of ground?
It's virtual ground, half rail (ca. 6v) coming from the two 47K resistor voltage divider.

Opamps need positive and negative power supply, say +6v and -6v. If you don't have this like in phantom power case here, you can ground the - side (pin 4) of the supply and put the + side to 12v(pin 8). But then you have to create this virtual ground which lies in between to give the opamp impression of dual supply. In this case audio input is biased to 6V (opamp sees it as virtual 0V, pin3), +12v supply (opamp sees it as +6v pin 8) and -0v negative supply (opamp sees it as -6v pin 4). Hope that makes sense.
 
Last edited:
Ah, ok! Thanks very much for this explanation and the time taken, @kingkorg (cool handle, BTW).
So virtual ground; here’s the noob part: where does it physically go - to its own ground plane or to the regular ground?
thanks again!!
:)
Ok - reading up on virtual ground/earth. Thanks for the point in the right direction - I’m betting this is why my breadboarding isn’t working. I had them all connected to regular ground.
 
Ah, ok! Thanks very much for this explanation and the time taken, @kingkorg (cool handle, BTW).
So virtual ground; here’s the noob part: where does it physically go - to its own ground plane or to the regular ground?
thanks again!!
:)
1G resistor and pin 5 are connected to the third green triangle. No connection to ground. There are already PCBs for this project, possibly even perfboard variant, so you can use them for inspiration and layout.

https://www.instructables.com/OPA-Based-Alice-Microphones-a-Cardioid-and-a-Figur/
JLI Electronics
JLI Electronics › ...PDF
“OPA-Alice” microphone electronics

https://www.pcbway.com/project/shareproject/Dual_OPA__Alice__Microphone_Circuit.html
https://www.pcbway.com/project/shareproject/OPA_Alice_Tom_Microphone_eaebcafc.html
 
1G resistor and pin 5 are connected to the third green triangle. No connection to ground. There are already PCBs for this project, possibly even perfboard variant, so you can use them for inspiration and layout.

https://www.instructables.com/OPA-Based-Alice-Microphones-a-Cardioid-and-a-Figur/
JLI Electronics
JLI Electronics › ...PDF
“OPA-Alice” microphone electronics

https://www.pcbway.com/project/shareproject/Dual_OPA__Alice__Microphone_Circuit.html
https://www.pcbway.com/project/shareproject/OPA_Alice_Tom_Microphone_eaebcafc.html
Just curios, what do you think about his pcb design?
 
@Khron is an authority on this topic.
I also need a capsule for that same schematic. I heard that you are an expert in that theme/sector. So you do know any good capsules that have appropriate shipping's costs to Europe and are not that expensive? Max price that I can spend would be about:50 Euro
Thanks 👍
 

Latest posts

Back
Top