Converting Collins 820D-2 AM broadcast transmitter to 4-400As.

[WD5JKO]

What would be the advantage of doing it that way?  Dropping the voltage on the unmodulated B+ line eliminates the need for the bypass oil capacitor, but otherwise the loss is the same. 

  Don, You make perfect sense, and there is no advantage in what I suggested. Thanks for presenting the case to make me realize that.

Jim
WD5JKO

[Gito]

Hi

The way I see it,the audio driver is design to drive a class AB1 final (a voltage source)
It has a 5k (pot) and a 15K resistor as a load for the collectors of 2N 3585 ,a small current flow make a large voltage drop in it.It is design as voltage follower source

As we know Class AB1 needs no power to drive it,so as you use the audio driver to drive a class AB2 or class B ,and current begins to flow at the grids of the 4-400 ,it also cause  the current flow at the transistor trough the 5k+15K resistor and make the audio driver distorted.

To my surprise looking to the data sheet of a class AB1 4-400 data sheet,
it has a - 137 volt dc ,so for full output power it needs a 274 peak to peak across the 4-400 grids,the power output is 1110 watt output ,more than enough to modulate a 1000 watt am transmitter.

With this setting Class AB1 ,the 4-400 needs only a voltage drive.

And not the usual way the class AB1 4-400 needs a - 137 volt bias
                                    Class AB2 4-400 needs a -   80 volt bias

So to make it class AB1 we need  a  - 137 volt DC bias

Gito.N

[DMOD]

I would be interested in knowing what the voltages are on the terminal strips for 12, 9, and 10.

Pin 9 looks to be +250Volts? and pin 10 appears to be -155 Volts?

Thanks

Phil

[K6IC]

Hi Brian,  and all,

Interesting thread.  You probably have looked at this transmitter and its modulation as much as anyone (short of Collins) ever has.

No Money Mike should be a wealth of info.

In addition,  several others in this region have put 820-D2s on the Ham Bands:
1.  Ted,  WA8ULG in A2 --  ted at vsb dot com (from QRZ)
2.  Jeff, WB6ZBX  I have no e-mail for Jeff

Will try pointing each of these guys to this thread,  in case they have any additional info on the comversion.

ZBX Jeff has also converted the Collins/Continintal 3-500 PDM rig to Hammie bands, and seems to be a Techie.

73  GL,  Vic

[WB6QED]

To my surprise looking to the data sheet of a class AB1 4-400 data sheet,
it has a - 137 volt dc ,so for full output power it needs a 274 peak to peak across the 4-400 grids,the power output is 1110 watt output ,more than enough to modulate a 1000 watt am transmitter.

With this setting Class AB1 ,the 4-400 needs only a voltage drive.

And not the usual way the class AB1 4-400 needs a - 137 volt bias
                                    Class AB2 4-400 needs a -   80 volt bias

So to make it class AB1 we need  a  - 137 volt DC bias

The audio driver stage is capable of about 500 volts p/p before clipping when running open circuit.  When connected to the properly biased modulators without the negative feedback connected, the audio driver will do about 450 volts p/p at the input to the modulators.  When feedback is connected, the audio driver output is less, but it still exceeds 300 volts before clipping.  I believe that there is plenty of voltage swing available from the audio driver.

As I have mentioned in my previous posts, at high power (1 kW), the transmitter is operating well.  The problem is at low power (250 watts).  My conclusion is that this is because the modulator screen voltage remains at 750 volts when the plate voltage has been reduced to 1500 volts.  This was probably ok with the 5-500A pentodes, but results in an inadequate amount of modulator swing being available with the 4-400A tetrodes.  Possibly if the modulator screen voltage was reduced to 375 volts the 4-400As would work perfectly at low power.

Brian, WB6QED

[WB6QED]

I would be interested in knowing what the voltages are on the terminal strips for 12, 9, and 10.

Pin 9 looks to be +250Volts? and pin 10 appears to be -155 Volts?

Pin 12 is ground.
Pin 9 is +290 volts (+285 volts measured)
Pin 10 is -155 volts.

[WB6QED]

No Money Mike should be a wealth of info.

