Quest for knowlege

[K5WLF]

Im sure that a careful use of a distortion meter while rebiasing can make any tube sound hard or soft.

In the ham world hard means splatter.

Just to clarify, Carl, all we did was a straight swap of tube brands of the same type with no rebiasing or other circuit changes of any kind.

[KB2WIG]

Weekend Update.

What were thought to be 6l6gc was actually 6l6GA. Oh well. Whatts an extra 120 volts on the plate gonna hurtz?

Pluging in the Ruskie tubes. Ive got much more swing on the modulator current. After asking for a few 'critical' audio reports ......  The audio sounds 'nice', the depth of modulation is still crappy. Time to switch out the D10-4 and see if a different mic helps things out....


klc

[AB2EZ]

Kevin

A lot of AMers use substantial amounts of compression (audio chain) to increase the average loudness of their signals. I use a moderate amount of compression; but even with that, I sometimes receive reports (from AMers using spectral displays or meters that provide peak and average audio power measurements) that my modulation depth is low. But, looking at my transmitter's modulated rf, on my digital oscilloscope... triggered by the input audio signal, I can clearly see that my modulation levels (positive and negative) are fine. Also, I can see this using my WA1QIX modulation monitor.

Unless you monitor your modulation with an rf sniffer and an oscilloscope, you won't get a good "picture" of what your transmitter is really putting out.

Also, using substantial amounts of compression significantly increases the distortion products ("splatter") in one's signal. People who use a lot of compression sound louder... but they have very wide signals.

Stu

[KB2WIG]

Stu,

Yes, the scope is the berries....  My bench set up is lacking. I've got hummm all over the patterns and its tough to read things with my Tek 453. Anyway, with a 100mV, 3khz signal feeding the Ranger, I'm getting about 60% mod with the modulator biased for 60mA static, it goes up to ~71% at a 70mA current. This is with the pot maxed out.

Compression....  Yes, its a good way for a lousy signal to sound even lousier. At least that is what Les Ismore says. Maybe I'll find a cheep CBS Audimax to play with, but not for now.

klc

[WD5JKO]

My bench set up is lacking. I've got hummm all over the patterns and its tough to read things with my Tek 453. Anyway, with a 100mV, 3khz signal feeding the Ranger, I'm getting about 60% mod with the modulator biased for 60mA static, it goes up to ~71% at a 70mA current. This is with the pot maxed out.

Kevin,

   The Hum on the scope needs to go!

If you are using a random wire as a scope pickup, then just shunt the wire to ground with a RF choke (like a 2.5 MH) at the scope input. Hum gone!

Or at the Ranger power levels, use a 10X scope probe right at the Ranger RF output. It will be safe so long as the Ranger sees a 50 ohm RF load. Use the scope ground clip too. At higher power, I just insert a 10 pf cap (3KV) in series with the scope probe tip.

Since you see an increase in modulation power as you increase the static modulator plate current, that suggests your modulator is limiting after class A operation and something is caving in as you transition to class Ab1 operation. More than likely the modulator screen voltage is falling, or the bias is shifting around (class Ab2).

I find it amazing how good some rigs can sound with nothing more than a D-104. Don't go crazy buying boxes outside the rig when the problem is inside the rig. 



Jim
WD5JKO

[AB2EZ]

Kevin

One way to increase the modulation depth is to lower the B+ on the 6146 (but not the B+ on the modulator tubes).

This will, of course, lower the average output power; but from the perspective of how you sound on the air, you will actually sound louder. This is because of the AGC/AVC action in a typical receiver. In a roundtable... after the listener adjusts the volume level... signals with a higher ratio of carrier to modulation will sound too soft.

There are a number of ways to lower the B+ on the 6146. I use a separate, adustable, external B+ supply in conjunction with the 9 pin socket on the back of the Ranger. By adjusting the B+ on the 6146, I can adjust the average power level. Since we are only talking about a maximum of around 600V at 125mA, this is not an expensive proposition. I use a small 120VAC=> 480VAC "control" transformer and a bridge rectifier. I use a small variac at the input to the transformer.

