Battery power for USB interface

"
I've got the output from the present inverting chip going to a 10k resistor and a 10mfd capacitor. Do I really need these? "

Won't hurt but 10 k in series? That will restrict HF on a long cable. Classic output config' is a series R of 74 to 120 Ohms* a series capacitor, 10mfd is fine into typical 10kZ inputs and a 'tie down' resistor of about 100k to keep DC from bleeding onto the OP and causing a click on connection.

Connecting the mic directly into the summing point of a shunt feedback amp means it 'sees' approx. zero Ohms and thus the whole thing resolves to 'current' drive and the response will be all over the shop! Mics are designed to drive a virtual open circuit or at least 5 X posted OP Z.
I think I shall have to post a couple of circuit ideas? BWT I think you are in UK? If so PM me if you like.

*You should always have a such an impedance (can be a small choke) on a chip's output into the 'world' because a 'cappy' cable can cause instability. Chips INSIDE gear, not driving a capacitance can simply be DC or AC coupled.

Dave.
 
Thanks again - I am in the UK - have just sent you a personal message!

The mic does sound amazing with the direct connection - lively with plenty of top and bottom!
 
Can't see how to attach files to a PM mate!

Anyhoos, the hybrid schem' you can see is pretty complex but is the only way to get good noise figures, even the 4562 is not as good as the best discrete transistors. There are slightly simpler but better versions in Duggie's book (sorry about the wedge! Local library?)

My effort is a bit of a rip from the book but with RFI components added. Using a 5534/2 or an LM4562 you could scale the feedback resistors down by a factor of ten and get a noise reduction of root ten =10dB. The circuit was originally designed for a 1:10 input transformer and has an overall max gain of 80dB! The series RC network across the traff secondary is to control out of band resonance and needs to be selected under test for a specific device but so long as the response is 1/2 dB down at 20kH ref 1kHz that should serve. I no longer have an audio genny but you can get 20kHz from most DAWs.

*Is to warn you that SOME OP AMPs, the 5534/4s e.g. will have ins and out pins NEGATIVE of the supply centre zero (assuming that IS bang on 1/2 Vcc*) Most chip are slightly +ve. I can find no data as to how the 4562 behaves. Modern electrolytics will not object to a few mV wrong polarity for the time it takes to check but left as such will probably have a short life.

For long term 'proper' builds I would use a full voltage supply then use a cheap chip, TL0, even a 741(!) to derive a low resistance 1/2 rail supply.

(found one traff, leads are bit tatty, sure I have another.)

Dave.
 

Attachments

  • hybrid mic pre.png
    hybrid mic pre.png
    50.7 KB · Views: 5
  • mic schen.pdf
    154.2 KB · Views: 1
Thanks Dave!

I have just sent you a pm with my email address. Wonder if you could mail me the circuits ? Have a few questions will post later if I'm?
 
Thanks again for the circuits Dave. I'm back home now and able to have a proper look. I couldn't resolve the detail on the hybrid circuit before but now it is really clear.
 
Back
Top