The AM Forum
March 28, 2024, 02:14:07 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: voice band filters, cathode followers in and out  (Read 11205 times)
0 Members and 1 Guest are viewing this topic.
Opcom
Patrick J. / KD5OEI
Contributing
Member
*
Offline Offline

Posts: 8308



WWW
« on: October 24, 2014, 10:15:34 PM »

Attached is a starting schematic. Triodes are simple, but I read somewhere Pentodes or beam tubes make better cathode followers.

Has anyone some cathode follower circuits for pentodes? I only need to pass a few volts at most. Power is not needed, but low impedance, insensitivity to load, and good linearity are wanted.

The reason for this is that I have several of those little can-type audio filters of various different impedance from 600 Ohm to 10K Ohm and I want to just use the same basic circuit for all of them. Several channels for the voice signal. I'll deal with differing gains or levels elsewhere.

The goal is to have a switch to select a 200Hz-4.5KHz filter, a 300Hz-2.5KHz one, a straight though, etc. to have options depending on band conditions.

The triodes might be things like 12AU7's, and the Pentodes I had in mind would be 6AQ5's which I have plenty of.

Suggestions from anyone who had tried a pentode cathode follower or knows if there are any suprises or secrets to it, would be a help.


* 1.gif (25.64 KB, 1365x655 - viewed 433 times.)
Logged

Radio Candelstein - Flagship Station of the NRK Radio Network.
AB2EZ
Member

Offline Offline

Posts: 1722


"Season's Greetings" looks okay to me...


« Reply #1 on: October 25, 2014, 11:21:26 AM »

Opcom

For this purpose, where you are dealing with signals whose peak amplitude is only a few volts, why not use a pair of solid state, single chip audio amps... which is what I have done in the recent past in a similar application.

You can use the RoHS version of the popular LM386. It will work with a single 4V-12V DC supply. These devices are available in 8-pin dip packages for about $1.00 each.

The LM386 has an output impedance of less than 4 ohms; and can be configured to pass very low audio frequencies, while being AC coupled (to block the DC value of the output... which is 1/2 of the single supply voltage), using a high value, 25V electrolytic capacitor in series with the output. [Refer to the schematics in the specification sheet]. You can, as your original post suggests, add a series resistor at the output, to match the input impedance of the filter.

The AC coupled range (with low distortion) of output voltages is  approximately +/-  0.5 x (the supply voltage -3V). If the single supply voltage is 12V, this corresponds to a low distortion output voltage range of +/- 4.5V.

The impedance, between each input and ground, is 50,000 ohms. As you suggested in your initial post, if you want to match the output impedance of the filter to the input impedance of the amplifier, place an appropriate value resistor from the selected input to ground.

The voltage gain of these amplifiers is 20 (26dB); and you can reduce that by using a voltage divider at the input, while still matching the desired input impedance with a resistor to ground, ahead of the voltage divider.

Stu
Logged

Stewart ("Stu") Personick. Pictured: (from The New Yorker) "Season's Greetings" looks OK to me. Let's run it by the legal department
DMOD
AC0OB - A Place where Thermionic Emitters Rule!
Contributing
Member
*
Offline Offline

Posts: 1768


« Reply #2 on: October 25, 2014, 01:55:58 PM »

If you want to stay with Hollow-state, you can use this Pentode circuit or a variation of it.

Adjust R5 for your specific B+.

You might want to compare the specs and curves between the 6AQ5 and the 6AU6 and tweek the circuit.

Phil


* 6AU6 VF.pdf (17.71 KB - downloaded 281 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
WBear2GCR
Member

Offline Offline

Posts: 4135


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


WWW
« Reply #3 on: October 26, 2014, 02:05:03 PM »

A few thoughts.

Afaik triodes make better followers.
But those pentodes will work fine.
They're followers.

I'd use a high Z (or another suitable value)  before the input cap notwitstanding the grid resistor.

Probably add another section, "flat" - through the followers.
Maybe also "bypass" - straight wire (maybe with an attenuator - see next comment)

Most filters of the type ur talking have insertion loss.
Probably the loss is not enough for you to worry about, but in the event it's 10dB or more, you'd want to "make up" the gain either at the input or output depending on what the max level is for the filter units and also what the necessary output level is WRT noise floor, etc.

Also, most of these prefab filters want to be loaded with the proper Z in and out in order to get the spec'd response. So just a cap in, won't work right, the cap will look like it is part of the filter. IF it needs a cap input.

Would be nice to find the original specs. What manufacturer?
What slopes are the skirts?