Separately... in my experience, by far the weakest link in the Ranger audio chain is the interstage audio coupling transformer. It produces a great deal of distortion, due to saturation. There are lot's of ways to replace it. One way... that I used... is shown here:

http://mysite.verizon.net/sdp2/id13.html

Stu

[KB2WIG]

Thanks fer the feedback.

I'm using a few turns of wire off of the 10X probe as a pick up. The choke thing is a better way to go. I'll recheck the  bias as I modulate the rig. The disturbing thing is having to max the audio pot to get apreciable output.  Something is wrong with the 12ax7 sections. This is a simple circuit and should be easy to fix. But not fer me.
 
I'm not going to do the add on thing untill this unit starts to act like it should.

Dropping the  B+ on the '46 is an option....  and its kind easy using the socket on the back.


klc

[AB2EZ]

Going along with what Jim said:

When using an analog oscilloscope to look the the modulated RF output... one will generally be forced to use a sine wave audio input signal.

The sine wave audio input signal, when increased to a peak amplitude that is higher than the grid bias voltage, will cause grid current to flow in each of the modulator tubes every audio sine wave cycle. But, the associated average grid current will interact with the grid bias supply to make the grid bias more negative... until the average grid current is reduced to (essentially) zero. I.e. the modulator will automatically re-bias to class AB1 if a sine wave audio signal is applied.

Therefore, if one is trying to operate the modulator tubes in Class AB2... which is typical for a Ranger operating with the same B+ on the modulator tubes and the 6146... the effect described above will make this impossible with sine wave audio applied to the Ranger.

With ordinary speech, peaks (both positive and negative) occur very infrequently (a small percentage of the total time)... the charge associated with the infrequent, short bursts of grid current will be bled off of the grid bias supply bypass capacitor by the various parallel resistors.... and, therefore, the capacitor across the modulator grid bias supply (15uF) will hold the bias voltage at the value it is set at with no audio applied. This will allow one to achieve larger audio peaks with voice than with a sine wave.

Stu

[AB2EZ]

Kevin

When applying a 3kHz sine wave... make sure that there is a capacitor (e.g. 0.1uF) in series between the output of the audio generator and the microphone input to the Ranger. If the audio generator looks like a modest impedance to ground (e.g. 600 ohms), and if you don't have the series capacitor... then you will remove the (grid leak) biasing of the 1st audio preamplifier stage of the Ranger... causing the plate voltage in the associated half of the 12AX7 to drop to a very low value (because of the 470k ohm and 220k ohm series resistors between the plate and the 300V supply), and signifcantly reducing the audio gain of the stage.

The capacitor is needed with any audio input source or any microphone (except an un-amplified D-104 or a microphone that has a high impedance to ground at DC)

Stu

[KB2WIG]

"  the audio generator and the microphone input to the Ranger. If the audio generator looks like a modest impedance to ground (e.g. 600 ohms), "

Duh.       Like I said earlier, it aint easy.  Lets see, that was explaind in class back in the 70's. And I still screwed it up.

The HP 200CD is transformer coupled, 600 ohm out .......   Well, I'll isolate the generator and see what things look like. I did hang a 100K pot off the generator to get the signal voltage down to 100mV. But that still doesn't do the bias any good.

 
"Going along with what Jim said: "

Thanks Stu, I wasn't following it too well.


"one will generally be forced to use a sine wave audio input signal"

I played with the 1khz cal output off the scope and fed it into the rig. But the scope really didn't like that waveform.

FWIW, Per Wa1HLR, I solid stated the bias rectifier, increased the cap to 100uF, and added a pot to vary the bias. Both the HV and LV were solid stated; the modulator is on the low voltage supply.   

I'll play some more tomorrow.  Thanks guys,

klc

[KD0HUX]

Ah, mo betta. Can't go wrong with the 6550s.

[Opcom]

Stu,

Yes, the scope is the berries....  My bench set up is lacking. I've got hummm all over the patterns and its tough to read things with my Tek 453. Anyway, with a 100mV, 3khz signal feeding the Ranger, I'm getting about 60% mod with the modulator biased for 60mA static, it goes up to ~71% at a 70mA current. This is with the pot maxed out.

