TOPIC:

@ Michele,

Thanks for the very good update! Unfortunately, it seems like the picture are not showing for some reason. We have a little bug on this forum under firefox it seems to post pics. We're doing a website upgrade these days so it will be fixed in new website. In the mean time, you can post the pics using Internet Explorer.

As for your comments, we agree that indeed adding a capacitor is indeed the easiest way to go at it.

DevTeam

I did various test on the Umik-1 to improve the noise floor.

First of all, I used a netbook (asus 1215B e-350) to do the test and my mic serial is 0514.

Second thing, after some while, I noticed that the mic is noisy itself but the supply provided from the USB port can really do some mess and increase the performance of the mic. WIth my netbook, I found that with or without the external power adaptor the things changed a lot.
So, I repeated some test, but not all because the cable that connects the mic to the circuit were damaged after many tests and I even break a part of the PCB for my fault.

Third, I think the best thing to do is to put a capacitor where George suggested. It's very easy to do and with an electrolytic 470uF 6,3V almost all the noise is suppressed. It's a simple and powerful mod. On my netbook, the noise is very low even with the power adaptor. I putted an extra capacitor (100uF 10V) after the ferrite bead used to filter the usb supply but it's not so useful I think. Anyway it should not make the things worse.

Fourth thing: I build up a low noise reg with a LT1763 3,3V to feed biasing circuit of the capsule (look on the web for miniVreg) but the results are worse than the simple capacitor. Really. For sure it's not a circuit designed for this particular chase but if you want to use a low noise reg it's up to you. It could be very useful with high internal gain (24/30/36dB) but only if projected ad hoc.

Last thing: I sat the internal gain at 12dB. For sure the noise floor increase but for me this is a good set-up so I can use the mic not only for loudspeaker measurements. You have to modify the value in the calibration files. For me I changed from "Sens Factor =-22.68dB, SERNO: 7000514" to "Sens Factor =-10.68dB, SERNO: 7000514" (12db gain).

Now the results and some photos. Let me know.

Original mic without any mod, no power supply:


Original mic without any mod, with power supply:


Mic with 220uF cap, no supply::


Mic with 220uF cap, with supply:


Mic with LT1763 3,3V, no power supply:


Mic with LT1763 3,3V, with power supply:


Mic with 470uF on mic power line + 100uF on 5V supply, no supply:


Mic with 470uF on mic power line + 100uF on 5V supply, with supply:


Mic enclosed (470uF on mic power line + 100uF on 5V supply), no supply:


Mic enclosed (470uF on mic power line + 100uF on 5V supply), with supply:


Photos of the mod:

LT1763 regulator (the two resistor inparallel are used to obtain 1950Ohm)




Two capactor mod (470uF 6,3V and 100uF 10V suggested)


the smaller capactor is soldered on the pins of the input 3,3V regulator


the bigger capacitor is applied on 3,3V supply of the mic capsule:



Hope it helps.....


Michele

Sorry Devteam, I'm working to made the pics visible... can you cancel this and the previous one post? Keep the first with the pictures not visible, I need the url of the pics....

miki_ada wrote:

Ok guys, here I am.
I did various test on the Umik-1 to improve the noise floor.

First of all, I used a netbook (asus 1215B e-350) to do the test and my mic serial is 0514.

Second thing, after some while, I noticed that the mic is noisy itself but the supply provided from the USB port can really do some mess and increase the performance of the mic. WIth my netbook, I found that with or without the external power adaptor the things changed a lot.
So, I repeated some test, but not all because the cable that connects the mic to the circuit were damaged after many tests and I even break a part of the PCB for my fault.

Third, I think the best thing to do is to put a capacitor where George suggested. It's very easy to do and with an electrolytic 470uF 6,3V almost all the noise is suppressed. It's a simple and powerful mod. On my netbook, the noise is very low even with the power adaptor. I putted an extra capacitor (100uF 10V) after the ferrite bead used to filter the usb supply but it's not so useful I think. Anyway it should not make the things worse.

Fourth thing: I build up a low noise reg with a LT1763 3,3V to feed biasing circuit of the capsule (look on the web for miniVreg) but the results are worse than the simple capacitor. Really. For sure it's not a circuit designed for this particular chase but if you want to use a low noise reg it's up to you. It could be very useful with high internal gain (24/30/36dB) but only if projected ad hoc.

Last thing: I sat the internal gain at 12dB. For sure the noise floor increase but for me this is a good set-up so I can use the mic not only for loudspeaker measurements. You have to modify the value in the calibration files. For me I changed from "Sens Factor =-22.68dB, SERNO: 7000514" to "Sens Factor =-10.68dB, SERNO: 7000514" (12db gain).

Now the results and some photos. Let me know.

Original mic without any mod, no power supply:

Original mic without any mod, with power supply:

Mic with 220uF cap, no supply::

Mic with 220uF cap, with supply:

Mic with LT1763 3,3V, no power supply:

Mic with LT1763 3,3V, with power supply:

Mic with 470uF on mic power line + 100uF on 5V supply, no supply:

Mic with 470uF on mic power line + 100uF on 5V supply, with supply:

Mic enclosed (470uF on mic power line + 100uF on 5V supply), no supply:

Mic enclosed (470uF on mic power line + 100uF on 5V supply), with supply:

Photos of the mod:

LT1763 regulator (the two resistor inparallel are used to obtain 1950Ohm)

Two capactor mod (470uF 6,3V and 100uF 10V suggested)

the smaller capactor is soldered on the pins of the input 3,3V regulator

the bigger capacitor is applied on 3,3V supply of the mic capsule:

Hope it helps.....

There is still space for improvement in 2022. Here are two files with and without USB power filters. The difference is audible.
I have no idea how noisy USB power should be, but in my case, noise from USB is enough to impact the signal path.
Files were recorded on the max gain setting of the win sound mixer.