In addition,  several others in this region have put 820-D2s on the Ham Bands:
1.  Ted,  WA8ULG in A2 --  ted at vsb dot com (from QRZ)
2.  Jeff, WB6ZBX  I have no e-mail for Jeff

Will try pointing each of these guys to this thread,  in case they have any additional info on the comversion.

ZBX Jeff has also converted the Collins/Continintal 3-500 PDM rig to Hammie bands, and seems to be a Techie.

Hi Vic,

Thanks for the note!

I have discussed this conversion with Mike, KO6NM, and Ted, WA8ULG.  Mike has converted his 820D-2 to 4-400As.  Ted is still running 5-500As.  

I am very interested in speaking with Jeff, WB6ZBX, not only about the 820D-2, but about his 828C-1/314R-1 conversion as well.  I am very curious how he modified the RF driver stages to 75 meters in both transmitters.  

Brian, WB6QED

[DMOD]

My conclusion is that this is because the modulator screen voltage remains at 750 volts when the plate voltage has been reduced to 1500 volts.  This was probably ok with the 5-500A pentodes, but results in an inadequate amount of modulator swing being available with the 4-400A tetrodes.  Possibly if the modulator screen voltage was reduced to 375 volts the 4-400As would work perfectly at low power.

I think you're on the right track there.

When connected to the properly biased modulators without the negative feedback connected, the audio driver will do about 450 volts p/p at the input to the modulators.  When feedback is connected, the audio driver output is less, but it still exceeds 300 volts before clipping.  I believe that there is plenty of voltage swing available from the audio driver.

For sure. I was concerned you might have bad coupling caps or leaky audio driver output transistors.

In BC rigs with SS drivers, whether RF or audio, I have seen cabinet heat increase cap and tranistor leakage.

My previous comments about not having enough audio drive are now nullified. 

Phil

[Gito]

Hi

I may be wrong,but why do you need so high peak to peak drive voltage?
if Your P&P drives exceed the  - bias supply ,than You go in class AB2 ,that means grid current flow in the modulator tubes,And I see a 100kohm grid leak at the grids of the 4-400 modulator tubes ,the grid current will also flows in the 100 Kohm ,a 3 ma trough it will developed a  300 volt voltage ?.

You wrote with the modulator tube connected the drive voltage is 300 v P&P,is it possible ?,that where the current begin to flow in the grids of the modulator tubes.


The question is, using a 4-400 with 3000 volt on plate  750 volt on screen and - bias 137 volt needs only 274 P&P to give a 1110 watt output.

Gito.N

[WB6QED]

I may be wrong, but why do you need so high peak to peak drive voltage?

I was only trying to convey that the audio driver stage has plenty of peak voltage output capability to drive the modulators properly.

I ultimately adjusted the audio input level and the modulator bias and balance controls for best symmetry and total harmonic distortion at 95% modulation.  The modulators are running about 200 mA of static plate current without modulation which is well into class AB1 operation. 

Brian, WB6QED

[DMOD]

Hi Brian,

I have somewhat of an academic request, but if you have time, could you capture the waveforms at Q3 base and collector and at say E13? 

I think the waveform at E13 may be a sine pulse of approx 120 degrees or possibly less.


As far as moving up in freq. to 75 meters, I would think that reducing C13 to about 100puffs and the grid tune to 5 to 245puffs might help. You might have to use a wafer switch to switch in and out varous capacities in these areas.

Is there a chance we can see the tuned output circuit?

Needless to say, this rig has certainly generated a lot of interest.   

Thanks

Phil - AC0OB

[Gito]

Hi

You are right Brian,

It's a class AB1 modulator with that static current, and will stay AB 1 if there's not grid current flow in modulator tube at maximum drive.

class AB1 3000 v plate 750v screen -137 v bias zero signal DC plate current 160ma
max signal DC plate current 635 ma.

class AB2 3000 v plate 750v screen -80 v bias Zero signal  DC plate current  160ma
max signal DC plate current  700ma.

