The AM Forum
March 28, 2024, 04:55:29 PM *
Welcome, Guest. Please login or register.

Login with username, password and session length
 
   Home   Help Calendar Links Staff List Gallery Login Register  
Pages: [1]   Go Down
  Print  
Author Topic: Homebrew microphone pre-amp  (Read 6670 times)
0 Members and 1 Guest are viewing this topic.
k7mdo
Contributing
Member
*
Offline Offline

Posts: 420


« on: September 27, 2018, 10:38:15 PM »

I just finished building a homebrew pre-amp for my Gates BC-1g that I am converting to 75 meter phone.

Unfortunately the amplifier seems to have a distortion in the output that I am afraid will eventually show up in the audio of the transmitter.

I developed the circuit from looking at about 20 different circuits in the late 1950's and -60's radio amateur handbooks,

It uses a 12AX7 as the tube and is a conglomeration of my junk box parts.

Overall it works!  But, when you look at the upper trace on the osc. you see a distortion that unfortunately gets worse as I turn up the input or the amplifier gain potentiometer.

Does this look serious?  Or, can one expect not to have linear amplification out of such a circuit, or, does anyone see a possible thing I could do to improve the circuit?

The plates are at about 200 VDC and the cathodes run about 9 VDC as it sits.

Thanks for any input!  73, Tom



* circuit.jpg (70.86 KB, 567x378 - viewed 319 times.)

* distortion.jpg (153.75 KB, 756x504 - viewed 298 times.)

* wiring.jpg (111.51 KB, 567x378 - viewed 254 times.)
Logged
PA0NVD
Member

Offline Offline

Posts: 615


Nico and Chappie (Chappie is the dog...)


« Reply #1 on: September 28, 2018, 12:26:40 AM »

Hi Tom
Seems to be over driven, I think that the second triode can't drive the transformer well.. Please give some info about the amplitudes of the signal, bu disconnect the transformer in order to measure the amplitude at the anode of the second triode. You can't expect much more undistorted signal that approx 100V t-t. (without e transformer )
You have a whole lot of gain there and I think a too-high impedance to drive the transformer..
Logged
KA2DZT
Member

Offline Offline

Posts: 2192


« Reply #2 on: September 28, 2018, 01:33:42 AM »

Try lowering the plate load resistor to 47K on the second stage and try raising the plate voltage to 250 volts
Logged
DMOD
AC0OB - A Place where Thermionic Emitters Rule!
Contributing
Member
*
Offline Offline

Posts: 1768


« Reply #3 on: September 28, 2018, 12:25:45 PM »

Is this for a communications application?

How much gain do you need?

Here is a general 12AX7A amp circuit that has high gain and reproduces waveforms very symmetrically.

As I see it, your cathode bias is way too high putting the stages into a very non-linear area of operation according to the tube curves.

12AX7 stages are not meant to drive low impedance inductive loads such as transformers.

I would recommend using this stage to drive a 12BY7 or some such if you are wanting to drive a line transformer.


Phil - AC0OB




* General 12AX7A Speech Amp and Driver.pdf (63.95 KB - downloaded 190 times.)
Logged

Charlie Eppes: Dad would be so happy if we married a doctor.
Don Eppes: Yeah, well, Dad would be happy if I married someone with a pulse.NUMB3RS   Smiley
WA2SQQ
Contributing
Member
*
Offline Offline

Posts: 1090


« Reply #4 on: September 28, 2018, 02:42:40 PM »

You might also want to use this circuit. This was a very popular "device" back in the late 60's for the high end CB'ers. I've used this with a D-104 and it delivers plenty of audio. As I recall it sounded pretty decent.

* Demco Modulator.pdf (120.19 KB - downloaded 232 times.)
Logged
k7mdo
Contributing
Member
*
Offline Offline

Posts: 420


« Reply #5 on: September 28, 2018, 05:52:06 PM »

I really appreciate a lot of the information you folks have given....  I tried raising the plate voltage on stage two of 12AX7 but no difference... yes, if I remove the transformer it seems to clean up the signal but I was pretty sure I needed the transformer to get my 600 ohm load fed....  


In fact, I think I solved the problem by doing something I had seen in the handbooks, I removed the cathode bypass capacitor from the second stage triode.  It immediately assumed the shape of the input signal... while also reducing substantially the gain of the circuit.  I see in the old handbooks where 12AX7's were run in modulator pre-amp circuits with no cathode bypass caps on either stage and with success....


For another try I replaced the -AX7 with an -AT7 and, yes, got lower gain and but no change in distortion until I removed the bypass cap.  I know that looking at an oscilloscope waveform is not a great way to "see" distortion but I can't see any more obvious glitches in the waveform....  hoping.