                        _-_-

Got pix? Cheesy
Logged

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

Posts: 1768


« Reply #4 on: October 26, 2014, 07:37:29 PM »

It would certainly be a good idea to sweep the filter and to determine it's input and output impedance for sure.

Most of those older audio filters had 600 ohm input and output impedances.

The above Pentode VF circuit has an output impedance of 365 ohms, which should drive any "Biblical" value 600 load. The input impedance is > 470k which will not load any 600 ohm filter.

Stage gain is approx. 0.75.

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
Opcom
Patrick J. / KD5OEI
Contributing
Member
*
Offline Offline

Posts: 8308



WWW
« Reply #5 on: October 27, 2014, 07:09:02 PM »

Stu, Phil, and Bear,

Thanks a lot on those; One thing I thought about cathode follower outputs was that it was needed to have a lot of current, 5-10mA, to make a low-Z cathode follower output. Maybe not. I picked the 6AQ5 for its higher current, but it is overkill. The 6AU6 may well work OK at 365 Ohms.  I will terminate the filters as they were designed to be. There is also a box of 100 6GH8A's here, so maybe some gain could be made up. 6AU6's I don't have so many of. I would like to use what is on hand.

The lowest Z is 600, the highest is 10K. The brands are mixed, the filters have come from junk piles, military stuff, almost anything. They are in a bin in the lab, each is tagged with the data points of dB loss when properly terminated. I have swept them needless to say, and I think that being able to choose one or another may give a flexible result. I am sure there will be a straight through selection, and among those filters is a 'space shuttle audio' filter, about a 500Hz HPF, that could go before any of these. It sounds like a telephone conversation but might really punch through the noise.

I want to use tubes on this project if possible. I know the op-amps would do fine and maybe be less trouble, but it is a personal preference thing.

About the 6AU6, why is a 47K resistor needed in the plate lead? Doesn't that reduce output or gain?
Logged

Radio Candelstein - Flagship Station of the NRK Radio Network.
DMOD
AC0OB - A Place where Thermionic Emitters Rule!
Contributing
Member
*
Offline Offline

Posts: 1768


« Reply #6 on: October 27, 2014, 08:30:25 PM »


About the 6AU6, why is a 47K resistor needed in the plate lead? Doesn't that reduce output or gain?

The R5 resistor is simply used to the drop B+ to a 150V plate voltage operating point. You'll notice it has a honkin' Electrolytic to stabilize plate voltage. It's manly there to decouple the B+ line and provde extra filtering.

The resistance value R5 = (B+ - Desired Plate Voltage)/0.0024 = (100)/0.0024 = 41.7k, so a 39k would work better than a 47k and give a slightly higher plate voltage. 2.4 mA is total tube current: 1.7 MA for the plate and 0.7 mA for the screen. Pr = 0.225 W so a 1/2 Watt resistor should suffice.

For a B+ of 310V, R5 = 66.6k or 68k to the nearest standard value.

For a B+ of 350V, R5 = 83.3k or 82k to the nearest standard value.





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

Posts: 1768


« Reply #7 on: October 27, 2014, 11:24:22 PM »

Here is a calculated (according to the RCA tube curves), non-tested stage using the 6GH8(A) .

The 6GH8 is a power hog and not the best toob for audio (voltage) ampilfiers since it was designed for the Horizontal section of TV's.

* 6GH8 Gain and Follower.pdf (26.91 KB - downloaded 156 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
DMOD
AC0OB - A Place where Thermionic Emitters Rule!
Contributing
Member
*
Offline Offline

Posts: 1768


« Reply #8 on: October 28, 2014, 12:48:13 PM »

Quote
Thanks a lot on those; One thing I thought about cathode follower outputs was that it was needed to have a lot of current, 5-10mA, to make a low-Z cathode follower output. Maybe not.

It depends on what voltage you place across the 600 ohm load.

The current through a 600 ohm resistance with a sine wave voltage of 2 volts rms across it is 3.3 mA.

The dissipated power is 7 mW for a 2 volt rms sine wave signal into a 600 ohm load,

I assume these filters are all passive types with mostly lumped L and C components?


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
W3NE
Contributing
Member
*
Offline Offline

Posts: 139


« Reply #9 on: October 28, 2014, 01:34:22 PM »

The primary purpose of a screen grid is to eliminate Miller Effect. Since the plate of a cathode follower is at a-c ground, there is no Miller Effect. Use of a tetrode or pentode as a cathode follower achieves nothing except increased circuit complexity and reduced reliability. There are plenty of low power triodes with sufficient plate current capability to drive an audio filter with 600-ohm input impedance. For example, even with a signal amplitude of 2 V. p-p, a quiescent cathode current of 5 ma. is more than enough to ensure linearity on negative-going signals. A 6C4, 6J5 or one section of a 12AU7 (or 6SN7 to keep it really vintage) would be adequate.

