Modulation transformer impedance ratio 2x811a / single 813

[PA4WM]

Hi all,

In slow, but steady progress of collecting the parts for a class b 2 x 811a / class c 813 TX set up.
I do have a little mod iron on hand, but the impedance ratio (2:1) does not realy match.

I know of someone here in the Netherlands, that winds modulation transformers on demand, so the question of the right impedance match came to mind.

Now, before I order something wrong, I would like to put my calculations below, to have the experienced guys have a look at it....

I plan to run the 813 at a moderate level, because of my QTH lies in the middle of town, and would like to keep a good relationship with the neighbors..

Here is the plan:


TX :Single 813 at 1250v to 1500v./ 150 mA. Output approx. 175 watt
 This will give an impedance of 8 to 10kohm

Modulator 2 x 811a at 1250v to 1500v. Zero bias. Output approx. 250 watt
This will give a P/P impedance of 9 to 12.5 kohm (According to tubedata)

To my opinion, that would mean an impedance ratio of 1.25:1

Am I correct, or did I miss something?

Regards, Martin

[WD5JKO]

Martin,

   You might review the posts at the following thread:

http://amfone.net/Amforum/index.php?topic=27244.msg207633;topicseen#msg207633

My post #8 goes into the math. I dwell on peak modulator capability, voltage ratio, turns ratio, and impedance ratio in that order.

With a common supply, it seems that ~ 1.4:1 turns ratio is often the best. Your 2:1 ratio tranny might play well if the modulator B+ is higher than the RF final B+.

Jim
WD5JKO

[KA2DZT]

Martin,

I'm running the same tube line up.  (1)813 x (2)811a both at 1500 volts.  I'm using a Stancor 175 watt mod xfmr.  Imp 9K (811a) to 5K (813).  Works fine.  I'm also using a 50hy mod reactor to carry the 813 plate and screen current.  At 1500 volts the 811s need some bias.  I use a 4.2v zener diode in the filament xfmr CT return to ground.

Fred

[The Slab Bacon]

Martin,
          An 813 modded by a pair of 811s with a common power supply basically is an ART-13. They start to really come to life around 1200v. Maybe you should look for an ART-13 mod tranny, they are pretty easy to come by from E-pay and other sources.

The ART-13 mod tranny also has a tertiary secondary winding to put a little shake on the screen grid of the 813.

just my $.02 worth

[KA2DZT]

Martin,

Your post states that your mod xfmr has a impedance ratio of 2/1.  That is a turns ratio of about 1.4/1. That xfmr should work fine if it is big enough.

Fred

[PA4WM]

Thanks for the replies so far.

@Fred: As mentioned in the post, I have a 2:1 impedance ratio transformer.
If I calculated it correctly, Yours is about 1.8:1. When I do the math on the tube set up, that is not a perfect match.

@Jim: I read the topic you mentioned, I think I do understand most of it, but I'm still puzzled.
As I understood it, turns ratio, hence impedance ratio has an effect on the low frequency respons as well

It looks like it's not that simple I thought it would be.

I have the opportunity to wind a 1.25:1 transformer, wich would match perfectly to the impedances, or use the Mod iron I have (2:1)

What I don't understood yet, is how bad it would be if I would use the 2:1 impedance ratio transformer, where a 1.25:1 would be ideal

Martin

[KA2DZT]

Martin,

Sorry I misread your post,  I saw that it was the imp ratio of 2/1.  I've since corrected my previous post, you may want to reread it.

Fred

[KA2DZT]

Martin,

The turns ratio is the square-root of the impedance ratio.  Your xfmr has a imp ratio of 2/1.  The turns ratio is therefore the square-root of 2 which is 1.414.  Your xfmr has a turns ratio of about 1.4/1.  It will work FB if is large enough and has a high enough peak voltage insulation.

Fred

[KA2DZT]

Martin,

When you have plenty of excess audio power, the turns ratio is not that critical.  It doesn't have to be exactly perfect.