With different  min bias the static plate current stay the same.(from Eimac)



Gito.N

[WB6QED]

I have somewhat of an academic request, but if you have time, could you capture the waveforms at Q3 base and collector and at say E13? 

Here you go!

Brian, WB6QED

[WB6QED]

Is there a chance we can see the tuned output circuit?

The RF output network is a bandpass filter that presents 3350 ohms to the plates of the final amplifier tubes when terminated with 50 ohms at the RF output port.

Brian, WB6QED

[DMOD]

The RF output network is a bandpass filter that presents 3350 ohms to the plates of the final amplifier tubes when terminated with 50 ohms at the RF output port.

Well, actually, it's more than that.

What you have there is a 4-node, bandpass circuit that also matches impedances.

The bandpass characteristics of the filter are that of a flat response hat with steep skirts.

This output design was also used in some of the tube-type Harris PDM, AM BC rigs.

For a switch-mode RF drive circuit, the grid signal appears very clean.

You have a very clean rig there, spectrum wise, and are very fortunate to have such an historic rig.

Phil - AC0OB


 

[n2bc]

The post from Gito above,

AB2 Screen voltage should be 500 rather than 750.   It didn't make sense that ZSAC would be the same in AB1 and AB2 with less control grid voltage.

[Gito]

Hi


You are right,I miss type The screen voltage,it must be 500 VDC,

Gito.N

[WB6QED]

I am back at it again!  I tried substituting the 1000 pF PA screen bypass capacitors with 100 pF capacitors to see if that would improve the THD at 10 kHz and unfortunately, it made little difference.  I then tried substituing the 1000 pF modulator screen bypass capacitors with 100 pF and got the same result.  No improvement.

This morning, I made some additional audio measurements with the modulator feedback circuitry disconnected to eliminimate it from the equation.  If you take a look at the attached .PDF you can see the results.  The Audio Driver output looks very good.  I think that it is safe to say that it is ok.  I need to get my hands on a second Tektronix P6015 high voltage probe before I can make differential measurements at the input to the modulation transformer.  I was able to make single ended measurements at the modulation transformer output with the one high voltage probe that I have and the results weren't very favorable.  I am thinking that I need to get a couple of 3K 225 watt resistors to test the modulation transformer into a purely resistive load and to isolate it from the RF PA.

I'd appreciate any thoughts and recommendations as to what my next steps should be. 

What can cause this sort of non-linearity at the high end with such good performance at the low end?

Sincerely,

Brian Henry, WB6QED

[WD5JKO]

Brian,

   At 10 Khz tone input , the harmonics will at some point get a boost since the modulation transformer will go through self resonance. My earlier post did touch upon this:

"You mentioned the Modulated B+ peak level rises at 10 Khz over that at 1 Khz. You also see audio phase shift on the 10 Khz trapezoid. All audio transformers have a self resonant frequency, and phase shift as we approach this resonant frequency. We like this frequency to be way above the highest audio frequency modulated though. A couple of things will help here. First a series R-C from modulated B+ to ground where the 'R' might be 2X the modulator load impedance, and the 'C' a value where the Xc matches the 'R' at 10 Khz. Just a guess, but an easy test to see what happens. The other thing would be to have a low level LPF to limit your upper audio frequency response to 5 Khz or lower."

   If you can do a FFT plot versus a meter indication of THD, then maybe that will shed light on the subject. In the absence of that a scope picture should reveal whether your fighting high even order versus odd order distortion products. I am no expert here, but high 2nd harmonic THD shows gain compression on one half of the sine wave, and high third harmonic might show a more abrupt clipping on both sides.

Jim
WD5JKO

[WB6QED]

If you can do a FFT plot versus a meter indication of THD, then maybe that will shed light on the subject. In the absence of that a scope picture should reveal whether your fighting high even order versus odd order distortion products.

The plot thickens......

This morning, I borrowed the modulation transformer from a Bauer 707 transmitter.  While not identical, I wanted to see if I could get the 10 kHz distortion problem to move.  It didn't.  I got nearly identical results with the Bauer modulation transformer installed that I get with the Collins modulation transformer.  Both modulation transformers were manufactured by Electro Engineering Works in San Leandro, California.