Bob - NE
Logged
Opcom
Patrick J. / KD5OEI
Contributing
Member
*
Offline Offline

Posts: 8308



WWW
« Reply #10 on: October 28, 2014, 07:10:33 PM »

Quote
Thanks a lot on those; One thing I thought about cathode follower outputs was that it was needed to have a lot of current, 5-10mA, to make a low-Z cathode follower output. Maybe not.

It depends on what voltage you place across the 600 ohm load.

The current through a 600 ohm resistance with a sine wave voltage of 2 volts rms across it is 3.3 mA.

The dissipated power is 7 mW for a 2 volt rms sine wave signal into a 600 ohm load,

I assume these filters are all passive types with mostly lumped L and C components?




The 6GH8 a power hog? I never thought of it that way. How so, compared to the 6EA8 or 6U8 used in audio, etc?
Yes, all passive RLC stuff.
Logged

Radio Candelstein - Flagship Station of the NRK Radio Network.
Opcom
Patrick J. / KD5OEI
Contributing
Member
*
Offline Offline

Posts: 8308



WWW
« Reply #11 on: October 28, 2014, 07:22:25 PM »

The primary purpose of a screen grid is to eliminate Miller Effect. Since the plate of a cathode follower is at a-c ground, there is no Miller Effect. Use of a tetrode or pentode as a cathode follower achieves nothing except increased circuit complexity and reduced reliability. There are plenty of low power triodes with sufficient plate current capability to drive an audio filter with 600-ohm input impedance. For example, even with a signal amplitude of 2 V. p-p, a quiescent cathode current of 5 ma. is more than enough to ensure linearity on negative-going signals. A 6C4, 6J5 or one section of a 12AU7 (or 6SN7 to keep it really vintage) would be adequate.

Bob - NE


That's very interesting about the screens. I'd thought more that the screen grids would help regulate the average plate/cathode current despite slight plate supply fluctuations, if supplied with a stiff screen voltage.
Logged

Radio Candelstein - Flagship Station of the NRK Radio Network.
DMOD
AC0OB - A Place where Thermionic Emitters Rule!
Contributing
Member
*
Offline Offline

Posts: 1768


« Reply #12 on: October 28, 2014, 09:56:23 PM »

Quote
Thanks a lot on those; One thing I thought about cathode follower outputs was that it was needed to have a lot of current, 5-10mA, to make a low-Z cathode follower output. Maybe not.

It depends on what voltage you place across the 600 ohm load.

The current through a 600 ohm resistance with a sine wave voltage of 2 volts rms across it is 3.3 mA.

The dissipated power is 7 mW for a 2 volt rms sine wave signal into a 600 ohm load,

I assume these filters are all passive types with mostly lumped L and C components?




The 6GH8 a power hog? I never thought of it that way. How so, compared to the 6EA8 or 6U8 used in audio, etc?
Yes, all passive RLC stuff.

Relatively speaking. As long as you have a LV supply that can supply the required sum of Ip's, then it is no big deal.

As far as the Miller Effect, I doubt you will have any problems with RC coupled class A amplifiers at audio frequencies.  Now when you get to RF frequencies = 455kc and above, the Miller Effect creates a roll off of gain as frequency increases, and circuit component values have to be modified.
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
W3NE
Contributing
Member
*
Offline Offline

Posts: 139


« Reply #13 on: October 29, 2014, 11:24:35 AM »

Quote
As far as the Miller Effect, I doubt you will have any problems with RC coupled class A amplifiers at audio frequencies.  Now when you get to RF frequencies = 455kc and above, the Miller Effect creates a roll off of gain as frequency increases, and circuit component values have to be modified

Previous comments about Miller Effect were directed exclusively to its significance in a cathode follower circuit -- regardless of signal frequency -- where the plate is at a-c ground. As stated, use of a tetrode/pentode, as had been proposed for use as a cathode follower,  results in unnecessary complexity and lower reliability without improved performance. Of course the secondary effect of plate-grid capacitance will generally have a negligible effect at audio frequencies in an a-c grounded cathode circuit, wideband feedback amplifiers being a possible exception.

