News:

"The phone is a remarkably complex, simple device,
and very rarely ever needs repairs, once you fix them." - Dan/Panther

Main Menu

Replacement part for old carbon mic

Started by Vesa, December 02, 2024, 03:55:20 PM

Previous topic - Next topic

Vesa

This is multipart post of replacement mic desing and building. Whole ide is to do condenser microphone based sparepart for old phones. First step is to create case for new mic electronics. I did 3d design with Fusion 365. For metal contact on bottom of the case is 2.9mm hole, normal 3mm bolt will screw directly. For other contact I use 0.2mm thick steel plate with soldering tab. Soldering tab has spot welded to metal ring. Metal ring fit to bottom part of the mic casing and can be glued there with super glue. A few images of te process below.

3D stl files are attached. Just rename .txt -> .stl

For microphone I try to use this from Digikey. This mic has integrated fet-amplifier. Now just waiting mic delivery. I will post new message about circuit when I have ready solution. I have two types phones: with audio transformer and without, the plan is to create universal circuit suitable for both models.

Vesa

#1
Part 2: mic ring. Mic ring outline is attached mic_ring.jpg. Measurements are in millimeters. Print it in 1:1 scale and use 2-side tape for metal plate. Then just cut plate in right size and bend slightly to right form.

countryman

Very nice, I'm excited to see the finished product!

Vesa

#3
Part 3: Equivalent circuit for carbon mic. First step is to do some measurement with carbon mics. I have couple different mics: one is from S&H Model 36, second from LM Ericson Dialog. First one has date stamb from 1962, so mic is replaced on the phone. Second one is probably original mig, date stamp points to 1970-decade. Both are Siemens mics. On measurement I use 500ohm series resistor, analog lab powersupply and osciloscope. Fully "calibrate" soud soure is me, just wording AAA! to mic. 8)

Result in the carbonMic.jpg, osciloscope measure over series resistor. 1v peak-to-peak voltage translates to 2mA AC-current. Both mics works quite similar way. We can estimate that this is "typical" carbon mic functionality level.

Mic what i'm plannign to use is this from Digikey. By the datasheet it gives around 1v amplitude with 94db(A) level. Human voice is around 80db near mouth so safe estimate is 0.1v amplitude. To proper voltage to current conversion needs separate amplifier, for that transistor is ok solution. With some fidling with values I got schematics with LTSpice simulator. Schemtics and simulation result as AC current on simu.jpg. So this equivalent circuit gives 2-3 times more current and creates 10mA dummy load to phone line loop. Values are in right ballpark.

Real circuit needs some more components, first one is diode rectifier for output. This solves polarity issue. Second one is mic powerin circuit. For that I will use resistor divider with zener diode clamping. Target for mic voltage is around 3v, so line voltage can drop near 5v volta and circuit works still with same audio volume level. Simulation gives same 5mA peak-to-peak values 5-20v range. Transistor drives current and that is good in this case.

Vesa

#4
Part 4: building the new mic. On the circuit.jpg is "final" mic circuit. D5 2.4v zener, Q1 BC846B, D1-D4 100v schottky. V1 is mic capsule and yellow box is external test circuit. Measurement result do not look promising, voltage swing in picture New_mic.jpg is lover than carbon mic. Real life test with S&H Model 36 gives opposite result: sound is laud and clear. It seems that carbon mic distordion is dominant on osciloscope measurement. Electric capsule mic gives only undistordet sound and this is the actual voice. Sound quality is superior over carbon mic.

Mic circuit with 3d printed parts on the picture mic_circuit.jpg. Assembly on handset is easy, just screw cap open and changes mic capsule to this the new one.

R5 is now tuned for low voltage use, Model 36 use transformer and mic voltage is only 6v. Some phones gives line voltage directly to mic and 5k is better for that use case.

countryman

Wow, that came out great! Very good job with those tiny SMD components, too.

I made a few boards with normal wired parts and fitted them into emptied carbon capsules. I followed the schematic shown below. The transistor is a Darlington type, to which the microphone is directly coupled. This way the mic only gets 1.2 ... 1.4 Volts which is low. Your schematic allows enough voltage to properly operate any electret mic.

Vesa

I spend some time for circuit testing and measurements and I was not happy with performance. Electrec mic voltage generation did not work correctly and NPN dc-current draw was just too high. So here is now fine tuned circuit on circuit.jpg. This circuit generates full 2.4v to mic even with 5v test voltage. Below is measurements with differenet voltages, voltage refers to circuit V2 test voltage source. With fine tuning this outperforms now carbon mic on all conditions. With 3v mic circuit receives around 1v and it stil produces good signal. Most of cases this fine tuned circuit gives 2x signal level when compared to carbon mic.

I will test this also using dual BC846B as darlington. When I have circuit ready I will do also comaprison on phones.

Vesa

One post more, now with darlington circuit_d.jpg. Test result on Darlington.jpg. More gain and performance is slightly better on low voltaga use. This version stays on use with Model 36.

countryman

Great research work, thank you for publishing the results in detail!
What do you think is the lowest voltage for successful operation? Might it also be used for local battery phones with 3 Volt supply?

Vesa

Yes, it should work with 3v local battery feed also. With the local battery polarity of the voltage is known, so there is no need for the bridge rectifier. Below is the circuit with polarity markings and you can leave diodes D1-D4 away. This will give 0.6v more voltage to the circuit, schottky diodes forward voltage is 0.3v. D5 can be leave away also, there is no risk for high voltages to the electrete mic.