Warning: Creating default object from empty value in /homepages/11/d132647312/htdocs/Amfone/mkportal/include/SMF/smf_out.php on line 47
Musings on bias in converted BC rigs, especially grid leak for 4-400A finals

The AM Forum
August 11, 2022, 05:20:14 AM *
Welcome, Guest. Please login or register.

Login with username, password and session length
   Home   Help Calendar Links Staff List Gallery Login Register  
Pages: [1]   Go Down
Author Topic: Musings on bias in converted BC rigs, especially grid leak for 4-400A finals  (Read 435 times)
0 Members and 1 Guest are viewing this topic.
Offline Offline

Posts: 310

« on: May 17, 2022, 09:45:52 AM »

Note:  all currents are quoted in terms of the thing being discussed. That is, if there are two finals with 10mA grid current each in the transmitter, the transmitter grid current will be stated as 20mA. If I mean a single tube (inside or outside of the transmitter), it will be stated as “per tube”.

I’m about done with converting an RCA BTA-1R1 to be used on 160, 80, and 40m. The method I used was tested on 20m as well, but my band-switching arrangement doesn’t include it.  The setup is to use a modern radio as a driver (in CW mode, grid drive) and receiver (in AM mode).  I set up the TR switch such that the radio is connected to the grids in transmit, and the antenna in receive. In my case the TR switch also connects the transmitter to the antenna in transmit, and grounds it in receive.  In the process, I had many conversations with others on the AMFone forum, on FaceBook AM groups, and with individual hams, an RF/EMC engineer friend, and a couple different retired broadcast engineers.  The best part of all of the talk, and the tests and trials, was the learning.  Striving to know how things work has always been a passion, and is part of the reason I enjoy radio.

First, some numbers to compare from manuals and data sheets:

RCA BTA-1R series (a pair of 4-400A finals):

Plate voltage 3000v, current 460mA, input power 1380w
Output power 1000w, efficiency (including tank/filters) 72%
Grid leak 20k, grid current 19.8mA, for a grid voltage of 396v.

Extremely similar numbers apply to Bauer and other broadcast transmitters. They all run 3kV, 20mA on the grids, and put out 1kW.

Eimac 4-400A datasheet values (one tube at 3kV, CCS /ICAS rating are the same at 3kV):

Plate voltage 3000v, current 275mA, input power 825w
Output power 630w, efficiency (tube alone/no tank) 77%
Grid current 12mA, grid voltage 220v, which would work out to 18.3k for one tube, 9.16k for a pair.

By the datasheet, a pair of 4-400A’s should do 1260w carrier.  RCA and others chose to run bias much higher. What happens when you increase the grid resistance?  Well, the voltage necessary to get the same current climbs. More voltage, means more grid drive. Transmitters like the BTA-1R series had no issue making drive – they needed about 7 watts, and RCA used a class C 6146 to drive the grids, good for north of 50 watts, and delivering about 20w in their conservative circuit.  What else happens?  The output power drops. This sounds like a bad thing, but it has a purpose, once you understand how big transmitting tubes wear out.  In a properly running transmitter, they don’t over heat, they don’t burn out, they lose emission.  What does that mean?  It means that the total amount of electrons coming off the filament isn’t enough any more.  With a plate modulated class C amplifier, there has to be enough emission to support double the standing current for 100% positive modulation to occur. As the tube ages, fewer and fewer electrons are available, and it does not matter what the bias, or plate voltage, or anything else is, there’s simply a maximum rate at which electrons are emitted, and when you exceed it, the tube can’t pass any higher current.  I have a pair of completely used up 4-400A’s on hand…with 3kV on the plates, and NO bias at all, they run under 200mA plate current per tube! They still light up, they still make RF (albeit at 1/10 rated power). They just don’t have enough emission.
So, what RCA and others did, was raise the bias, lower the power, and in doing so, extend the time the transmitter would make rated power on a set of tubes.  In the commercial broadcast industry, tube life was an inescapable cost – they were always expensive, and you were off the air without them….