Compression....  Yes, its a good way for a lousy signal to sound even lousier. At least that is what Les Ismore says. Maybe I'll find a cheep CBS Audimax to play with, but not for now.

klc

hum on a scope is almost always filter capacitors.

[KB2WIG]

P,

I shuda been clearer.... The hum is only humming when i'm doing modulation monitoring.


klc

[AB2EZ]

Kevin

If you have a ferrite core, try making an isolation transformer that will let through the modulated RF, but block 60 Hz and harmonics of 60Hz:

Use a few turns on the primary and an equal number (actually it doesn't matter that much if the numbers of turns are equal) on the secondary.

The output of the RF sniffer should connect to one of the windings, and the scope to the other.

Make sure that there is no ground connection directly between the scope and the Ranger. Make sure that this ferrite core transformer is located outside of the Ranger, at least several feet from the Ranger's AC supplies (transformers). Use a twisted pair of wires to connect the pickup to the ferrite transformer,

Note: when you use a few turns of wire... located inside the Ranger or near the Ranger's power supply transformers... as a pick-up for the scope...   you have a problem with the magnetic fields of the Ranger's power supply transformers inducing hum into the pickup . [This is exactly the same as hum induced into guitar microphone pickups or ribbon microphones]

For that reason, it is best to use an RF sniffer that is coupled to the output of the Ranger somewhere along the coaxial cable (external to the Ranger).... but if you add the extra transformer (made from a ferrite core), as above, you will block the low frequency (60 Hz, 120 Hz, ...) hum components that have been induced into the pickup.

The use of an RF choke across the scope input, as suggested by Jim, performs a similar function... but the transformer will also isiolate the scope's ground from the Ranger's ground.

Stu

[KB2WIG]

What I did was take some #12 Awg and make a 2 turn, 1/2 inch coil. The scope gnd on one end, the 'hook' on the other. Placing the pick up coil near the tank, I would adjust the un modulated trace to display 2 grads on the screen.... With the cover off, ther's rf floating all over the place. 

The ferrite isolation coil is a gud idea. I'll try that....  I do have a pick off point on the dummy load, maybee i'll just mount a  toroid off of it....

back to the bench.


klc

[AB2EZ]

Kevin

You wrote:

"FWIW, Per Wa1HLR, I solid stated the bias rectifier, increased the cap to 100uF, and added a pot to vary the bias. Both the HV and LV were solid stated; the modulator is on the low voltage supply."

These are good things to do... but even with the 100uF capacitor, application of a periodic input signal (like a sine wave) will cause the bias voltage to become more negative (i.e. the capacitor will charge up to a more negative value) if you increase the amplitude of the audio signal to the point where grid current flows on audio peaks. This effect is enhanced (made worse) by the use of the resistive voltage divider.

Again, the effect will be negligible with voice input, because the statistics of voice are such that the signal is at peak amplitude only a tiny percentage of the time. E.g., there are long periods of silence between words, etc.

Therefore, you need to be careful about drawing conclusions regarding what your Ranger's peak modulation capabilites are (for voice) when you make measurements using a sine wave.

I had this problem with my KW-1, and with my Ranger. To solve the problem, I made a bias supply (for each of them) using a string of Zener diodes ... whose values were chosen to produce the desired resting current in the modulator. Since there is essentially zero series resistance (for a bias supply that employs Zener diodes), drawing average grid current during modulation does not have any effect on the bias voltage.

Stu

[KB2WIG]

Stu,

Thanks for the input.  This stuff is interesting - the small stuff can sneak up and bite. I've got a 43 volt zener (1N4755) accross a 10k pot. This takes the place of the 'bottom' voltage divider resistor in the bias rect. circuit.

I think i'll try the zener ckt after I get this Ranger checked out and working better. There's a Viking 2 after the Ranger 2.


FWIW, I've got the Viking loaded up on 7290 if anyone is interested.....

klc