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

Posts: 1768


« Reply #14 on: October 29, 2014, 09:52:14 PM »

Quote
As far as the Miller Effect, I doubt you will have any problems with RC coupled class A amplifiers at audio frequencies.  Now when you get to RF frequencies = 455kc and above, the Miller Effect creates a roll off of gain as frequency increases, and circuit component values have to be modified

Previous comments about Miller Effect were directed exclusively to its significance in a cathode follower circuit -- regardless of signal frequency -- where the plate is at a-c ground. As stated, use of a tetrode/pentode, as had been proposed for use as a cathode follower,  results in unnecessary complexity and lower reliability without improved performance. Of course the secondary effect of plate-grid capacitance will generally have a negligible effect at audio frequencies in an a-c grounded cathode circuit, wideband feedback amplifiers being a possible exception.

Bob - NE

I suppose Pat would appreciate any alternative circuit schematics you may have to offer.
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
Opcom
Patrick J. / KD5OEI
Contributing
Member
*
Offline Offline

Posts: 8308



WWW
« Reply #15 on: October 29, 2014, 10:23:25 PM »

Those are all great schematics and advice, and I would appreciate any others.

It may be as well to use the 6GH8 circuit as the input side with the pentode driving the triode so there is some adjustable gain from it to compensate for differences in filter loss, and a 12AU7 triode cathode follower to buffer the filter output. I had in mind maybe using it as a pentode rather than with the screen tied to plate.

A 12AU7 section with plate voltage 200V would conduct 5mA with a bias of 9V. That bias could be had from an 1800 Ohm cathode resistor.

I noticed that on the cathode follower of the 6GH8 and 6AU6, the cathode resistor is tapped and the grid returned to a point partway up the resistor.
Is this to set the bias properly while allowing the voltage to swing negative enough?

Is the triode part of the attachment acceptable since there is 9V there, or should the cathode resistor also be tapped?


* triodepentode3.png (142.02 KB, 1126x1076 - viewed 388 times.)
Logged

Radio Candelstein - Flagship Station of the NRK Radio Network.
DMOD
AC0OB - A Place where Thermionic Emitters Rule!
Contributing
Member
*
Offline Offline

Posts: 1768


« Reply #16 on: October 30, 2014, 01:49:46 AM »

Quote
Is this to set the bias properly while allowing the voltage to swing negative enough?

It was for biasing.

The set points were from the tube curves for 150V.

For 1.7mA Ip and 0.7 mA screen current the grid bias, which is always with respect to the cathode, had to be about 2.4 volts.

Look at the voltage at the top of R3.

Now Look at the cathode potential. What's the difference?

* 6AU6 Pentode Class A VF .pdf (27 KB - downloaded 153 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
DMOD
AC0OB - A Place where Thermionic Emitters Rule!
Contributing
Member
*
Offline Offline

Posts: 1768


« Reply #17 on: October 30, 2014, 02:00:32 AM »

Quote
Is the triode part of the attachment acceptable since there is 9V there, or should the cathode resistor also be tapped?

For the 12AU7 the 9 volt bias sets the plate current about right.

However, there is no need to use 10k grid resistors in your circuit. 100k is minimal.


For the last stage voltage follower in your circuit its input load impedance needs to be about 10X times the output impedance of the filter. Since this is a voltage transfer circuit and not a power transfer circuit, matching impedances here is not necessary.

R2 is not needed as well for the same reason.

Voltage followers have a gain of less than 1, so a gain block ahead of the filter may be needed depending on the filter loss.

Good luck.

Phil



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
WBear2GCR
Member

Offline Offline

Posts: 4135


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


WWW
« Reply #18 on: October 30, 2014, 05:49:46 PM »


Please correct me if I have this wrong, but it seems to me that I have read where a passive filter network wants to have a particular load Z on the output side for proper transient response and filter shape?

                  _-_-
Logged

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

Posts: 1768


« Reply #19 on: October 30, 2014, 11:04:56 PM »

Pat,

Here is another possible circuit but running at 250V B+ with a slightly higher plate current, below max power, and in the more linear areas, for 10k filter input and output impedance.

BEAR brings up a good point.

My info says it depends on the filter design. Some filters may have the internal 10k resistive load, some may not, which is why you need to put a signal generator on the input of the filter and load the filter output with a 10k resistor and without.

If the filter does not have an internal load termination, simply add a 10k resistor to the output pin of the filter.

You could take your audio generator terminated in its characteristic impedance, and in series with a 50k variable resistor and measure the resistance across the pot with and without a 10k load. The old, "half voltage" method. Smiley




* 12AU7A Triode Class A VF.pdf (24.83 KB - downloaded 161 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
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 0.103 seconds with 18 queries.