Designing a low power rig around the mod. transformer.

[stevef]

I am developing ideas for a homebrew rig, but I am starting with (and building outward from) the modulation transformer I already have on hand:

Triad M-12AL Multi-Match
4000-2000 ohms impedance range
125 watts
300mA max dc pri. and sec.

I am thinking about the modulator and PA at this time and not concerned about the speech amp or exciter just yet.
Plate-modulated Class C PA.
Probably single-band (75m)

To provide some headroom on the modulation iron, I am supposing around 100 watts for the modulator.
Using the “50% of Class C input power” rule of thumb, I figure 200 watts input power on the PA.

My first thought was a pair of 6146s for the RF PA.
600v at 280mA.
160 watts in, 124 watts out.
2143 ohms modulating impedance.
**My mod. transformer only matches down to 4000 ohms so may not work well here.

My second thought was a single 813 for the RF PA.
1250v at 150mA
187 watts in, 140 watts out
8333 ohms modulating impedance.
**The mod. transformer will work here.

For the modulator, a pair of 807s in push-pull, triode-connected.
600v at 240mA (max signal)
91 watts audio power out
5050 ohms plate-to-plate
**The mod. transformer will work here.

It appears I have some wiggle-room to tweak voltages/current/power a bit and keep the modulation transformer happy.  My reading indicates this is similar to the ART-13 setup.

Any thoughts or suggestions before I start designing the rest of the rig?

Thanks,
Steve KK7UV

[W3RSW]

Don't forget the 4d32 Raytheon for the RF final.

I've seen several articles where 6146 s are purported to be somewhat distorted AF amplifiers when used as beam tetrodes, so your choice of 807s in triode config. sounds good for the modulators.

I have a stand alone modulator with 6146s into the same M12 AL but never used it much. It has a crappy small driver xfor from 12au7 into the 6146 grids as was the practice then and it's ok for comm quality but raspy and thin.

Btw, Steve put the impedance data for the m12aL that I sent him on his AM Window site if anyone else needs it.

[kb3ouk]

If you go with the 813 modulated by a pair of 807s, I would drop the plate voltage on the 813 down to about 1000 volts, that should make it easier to get 100% modulation with the 807s as the modulators, with the voltages you have right now as a start, those 807s are just barely going to do it, especially when you consider that there will be some loss in the mod iron. Something I've wondered about with tetrodes is if you had a source of modulation for the screen that could be varied, could it be used to make up for a lack of modulation power from the plate modulator.

[stevef]

>> if you had a source of modulation for the screen that could be varied<<


The 813 plate and G2 screen could be modulated together as shown on W8JI page under "plate modulated tetrodes":

http://www.w8ji.com/amplitude_modulation.htm

The R1 and R2 values on the screen would allow you to vary the degree of screen modulation.

[W2VW]

Some interesting reading here concerning Mr. Screen:

http://amfone.net/Amforum/index.php?topic=36983.0

[KA2DZT]

The 807s are not going to give you the AF power you need.  Design for as much audio power as the PA.  A single 813 sounds good but try to use 811As for modulators.  Build one HV supply at 1200V to 1400V (300ma) to run both the PA and the modulators.  You should get about 150W to 175W RF output.  This could be easier to build than with 6146s and 807s.

Fred

[Steve - K4HX]

He did include in the title, low power. Low power has different meanings to different people. But let's just say in this case, it's 100 watts RF output or less. If so, the 807s are fine.

If the design approach is that the mod tranny is rated for 125 watts, therefore the final can run 250 watts input, then yes, the 807s are not appropriate.

Personally, even though the tranny is rated for 125 watts, I'd design the transmitter for 100 watts RF output. This means the tranny would be loafing along and would tend to last forever. In this case, two 6146s or a single 4D32 would be good RF finals. The 807s would be good modulators.

Another thought, the triode connected 807s idea is cool, but maybe just go with actual triodes like 809s or 811/811As.

Lots of possibilities. 

[WA5VGO]

I have a 4D32 modulated by 809's that I built about 10 years ago. It also uses a 125 watt poly impedance transformer although it's a Thordarson. With 750 volts the transmitter will easily delive 100 watts with plenty of headroom. The VFO tunes 160 meters. The first buffer (5763) is broadly tuned to 80 meters and the grid of the 4D32 has a wide range circuit that will resonate on either 80 or 40 meters. The only thing the band switch does is short out a portion of the coil in the pi-network output. It's a very simple way to achieve two band coverage. In my experience the 4D32 is much more rugged than a pair of 6146's in spite of the comparisons we frequently see. The 809's will operate zero bias with 750 volts, but I add 4.5 volts with a 5 watt Zener diode in the cathode circuit. The 809's will drive much easier than you might think. I'm using a 6BL7 in push-pull and have plenty of reserve drive. 

Darrell

[N2DTS]

That sounds like a great rig, got a diagram for it?
I am interested in the broadly tuned stuff.

I built an exciter but have to switch the grid between 80 and 40 meters.

I have a 4D32 modulated by 809's that I built about 10 years ago. It also uses a 125 watt poly impedance transformer although it's a Thordarson. With 750 volts the transmitter will easily delive 100 watts with plenty of headroom. The VFO tunes 160 meters. The first buffer (5763) is broadly tuned to 80 meters and the grid of the 4D32 has a wide range circuit that will resonate on either 80 or 40 meters. The only thing the band switch does is short out a portion of the coil in the pi-network output. It's a very simple way to achieve two band coverage. In my experience the 4D32 is much more rugged than a pair of 6146's in spite of the comparisons we frequently see. The 809's will operate zero bias with 750 volts, but I add 4.5 volts with a 5 watt Zener diode in the cathode circuit. The 809's will drive much easier than you might think. I'm using a 6BL7 in push-pull and have plenty of reserve drive. 

