The AM Forum

THE AM BULLETIN BOARD => Technical Forum => Topic started by: KI4YAN on July 26, 2020, 11:32:07 PM



Title: RF Deck layout thoughts
Post by: KI4YAN on July 26, 2020, 11:32:07 PM
I started laying out the RF deck for my 4-65A transmitter tonight. Before cutting metal, I set parts on the top and tried to figure out how I might place them all.

(https://i.imgur.com/8Us0Ndu.jpg)

The idea is (as you might see) to make sure that in the future, I can swap the socket, the filament transformer, and the 4-65A out for a 4-125A later on if  I get into that position. The cutout for the socket and the filament transformer is identical, so chassis rework would be minimal.

I have this single capacitor that's 114pF, and spaced 0.070" between the rotor and stator, and I have a capacitor that will take up the entire section marked off there and is 120pF in two sections, with 0.180" between rotor and stator plates. Either cap has enough capacity for 80 to 10m with the 4-65A or 4-125A.

As for the loading capacitor, I have a receiving type variable that's a 365pF+365pF dual section. That is only 30pF more than the needed amount for an 80M match to 50 ohms.

Looks like 160M is not feasible without switching in capacitors on both variables AND inductance. That's alright. I can be satisfied with 80M-40M-20M-10M.

Here's what I am proposing for a general layout-things are fairly tight on the top deck. I moved the 4-65A tube from dead-center over a bit, centered up on the X. I can jog it forward a little, towards the front of the chassis, to make more room for the plate choke.

(https://i.imgur.com/dwlJ0iz.jpg)

You see on the far right hand side, I have cordoned off a section for the 2E26 driver. Input to the driver's grid will be a passive 50R input, and will be driven by a DDS-generated source. The output of the 2E26 only needs to be around 5W. I figure a switched-parallel plate tank for the 2E26, with the passive grid, would be enough to do the job. I plan to make the parallel L-C tanks around a pair of switch wafers.

Does anyone see any major, layout-changing failures in this plan?


Title: Re: RF Deck layout thoughts
Post by: KI4YAN on July 27, 2020, 02:49:03 AM
This is the basic idea of how I plan to implement the RF deck. The 2E26 is plate-keyed and the 4-65A is screen-grid keyed. Both the 4-65A and the 4-125A are suitable for this method of keying up to 2000 volts, so it won't have to change if I later on change tubes.

(https://i.imgur.com/h5H8Go2.png)

The "Bandpass Tank" block is a pair of rotary switch wafers that will swap out the tank components, and the output pi-tank is the normal switched inductor and variable air capacitors.

Some cathode bias is intended on the 2E26, and a bit of grid leak bias is intended as well. Drive will be about 200mW going in, so I think it'll mostly be cathode bias.

The excitation control varies the screen voltage on the 2E26, which should provide good control of the grid drive-about 2.5W will be needed to drive the final.

I'm sure it'll burst into flames the first few power-ups, but hopefully I can minimize that.


Title: Re: RF Deck layout thoughts
Post by: ka1tdq on July 30, 2020, 12:20:34 AM
Here's a 4-65A transmitter I built in 2010. My wife had her first AM QSO on it. I wish I still had that video. I sent a CD-ROM of the video to the ham in Mass who was talking with her. I forgot his call, but if you're out there could you dropbox it to me?

Anyway, tubes are fairly forgiving to layout, especially at 75 meters, unless leads get really long. What's more critical is design. My tube lineup was a 5763 -> 7984 -> 4-65A. I probably didn't need the 7984 in the middle but I was learning. It was modulated by 6L6's or KT-88's (take your pick).

There's a million ways to do things and this was my path. I hope this helps.

Jon


Title: Re: RF Deck layout thoughts
Post by: KI4YAN on July 30, 2020, 08:29:17 PM
Layout continues. I'll be building a plate-circut box for the 2E26, to make sure that the plate circuit of the 4-65/125 doesn't see the plate circuit components of the 2E26, and also that the grid circuits for both tubes are segregated as well. That's 4 compartments plus the output tank circuit, so 3 divider panels and a screened box.

Under chassis:

Divider to section off the 2E26 grid circuitry and the drive level control
Divider to cordon off the 4-65/125 socket and filament transformer so that a small fan can pressurize the chassis and blow up through the socket

Above Chassis:

Divider to box in the 2E26 tube and plate circuit, solid aluminum panel between the rest of the chassis and a screened aluminum panel to the top and outside for airflow.

Front Panel Layout:

Above Chassis:

Plate Tune and Plate Bandswitch on left side
Tube Monitoring Window Middle-Right
Grid Bandswitch on Right Side

Inside Chassis:

Loading Control Right Side
Drive Level control Left Side

Below Chassis:

Plate Current Meter, Filament Voltage Meter, and Grid Current Meter

I'd move the meters up into the chassis but they're too big to fit-they're the Simpson model 27 that require a 3" diameter cutout. I had one that took a 1.75" cutout but I can't find any more like it that are appropriately scaled, and I have a 250DCmA, 10VAC, and just need to find a 50DCmA to replace the 50DCuA meter I have now.

All measured, the front panel will need to be 7 rack-units tall, or 12.25".
AMfone - Dedicated to Amplitude Modulation on the Amateur Radio Bands