Now, the confession is that I am not sure what I need to feed the Inovonics 222 peak limiter that goes between the Gates and the pre-amp but 600 ohms seems to be a standard impedance.... in any event I know I need more audio from the old Shure 444D to get enough to the Gates and I will see if the "limiter" light goes on when I try it on the Inovonics.

I have the Gates up and making power by itself into its dummy load... now I will hook in the audio and see if I get modulation.

Thanks again, I looked at the other circuits but as you know, just getting all the parts and a chassis together is a lot of work and I am not ready to dump my single tube pre-amp until I just can't make it work or ??

 
Logged
DMOD
AC0OB - A Place where Thermionic Emitters Rule!
Contributing
Member
*
Offline Offline

Posts: 1768


« Reply #6 on: September 28, 2018, 06:14:33 PM »


...Thanks again, I looked at the other circuits but as you know, just getting all the parts and a chassis together is a lot of work and I am not ready to dump my single tube pre-amp until I just can't make it work or ??

 

I don't think you're going to get a decent symmetrical waveform unless you change the biasing. The bypass capacitor won't solve your problem, as it would only increase gain.

Linear class A circuits with a dual triode are highly dependent on the plate current and proper cathode biasing.

Look at page 3 of https://frank.pocnet.net/sheets/093/1/12AX7.pdf and you will see your bias voltage doesn't even show up in the curves.


Phil - AC0OB


Logged

Charlie Eppes: Dad would be so happy if we married a doctor.
Don Eppes: Yeah, well, Dad would be happy if I married someone with a pulse.NUMB3RS   Smiley
PA0NVD
Member

Offline Offline

Posts: 615


Nico and Chappie (Chappie is the dog...)


« Reply #7 on: September 28, 2018, 06:48:01 PM »

If you have sufficient gain, put the second triode as a cathode follower and drive the transformer from the cathode via 0.1 uF or more. Anode direct to the positive without a resistor.
Logged
DMOD
AC0OB - A Place where Thermionic Emitters Rule!
Contributing
Member
*
Offline Offline

Posts: 1768


« Reply #8 on: September 28, 2018, 08:49:42 PM »

I was thinking of a Cathode Follower as well and so here is a suggested circuit that uses two 12AX7A's and has an output impedance of ~ 100 ohms and has plenty of drive to an XLR output connector for an Inovonics 222 input.


Pheel

* General 12AX7A Speech Amp with CF.pdf (47.64 KB - downloaded 137 times.)
Logged

Charlie Eppes: Dad would be so happy if we married a doctor.
Don Eppes: Yeah, well, Dad would be happy if I married someone with a pulse.NUMB3RS   Smiley
k7mdo
Contributing
Member
*
Offline Offline

Posts: 420


« Reply #9 on: September 28, 2018, 11:18:39 PM »

It is starting to appear I should have asked before I designed and built this little project! 

I think I will try to convert the existing circuit to a cathode follower as suggested before I start adding tubes and drilling more holes!

This makes my attempt to avoid buying a Symetrix seem like a bad idea😕.

As before, thanks for the ideas.  73, Tom
Logged
PA0NVD
Member

Offline Offline

Posts: 615


Nico and Chappie (Chappie is the dog...)


« Reply #10 on: September 29, 2018, 09:47:31 AM »

If you don't want to add tubes, you can add a simple HV mosfet as a source follower to the anode of the second triode as well. Gate directly to the anode, drain to the +HV and source via e.g.  4k7 1Watt to ground and via a cap to the transformer. The Mosfet should be able to withstand 300VDC, but that is no problem, there are plenty and really cheap.
No holes and hardly any extra parts.
Logged
k7mdo
Contributing
Member
*
Offline Offline

Posts: 420


« Reply #11 on: September 29, 2018, 10:23:12 AM »

Have you got a typical part number for the mosfet?  I see there are quite a few choices...  I like the idea but would be on new ground.

Tom
Logged
DMOD
AC0OB - A Place where Thermionic Emitters Rule!
Contributing
Member
*
Offline Offline

Posts: 1768


« Reply #12 on: September 29, 2018, 02:05:21 PM »

Possibly this is what Nico was referring to ? and he can speak to it.

My recommendation for a Mosfet is this one that has been used in ClassE designs:

http://www.retelektronika.hu/DataSheets/25_FET-EK/FAI_116/FQA11N90_F109.pdf



Phil - AC0OB

* General 12AX7A Speech Amp Mosfet CF.pdf (44.8 KB - downloaded 120 times.)
Logged

Charlie Eppes: Dad would be so happy if we married a doctor.
Don Eppes: Yeah, well, Dad would be happy if I married someone with a pulse.NUMB3RS   Smiley
PA0NVD
Member

Offline Offline

Posts: 615


Nico and Chappie (Chappie is the dog...)