The Collins transformer has these specifications: Primary: 4240.4 VCT, Secondary: 3394.2 (1.24:1 step down)

Collins modulation transformer THD at approximately 90% modulation:

100 Hz:  1.1%
1 kHz:    0.64%
10 kHz:  4.0%

The Bauer transformer has the following specifications: Primary: 3336 VCT 40 Hz-10 kHz, Secondary: 2170 V 0.35A (1.58:1 step down)

Bauer modulation transformer THD at approximately 90% modulation:

100 Hz:  1.0%
1 kHz:    0.72%
10 kHz:  3.8%

I also carefully retuned the output network using a UHF AIM and achieved a very flat impedance across the 20 khz bandpass.

I have verified the component values in the PA and modulator sections and everything looks correct.

I was able to reduce the 10 kHz THD by about 0.5% by changing the modulated B+ bypass capacitance from 500 pF to 50 pF (C34 & C35).

The 10 kHz THD improves to about 2% at 50% modulation.

All measurements were made the audio feedback disconnected.

Brian Henry, WB6QED

[WB6QED]

Here is the impedance plot of the RF output network.

Also, 10 kHz modulation performance at 50% THD.

[WB6QED]

I had a major breakthrough earlier this week and was able to make a signficant improvement in the 10 kHz THD.  It is now under 1.8%!  I increased the screen voltage to the PA tubes by reducing the value of the screen voltage dropping resistor from 3K to about 1.6K.  The PA screen voltage is now a little over 500 volts and the total screen current is a little over 130 mA which is right at the screen dissipation maximum of 35 watts per tube.  If I push the screens over the dissipation limit, I can get even better performance.  With the 5-500As, Collins tied the screen and the suppressor grids together.  This probably gave the tubes a little extra dissipation capability.

With the audio feedback disconnected, I have now achieved less than 1% THD at 100 Hz and 1 kHz and less than 1.8% at 10 kHz at 95% modulation with 1 kW of RF output power.

I now have a balance issue to resolve when the audio feedback is connected to the input of the audio driver.  When the feedback is connected, the MOD 2 DRIVE potentiometer wants to be fully CCW for minimum THD.  I have swapped the modulator tubes with the PA tubes and there is no change.  I also checked the components in the feedback ladders and everything appears to be ok.  However, I did find several 1/4 watt carbon composition resistors on the audio driver circuit board that exceeded 10% of their specified tolerance.  I intend to replace those which should get the audio driver into better balance.

More to come......

Brian Henry, WB6QED

[WB6QED]

Just for fun, I swept the frequency response of the RF output network with the tracking generator in my 7L5 spectrum analyzer.  See the attached photos.

Brian Henry, WB6QED

[WB6QED]

Here are two more photos of the "Q-Taper" RF output network response.

[w3jn]

I had a major breakthrough earlier this week and was able to make a signficant improvement in the 10 kHz THD.  It is now under 1.8%!  I increased the screen voltage to the PA tubes by reducing the value of the screen voltage dropping resistor from 3K to about 1.6K.  The PA screen voltage is now a little over 500 volts and the total screen current is a little over 130 mA which is right at the screen dissipation maximum of 35 watts per tube.  If I push the screens over the dissipation limit, I can get even better performance.  With the 5-500As, Collins tied the screen and the suppressor grids together.  This probably gave the tubes a little extra dissipation capability.

That's pretty much what I said on the first page of this thread; apparently, you missed it.

-225 is about right for a 4-400 class C, and 80 mA is just about right for screen current.  What are the values of the screen bypass caps?  If the time constant is too long the screen won't self modulate.  Screen voltage seems a bit low also.  You should tune the transmitter for a peak in screen current, rather than the absolute dip in plate current.  The peak in screen current is a bit sharper.  Try increasing the screen voltage to about 500-600V, then adjust the drive for 80 mA screen current.

Take a peek at the GPT-750 manual and see how TMC did it (4-250s in the final, but 4-400s are drop in replacements) http://bama.edebris.com/manuals/military/gpt750/

Another thing you can try is reversing the plate caps on the modulators, see if the peaks improve.