Darrell

[stevef]

Putting the discussion back on track...

P-P 809 modulator seems like a good match for the single 813 PA and the 125w modulation transformer, at least on paper.

Some suggest 811's for the modulator, but it looks like that is up around 300 watts under the 'typical operation' values on the RCA data sheet.  To bring them down in the 125-watt range (my modulation transformer power rating), is it a simple matter of using a higher plate load impedance via the transformer taps?

I have the HamTips vol.8 no.1 which displays the calculations for doing so for the 811.

Should I load them down even more to stay under the 125w transformer spec?  Say 100 watts?

My transformer is spec'd at 300-3000 Hz.  While not hi-fi, would this be objectionably 'tinny' audio?

Thanks,
Steve KK7UV

[KA2DZT]

Those multi-match mod xfmrs are usually a step-up or down ratio using the entire windings on both sides.  Meaning both sides of the xfmr are not the same.  Everything depends on the plate voltage you use.  813/811s at 1200V-1300V on both, that xfmr will work FB.  Don't worry about the power rating as much as the voltage you put on the xfmr.  For a nice low to medium power rig this line-up will work FB.  Another thing, use the entire winding (no taps) on the mod side and try to do the same on the PA side.  Put the higher ratio side of the mod xfmr on the modulators and you'll have a step-down going to the PA.  Put the two windings for the PA side of the mod xfmr in series.  You may end up using less winding on the PA side.  You can do this by putting the HV B+ on one tap up from the B+ end.  Never tap down on class C end of the xfmr.

Start building

Fred

[Steve - K4HX]

Or just don't drive the 811s as hard.

To bring them down in the 125-watt range (my modulation transformer power rating), is it a simple matter of using a higher plate load impedance via the transformer taps?

[stevef]

Sounds like I should just get some 811's, use the mod tranny I have, and adjust impedance taps and drive to find the best operating point.

I'll be going with the 813 PA and I can adjust its plate voltage via power transformer taps and variac to anything between 1000 and 2000 volts.

 

[N2DTS]

813's are great tubes in plate modulated service, last forever.
One is good for 300 watts carrier at 2000 volts and 200 ma.
You can run it at less then that also.
A pair of 4D32's will give 200 watts output, 500 to 1200 volts on the plates works fine, they are $24.00 each new.

I have three 4D32's modulated by 811,s off a 1200 volt supply, 300 watts carrier. Works well.
Lower voltage makes for smaller RF parts, higher voltage makes for more power and less current through the mod transformer.

Its a good idea to have AS MUCH audio power as RF output power, that way the audio runs very clean.

For modulators, 807's, 809's, 811.s or some sort of HIFI tubes.

[w1vtp]

My good friend Paul, WB2SKC generated this spreadsheet for the M12AL.  Paul is a real clever guy.  Just enter the data into the blue fields and up comes the data

Al

[stevef]

Thanks for the spreadsheet!  Very helpful.

I will look into the 4D32 option also.

Steve

[N2DTS]

Tim (hlr) always said to use ALL the transformer, forget the taps, just get the best ratio you can and have no windings hanging in the wind and use all the inductance of the transformer.

Are you going to build the vfo and driver stages also?
One band makes it very easy to do.

[KA2DZT]

Tim (hlr) always said to use ALL the transformer, forget the taps, just get the best ratio you can and have no windings hanging in the wind and use all the inductance of the transformer.

Are you going to build the vfo and driver stages also?
One band makes it very easy to do.

I agree, I already mentioned using the entire windings in my first reply.

[stevef]

The transformer is 20k:20k on the full windings, which throws a wrench into my calculations.

Please provide some rationale and theory with recommendations.  I don't find "do it this way, works great" all that useful otherwise.  I'm using what I can find in the handbooks and such (Radio Handbook 15th ed, HamTips, RCA TT-5, etc.) to understand how to put it all together.

Yes - I will build my own speech amp and VFO/exciter too.  Most likely it will be single band (75m) with the possibility of 40m too but not likely since all I ever hear on 7290 is Chinese and Russian shortwave.

Steve 

[KA2DZT]

My HB 813/811s operates on only 75M and 40M.  I rarely ever operate on 40M.  It's a two stage xmtr it also has a 6146/6550s.  The 6146 also drives the 813.  It's a complete xmtr from the VFO.  So your idea of just building for 75M is going to be a lot easier.  If you build a VFO start your oscillator on 160M and double up to 75/80M.  This is a must or you may end up with FMing problems.

Don't worry about the impedance of the mod xfmr, it's the turns ratio that's important.  My mod xfmr is a Stancor 175W unit that was shorted when I got it.  I open the xfmr, removed the shorted outer winding and re-insulated it (it wasn't a multi-match type).  I ended up with 9000ohm primary and 5000ohm secondary (turns ratio of about 1.3-1 step down, just right). It works FB, I also have a 50hy mod reactor that I use with it.

Your mod xfmr is 20K-20K.  If you load one side at 10K then the other side will also reflect 10K.  It the turns ratio that's important.  Just use the whole winding on the mod side and use what ever portion of the secondary to give you a small step down ratio.  The impedance on the mod side will depend on the load presented by the PA side.  It's the PA side load impedance that reflects back the impedance seen by the modulators.  The PA load impedance is simply the plate voltage/plate current+screen current (if you modulate the screen with a dropping resistor from the modulated plate voltage).

Fred