The AM Forum

Looking at Tom's (K1JJ) 813 Mod'ing 813's circuit I have questions about the audio transformer.

What type of transformer should be used to drive the 813 grids?  Would a different transformer be used if using a tube speech amp vs a SS amp?

Thanks

Rich

Those are kind of open ended questions and hard to answer.

Some of the designs in handbooks show 813's in triode mode driven by "standard for the 1950's" circuits using 2a3s through multimatch driver class B driver transformers.

Orr has, in the 1959 Radio Handbook, a cicuit for driving an 813 also with a single tube (can't remember which).

If you use an off the shelf solidstate amp, you could perhaps use a plate to voice coil output transformer hooked up in reverse.

If your amp has high level line outputs, then you could use a line to grid driver transformer like a UTC CG-59ax.

If you use the 813's in AB1 you wouldn't need a driver transformer and could drive them with a pair of 6sj7s

Probably lots of other choices.

Rich,
I'm using a hi-fi audio transformer from a gutted Fisher audio amp. It's probably good for 25watts.  I don't remember what the primary impedance is but the secondary is multitap 4-8-16 ohms run in reverse. It originally was driven by a pair of 7591's. In the 813 rig it seems OK.  I'm going to try a different audio transformer though because I think response can be improved.  In the 813 rig, I'm driving the transformer with a homebrewed solid state audio amp. It can deliver 30watts or more into 8 ohms which is more than enough to drive Tom's 813 rig. And certainly more into 4 ohms. 

The solid state driver I did for Tom should work great driving 813s. Tetrode or Triode connected. I would think it could be scaled back to a lower voltage supply and lower voltage FETs if desired.

Rich,

The requirements are very different if ur using a backwards connected transformer (speaker impedance - "8ohm") driven by a power amp as compared to using a classic "speech' amp.

The latter generally uses the plates of driver tube(s) which are high - Z to match the grid of the 813.

Depending on AB1 or AB2 you will want high Z (voltage swing and not that much current) and lower Z (voltage swing plus a few watts to handle grid current) respectively.

I prefer to run tubes in A2, AB2 or B2 myself.

Running the power amp into a backwards connected transformer in general terms the output Z of the power amp is very low compared to the load, and there is ample current available to run it into grid current or not depending on if you have cap coupled to the grids or not...

For tubes run without a transformer and still run in "2", the cathode follower is often used.

I think GFZ refers to a direct drive solid state driver circuit that uses high voltage Mosfets, which have both low Z and the requisite voltage swing...

              :D

                                  _-_-bear

The present circuit will do about 200 V P-P at least 100 ma. Frequency limited by the modulation transformer.
This should drive the pants off an 813.
Tom can provide performance numbers.

Another option would be to use a high level line to voice coil transformer.

I have a UTC LS-33 which is a 20 watt transformer. But all the other manufacturers produced them.

Using it back wards, you have 8 ohms in and 600 ohms out. Impedance ration is 1:75 and turns ratio is 1:8.7.

Assuming 20 watts input, you have P = E*E / R. or E = sqrt(R*P)

Therefore the RMS voltage on the 8 ohm side is 12.6.

This gives 110 volts RMS on the secondary.

The peak-to-peak seconary voltage is 2.8 * 110 = 308.

According to the design in the Orr handbook, you need 160 V p-p to drive a pair of triode connected 813's so this transformer would work, unless I'm missing something in this analysis.

Now larger tubes like 810's require p-p drive of 380 volts for full output so this approach wouldn't work there, but it seems like it would work for 813s

In fact, the lower secondary impedance would be a plus in terms of low distortion as compared to using an output transformer backwards.

Thanks for showing me "the math", that helps me know what to look for.  It looks like a generic impedance matching transformer except that it has to handle 20 Watts.  That make make it harder to find.

Rich

After pursuing the audio/modulator designs of others it seems that I can either drive the 813 grids with and external amp using a transformer capable of handling ~20 Watts.  The transformer in this case acts to match impedance.

Alternatively, if I decide to do my own speech amp, then I would require a transformer to drive push-pull grids from a single plate.  The push-pull tubes would drive the 813 grids.

e.g.
http://216.77.188.54/coDataImages/p/Groups/31/31530/folders/15092/927744-125or813Modulator.gif

Then there is this design with THREE audio transformers...

http://216.77.188.54/coDataImages/p/Groups/31/31530/folders/15092/98599WA4KCYSpeechAmpandModulator.gif

I do like the idea of building my own speech amp into the xmtr.  More buzzardly and who wants a SS amplifier sitting on top of those glowing 813s?

Any opinions on the first circuit?

Rich

BTW, I ordered a copy (1959) of Mr. Orrs handbook.   :)

Rich

Hi Rich,

Here's the MOSFET audio driver which uses negative feedback - designed by Frank/GFZ. It would be my first choice for your triode-connected 813 modulators.  

http://amfone.net/Amforum/index.php?topic=23632.0


(As the guys already said) The 2nd choice is to build a DIRECT-Coupled tube audio driver. The BC rigs all used some form of this and the designs are available.  Use negative feedback.


The third easy choice is to get a modern solid state audio amp and find a regular tube ~~2K to 16/8/4 ohm hi-fi transformer to use as a driver for the 813's.  Chuck/K1KW has always used this method and his rigs always sounded pristine.  I used it too until I finally broke down and built the GFZ MOSFEt driver.

In my opinion, the MOSFET driver is by far the best way to go, efficiency and fidelity-wise.

T

Tom,
   Time-wise, it will be all I can do to get this going in the next year.  While I am sure your fet amp is FB, I am hesitant to add that much work to the project.

Perhaps I can start with a output impedance matching transformer driven by my Bogen amp to begin with.  I have some Thordarson 500 ohm to 4/8/16 (I think) but they may not have the power handling capability.  They measure about 3" across.

Rich

FWIW

In reading about my LS-33 transformer on an audio hobbiest forum I ran into a post by a guy who used the transformer in an interesting way.

This transformer is made to match a 500 ohm line (with taps for lower impedances) to either a voice coil, or another 500 ohm line (also with taps).

It has two secondaries -- a 500 ohm secondary and a voice coil low impedance secondary. And the two secondaries are not connected.

He attached the 500 ohm primary in series with the 500 ohm secondary and then used the combined 1000 ohm windings on the plate circuit of a tube.

In effect, by hooking the transformer up this way, he used it as an output transformer for some low impedance tubes in push pull.

This brings up the possiblity or reversing his idea and driving the voice coil input and driving the class B grids from the two 500 ohm windings hooked in series.

This  would give you a higher step up ratio to get more grid voltage swing without driving the transformer so hard.

I don't know if the Thordarsons you have have multiple secondary windings like the LS-33 -- but it's an interesting idea this audio guy had.

Two 500 ohm windings in series will give you a total of 2000 ohms, not 1000. Hmmm. OTOH, that fact doesn't really change the point you were making.  Sorry, forget I said anything <g>
73, Al

"Sorry, forget I said anything "

Actually I appreciate the correction.

You would get twice the voltage swing because of twice the turns, but four time the impedance.

Most of the line transformers that have 500-600 ohm outputs do not have a center tap necessary for bias in push pull grids. It would be ok for use in a large class A jug scheme.

How would you get bias to the two tubes using one of these transformers?

Some have taps on them but it's difficult to get the center of the transformer for balance.

   

The LS-33 has two separate 125 ohm windings on the primary. You are intended to hook them in series for a 500 ohm input.

So this issue wouldn't be a problem with this tranny.

Dave