The fact that your turns ratio may be a little high is better than it being too low.  The higher plate imp on the 811s helps to reduce the peak current the tubes will pull.  You'll still have more than enough audio power to modulate the 813.

Your xfmr should work.

Also, I run my 813 at 200ma plate and screen current combined.

Fred

[The Slab Bacon]

A simple test to get a feel for the turns ratio of the transformer is easy to do.
Just simply apply 120vac (or whatever your mains voltage is) to the full primary of the transformer. Carefully measure the voltage that you are applying, and then measure the AC voltage present on the full secondary of the transformer.

The amount of voltage step up or down should give you a real good feel for what the turns ratio of the transformer in question is.

the actual impedance that the modulator plates are looking into is directly affected by the load on the other side. (Ip and Ep of the final amplifier)

You should be more concerned by the turns ratio and breakdown voltage than the "rated" impedance of the transformer since the impedance will change depending on how heavily you load the final amplifier.

[KA2DZT]

Good post Slab.

Better than your usually olive oil and tomato sauce posts.

Fred

[The Slab Bacon]

Good post Slab.
Better than your usually olive oil and tomato sauce posts.
Fred

Well, if you insist..............

[k4kyv]

Running off a common power supply for modulator and final, a step-down greater than about 1.4:1 (2:1 impedance ratio) fails to produce enough peak audio voltage to modulate the positive (or negative) peaks beyond about 100%. In other words, little or no "headroom" beyond 100% modulation.

No matter how hard the grid is driven, the instantaneous plate voltage of a tube in class-B modulator service cannot be pulled down below about 20% of the DC power supply voltage before the plate current reaches saturation.  Therefore, a tube can supply a maximum peak a.c. voltage of about 80% of the DC power supply voltage.  With two tubes working in push-pull, the maximum peak plate-to-plate voltage that appears across the entire modulation transformer primary is approximately 1.6 times the DC plate voltage.  With a common power supply, a turns ratio of 1.6:1 will theoretically just barely be capable of modulating 100%.  Given that the transformer has some loss, the maximum step-down turns ratio for 100% modulation capability is more like 1.55:1.  To achieve some headroom and allow positive peaks to extend a little beyond 100%, the turns ratio needs to be more like 1.4:1 or lower. A 1.2:1 turns ratio will provide even more positive peak capability, as long as the the modulator tube peak ratings are not exceeded.

The only way to achieve greater peak voltage from the modulator as the tubes are driven near saturation is to increase the power supply voltage.  If the modulation transformer has too much step-down, the modulator plates must be run at higher DC voltage than that of the RF final plates. This precludes a common power supply, unless a voltage-dropping resistor is inserted in the plate supply lead to the RF final.  This resistor must be by-passed with a capacitor to in order to reduce the DC plate voltage without decreasing the audio voltage.  If a modulation reactor is used, the resistor can be wired in series with the reactor (preferably at the +HV end of the reactor), and no by-pass capacitor is needed.

The trade-off in return for the greater peak output capability from the modulator is that a lower step-down turns ratio reduces modulator efficiency and increases the peak plate current of the modulator tubes.

Many commercially built amateur transmitters, military, and older broadcast transmitters deliberately used too much step-down in the modulation transformer to make it impossible to overmodulate the final.  Some transformers have as much as 1.7:1 step-down ratio (3:1 impedance ratio). This allows the tubes to run cooler and more efficiently, but does little to prevent splatter and distortion, since driving the modulator tubes to saturation causes clipping of the audio peaks, producing exactly the same kind of splatter and distortion as overmodulating the transmitter in the negative direction. Since the modulator tubes are driven to saturation to just barely approach 100% modulation, this much step-down may actually result in more distortion.

[K5UJ]

FYI:

http://www.maxmcarter.com/classecalcs/tratiocalc.html
http://www.amwindow.org/tech/htm/modtran/misc/art13.htm