120 Watt AM/CW Power Supply

[W9ZSL]

Everyone has enough plans to last a lifetime, yours truly included.  I hit the big 70 on the 10th and if the Doc gives me a decent bill of health tomorrow, maybe I can build a couple.  I've tossed a bunch out at you over the past few years; the latest being a rebuild of a Heathkit AT-1 to use as an exciter.  That's project one. It's being done.

Project 2 is a speech amp driver on a separate chassis with a pair of 6V6s that can be switched between two separate modulators.

The intermediate project is shaping up to be the 807 amp shown in the '57 ARRL handbook.  I have several 5933s I can use with a pair for the RF and another for the Modulator using an ART-13 mod iron.  This will be a "warm-up" for project 4, the ultimate 813/811A amp.

Here's the question.  Concerning the "807" classic, here's the possible power supply.  Using the plate iron from an Apache (1800 VCT) and a couple Thordarson chokes along with HV caps consisting of 2 x 40 MFD in series for each branch, would this lash up be practical?  The two rectifiers are National SSK-13 / S866A.

[w7fox]

OM,

Regarding the 807 transmitter and modulator, your transformer is going to give you about 800 volts with a choke input circuit.  This is too much for the 807, which wants 600 volts max when plate modulated.  I am running mine at around 600 volts and it works fine.  On one occasion, my input choke shorted, and before I could figure out why my power output was way up, both finals fried.  I am also using a modulation transformer similar to the art13 with a pair of 807s as modulators.  I am running 700 volts on the modulators in AB2, which actually gives me more audio than I need.  Its nice to have a little extra.  One thing about the modulation transformer, I scrapped one from an art13, and another from a later piece of surplus that also had an 813 modulated by 811s.  These transformers have a separate winding for the screen of the 813.  The newer transformer put out more audio on this winding, so I used it in my rig.  It works fine, but I have never tried the art13 transformer to compare performance.  If you try it with yours and don't get enough audio to the screen, I would feed the screen through a resistor as is ususually done.  Good luck with your project.
73,
Chris W7FOX

[W3GMS]

Sounds like a fun project. 

If I was doing it, I would tend to use just a single supply providing you have a filter choke that is rated for the total current you need to deliver for the final and modulator.  With the load of the class C final on the supply, you should have decent regulation for the modulators.  Yes, the voltage even with the choke input filter is going to be a bit high.  I would use a small variac in the primary to reduce the voltage.  The nice thing about have a common supply is when you reduce the voltage on the final your reducing both the voltage and current so the plate impedance remains nearly the same.  The other alternative would be use a filament transformer in the primary as a bucking transformer to reduce the secondary voltage. 

Good luck with the project!

73,
Joe-W3GMS         

[W9ZSL]

Here's the deal Joe.  I have a nice UTC tranny that I used to build a solid-state supply with a bridge.  The iron is good for about 300 MA easy.  See the pix.  I already experimented using this supply on a parallel 5933 amp and the voltage dropped to about 625.  My question involves splitting the output into two separate branches as shown in my diagram but substituting a bridge for the National pair of rectifiers and using one power transformer.

My diagram WAS theoretical based on obtaining the TX-1 plate iron.  Practically speaking, rebuilding the supply shown makes more sense PROVIDED THAT THE VOLTAGE CAN BE DIVIDED INTO TWO FILTER NETS.  That is the basic question.

Is my voltage divider using two separate filters as shown realistic?

This supply can be re-built.  As for the TX-1 iron, it clearly would be overkill.

[W9ZSL]

In my diagram, substitute the TX-1 iron with a UTC transformer built for a bridge that puts out the 775 volts as shown with a single choke input supply.  I propose to split the output into two chokes and two filters one with a swinger and the other with a smoothing each rated at 200MA.  The chokes are matched Thordarsons: 19C35 and 19C42.  Gotta use them for something!  Caps can be whatever.  I have a slew of 100K resistors at high wattage that can be configured to whatever works.  They are the small aluminum jobs.  This has variable screen and bias supplies.

Main question; will the voltage divider I suggest work? 

[w1vtp]

I think redoing the PS so that you had the swinging choke first and then the smoothing choke.  Not sure about the value of the caps.  That would bring the output down a bit.  Not sure if it would be enough.  I think the 20 MFD cap values might be a bit high - more like 8 to 10.  Just a single filter system feeding everything

Just a thought

Al

[KA2DZT]

In my diagram, substitute the TX-1 iron with a UTC transformer built for a bridge that puts out the 775 volts as shown with a single choke input supply.  I propose to split the output into two chokes and two filters one with a swinger and the other with a smoothing each rated at 200MA.  The chokes are matched Thordarsons: 19C35 and 19C42.  Gotta use them for something!  Caps can be whatever.  I have a slew of 100K resistors at high wattage that can be configured to whatever works.  They are the small aluminum jobs.  This has variable screen and bias supplies.

Main question; will the voltage divider I suggest work? 

Mike,

The dual supply you are suggesting will work OK FB.  I've used that type of setup and there shouldn't be any problem.  But, you need to reduce the overall voltage output to about 700 VDC for the 807s modulators and even less than 700 VDC for the RF PA.

Fred

[nq5t]

I can't help you with the PS on this.  But this amp was one of the (unfulfilled) projects I dreamed about building as an early teenager.  Hope you get it built and running ….

Grant NQ5T

[W3GMS]

Mike,

Your scheme will work, its just a question of is it necessary based on the rating of your filter components.   My thought process would go something like this: 

1) Access the current for the final and the current for the modulators.   It that current is more than a single filter network can handle than split it up like you proposed. 

2) Based on the transformer you have, determine if a choke input filter or a cap input filter will deliver the voltage that you require.  Keep in mind that you have to satisfy the magnetizing current of the filter choke otherwise your choke input filter will look like a cap input filter and as a result your voltage will likely be way to high and your no load to full load regulation will be bad.  Sizing your bleeder resistor properly will guarantee that will not be a problem.  If you had a common filter scheme since your final steady state current would guarantee that enough current would be passing through the choke. 

ie, Swinging chokes do have an advantage for loads that are dynamic. 

3)  When you have voltage droop, its important to know what is causing the droop.  If the secondary current of the transformer cannot handle the total load current, then it will droop beyond what you expect.  Chokes beyond their ratings will saturate which will dramatically effect regulation since your running out of inductance. 

As always, many schemes will work and your proposal will definitely work.  Its more of a question if its necessary for your application and that answer lies in the filter components and your application, since the main power transformer is a common item. 

I have used two separate rectifier assemblies off a common transformer along with their individual filter networks for the purpose of delivering two separate and different voltages.  I would use a cap  input filter off one rectifier to feed the modulators.  I would then use a choke input filter on the second filter network.  So the final would have less voltage on it compared to what was on the modulators.  This helps  when the turns ratio of the modulation transformer is not sized properly making 100% modulation challenging on a stock design.  This higher voltage on the modulator stage allows a greater peak to peak voltage going to the final and allowed 100% modulation.   

So there you go!

Joe-GMS                   

[W9ZSL]

The main reason I want to split the supply into two separate filters is because I have a pair of 200 MA chokes.  Neither the RF or Modulators should pull more than 200 MA each.  I also have some rated at 350 and 400 MA but I'm saving them for another project...813 / 811A with two separate supplies.  This one with 5933 bottles will give me some practical experience in an amp build before I try anything bigger.  As for the filter caps, I can use what ever works best.  That supply with the UTC "as-is" works really great and is super-quiet.  As mentioned, I already used it with a similar amp using parallel 5933s and the voltage dropped to about 625 DC under load.