« Reply #13 on: September 29, 2018, 02:46:06 PM »

All relative small mosfets that can sustain the voltage
IRF710, IRF720 etc
It is non critical, smaller normally is better, less capacitance and no need for high currents
Logged
WBear2GCR
Member

Offline Offline

Posts: 4135


Brrrr- it's cold in the shack! Fire up the BIG RIG


WWW
« Reply #14 on: September 29, 2018, 04:02:55 PM »

The unbypassed cathode compared to the bypassed cathode has increased feedback, therefore less gain, but
more linearity. I think that is what you are seeing.

Adding a follower to drive a 1:1 (or similar ratio) of a low Z (~600ohm range) primary is a good idea since
it is a bad idea to drive a low Z load from a high Z source. I think this is what you are also seeing.

You could use something like a 10k:600ohm transformer, and possibly insert the iron in place of
the plate resistor, assuming the transformer is designed to take the voltage and the current (doesn't
saturate). But you could also run it "modified heising" which would be a choke for the plate resistor and
a cap feeding the transformer primary. But, you'd need ample gain since the 10k:600ohm ratio is a pretty
good step down, so the voltage will be reduced proportionally. But, how much voltage swing at the output
are you looking for, and what is the output going to drive?

And, yes, definitely get the bias correct.
You should set the bias based on the vdrop across the cathode or plate resistor, use Ohm's law to determine
the current, and set it per the mrf's spec sheet for operation at your plate voltage. The tube curves can also
show the proper operating point... it's not hyper-critical, but it ought to be close. Ultimately you may want to
fine tune it by watching the clipping on the positive going and negative going waveform - I use a triangle wave,
not a sine.

Fwiw, this is pretty much how circuit design goes... Cheesy
Logged

_-_- bear WB2GCR                   http://www.bearlabs.com
DMOD
AC0OB - A Place where Thermionic Emitters Rule!
Contributing
Member
*
Offline Offline

Posts: 1768


« Reply #15 on: September 30, 2018, 07:26:41 PM »

All relative small mosfets that can sustain the voltage
IRF710, IRF720 etc
It is non critical, smaller normally is better, less capacitance and no need for high currents

But we don't need 10 nanosecond switching rise times here.

We're passing analog audio with slow rise times.


Phil - AC0OB
Logged

Charlie Eppes: Dad would be so happy if we married a doctor.
Don Eppes: Yeah, well, Dad would be happy if I married someone with a pulse.NUMB3RS   Smiley
PA0NVD
Member

Offline Offline

Posts: 615


Nico and Chappie (Chappie is the dog...)


« Reply #16 on: September 30, 2018, 08:19:00 PM »

It is not a matter of risetime. The drive impedance is very high, so yo don't want a capacitive load. It is a source follower, so not too critical, but never the less. I should even use a smaller Mosfet, but they are hard to get and more expensive. An IRF710 is cheap and rugged.
Logged
DMOD
AC0OB - A Place where Thermionic Emitters Rule!
Contributing
Member
*
Offline Offline

Posts: 1768


« Reply #17 on: October 01, 2018, 02:11:45 AM »

The advantage of Source Followers or Cathode Followers is they have high input impedances and low output impedances.

The Common Drain or Source Follower MOSFET is a high frequency stage with a gain of < 1.0 and does not suffer any substantial frequency roll-off problems from Miller-Effect capacitances.

Any of the suggested MOSFETs with > = 400 volt Vds' will work.

If one is not familiar with SS components and their operation, then the original circuit with the two 12AX7A's will work well.


Phil - AC0OB


 

* General 12AX7A Speech Amp Mosfet CF.pdf (45.4 KB - downloaded 120 times.)
Logged

Charlie Eppes: Dad would be so happy if we married a doctor.
Don Eppes: Yeah, well, Dad would be happy if I married someone with a pulse.NUMB3RS   Smiley
k7mdo
Contributing
Member
*
Offline Offline

Posts: 420


« Reply #18 on: October 01, 2018, 10:19:54 AM »

Ok, I  can give this circuit a try with my existing layout.  I had the 600 ohm collins transformer and stored it for years hoping to put it to use...  This may not be the time.

Thanks again for the explanations and advice.

Tom
Logged
wa1mtz
Member

Offline Offline

Posts: 29


« Reply #19 on: October 03, 2018, 07:47:24 PM »

look at guitar amp circuits, lots online for free..fender champ, or Princeton reverb (fender) the handbooks usually go for restricted audio and max gain of stage. not what you need. The tx will want something like +10 db at 600 ohm balanced line. a 12au7, 6sn7 push-pull =250v will drive it. By the way, you certainly need a limiter compressor, that will have the right output, & Z      (AM-864/U)
Logged
Pages: [1]   Go Up
  Print  
 
Jump to:  

AMfone - Dedicated to Amplitude Modulation on the Amateur Radio Bands
 AMfone © 2001-2015
Powered by SMF 1.1.21 | SMF © 2015, Simple Machines
Page created in 0.062 seconds with 19 queries.