In experimenting and talking with others, there ended up quite a bit of debate between RCA’s 400v bias, and the datasheet’s 220v bias.   I tried both. At 220v, I got higher plate current and 1250 watts, at 400v, I got RCA-spec plate current and 1000w.  I also found, when using direct grid drive (as I am doing) there are issues at higher frequencies with the higher grid drive requirement.
As we should know/have realized, a given tube in a given circuit will have progressively less output at higher frequencies. We no longer have drive to spare, we have to peak the grid circuit more precisely, etc. The same holds true for a 4-400A. It can run at 75MHz. But it is going to take all the drive you can stick into it up to the grid dissipation limit.  The broadcast rig is dealing with about 1MHz. it needs very little drive. So, what happens when you go with the RCA design and try to get the needed bias current at higher frequencies?  You have to push the drive up, eventually past the grid limits. RCA’s design works fine with fresh tubes and plenty of emission on BC or 160m. It works ok on 80m. It pushes the tubes and grid circuit elements really hard on 40m. On 20m, you are well beyond the grid dissipation limits.  If you use the datasheet grid voltage, it is easy to drive the tubes within the grid limits all the way up to 20m. Maybe higher as well – but I did not test beyond 20m, as my output network can’t adjust to any lower values to support higher frequencies.

In the end, my conversion uses neither of the values.  I am using a 10k grid leak, a little above the data sheet, with 4-400A tubes. It was convenient; I had two of the original stock 20k resistors, so I simply paralleled them.  I also ended up adding an RFC in line (as one typically sees in a grid leak bias circuit). RCA depended on the resistor's natural inductance; it is fine at BC frequencies, but can have self-resonance elsewhere, so I made sure there wasn't any RF on the grid leak.  

For the time being, I will be using 4-250A tubes instead. Eventually, I will get my hands on a good pair of 4-400’s and go back to them but all of mine are a bit weak. The 4-250A’s call for a different grid leak, and will drop the actual possible power level to be right at 1kW, vs. the possible level with 4-400A’s of 1260w. But, being new and having plenty of emission, they’ll make 100+% positive modulation – the old 4-400A’s I have barely make 75% positive.

I wrote this up to share my experiences and what I learned. I realize there are a bunch of more knowledgeable people out there. If you see an error, please let me know – I’d like to get it right.


Ed, K8DI, warming the air with RF, and working on lighting the shack with thoriated tungsten and mercury vapor...

Offline Offline

Posts: 181

« Reply #1 on: May 17, 2022, 10:49:34 AM »

I am certainly not more knowledgable, but I am posting to compliment your thoughtful and systematic testing and analysis of the issues that one meets when building a transmitter with tubes that may have been pulled from service while still showing themselves to be “good” and also the challenges of the adaptation of a mw design to transmit at short wave frequencies.

Of special interest is the point about the reserve capacity of the tube to provide peak modulation envelope power in a high level plate modulated circuit.

If one put those used 4-400As in a linear amplifier and drove them to maximum output, the peak power available would be readily available when tuning up and likely show a lower Peak carrier at that time.
In Class C, the carrier could appear to be ok, but the peak envelope power capability would be unknown unless modulation power tests were done with peak reading watt meters and also observation of modulation distortion would also be unknown with just carrier available.
Very interesting research. I look forward to further comments from real experts in the field. We have a number on this forum that I know can provide further explanation.
Offline Offline

Posts: 8768

"Let's go kayaking, Tommy!" - Yaz

« Reply #2 on: May 17, 2022, 11:10:03 AM »

A very interesting summary, Ed!

It builds a good case for testing tubes at the highest frequency available in your rig rather than at a lower one to show emission.  I understand dropping the filament voltage a small amount will also help expose fil/cathodes that are on the edge of acceptable.

When I have a bunch of similar tubes to test, I load up a good "known" tube to maximum parameters and then use tone tests (AM mode) to find how well it handles peak power and where it flattops.  Make a list of the tubes and max parameter performance and it will tell you the best to worst tubes in the harem.  

Congrats on completing the 4-400 BC rig.

I'm in the tower climbing mode right now, putting up and rebuilding some new antennas for 75 and 40M.  Also adding 10-20M arrays to take advantage of this questionable solar cycle. Yes, it's not easy to get the minimum L/C tuning values on 10-20M from a 160-40M rig.

So hope to catch ya on soon.


Use an "AM Courtesy Filter" to limit transmit audio bandwidth  +-4.5 KHz, +-6.0 KHz or +-8.0 KHz when needed.  Easily done in DSP.

Wise Words : "I'm as old as I've ever been... and I'm as young as I'll ever be."

There's nothing like an old dog.
Pages: [1]   Go Up
Jump to:  

AMfone - Dedicated to Amplitude Modulation on the Amateur Radio Bands
 AMfone © 2001-2015
Powered by SMF 1.1.21 | SMF © 2015, Simple Machines
Page created in 0.051 seconds with 18 queries.