RCA BTA-1R1

[K8DI]

I've been working towards a heavy metal rig, and there's one finally standing up in the garage.  I picked this up two weeks ago, but I needed to clear a spot for it, and then install some castors so I could move it out from the wall to work on it.  I haven't gotten the iron or tubes back in nor the doors on, but it's here!

A couple thoughts, I wonder what the collective opinions are:

This is a solid state rectifier setup. Its a FW bridge, using a diode rated at 400PIV, 750mA, 1.2v drop at 750mA.  There are 18 of them in series, with balancing resistors, on each card. There's also a resistor in series. I don't know if that was to limit inrush, or to balance out the cards.  This transmitter model was designed with ConElRad in mind, so it would end up being subject to rapid power cycling.  Each leg of the bridge has two cards in parallel, which makes sense in that at 100% modulation, the total of mod and RF plate current is specified at 945mA. 

I have done no testing yet, but if I find any issues, I am inclined to strip the boards and replace the 18 diodes (well, 36 because there's two per leg) with 8 of 1N5408 3 amp, 1000PIV diodes.  The current rating doubles, the total PIV is about the same, and the forward drop is going to be less than half.  Note that there is a power up/down control on this transmitter that varies a series resistance in the B+ line, so the final plate voltage can be adjusted back to spec.

At the end of the day, when this thing is converted, I intend to key the plate power for PTT. I realize I'll have to deal with the oscillator and driver sections as well, but those voltages and currents are much more manageable. Whether it's the primary or the secondary of the plate transformer that gets switched remains to be decided.
With all that in mind:

1) am I unwise to swap the diode string out entirely, vs replacing only what is blown?
2) What about the series resistor?
3) Any immediate cons to switching plate power, vs removing RF drive and switching on some bias to keep the finals happy (grid leak design/no bias on finals)?

I am some distance away from having to decide on all this. I have some cleanup to do, iron to install, check a thousand screws for tightness, and a 240v circuit to run, before I can even turn it on.  But, I like to plan and decide before I start working....


Ed

[W7TFO]

Keying the HV is a popular way of PTT, but I highly advise you just run that on a timer set (say 10 to20 minutes) from the last activation.

To key along with that, do the drive.  It keeps things much happier without all the power jumping around with HV keying.

Normal bias will hold the finals in low current whilst waiting for the command, and the RF input will set things in action without fail.

73DG

[Opcom]

To try to adderss the three questions,

In the picture it shows eight boards 1Z401-1Z408, each with 18 rectifiers. These are in series-parallel for the bridge. Each rectifier has a 27K resistor across it for old-rectifier-technology reasons, but it cost nothing to keep them all on general principles when many diodes are in series, unless there's an objection.

Looking at rectifer board assemblies 1Z401-1Z408 in the schematic, each of these also has a 20 ohm resistor in series with its diodes at the (+) side, which is there to balance current among the paralled boards. It is also an old school thing, but as well to keep them unless there's an objection.

The total resistance with four 20 Ohm resistors in series parallel each half cycle is 20 Ohms which won't meaningfully affect inrush current, but improves reliability in case of a shorted diode. If a diode on one board shorts and therefore that board's voltage drop is lower than the other, the 20 Ohm resistor will try to maintain reliability by reducing the current imbalance between parallelled rectifier boards.

The rectifier feeds 1L502 the HV choke for excellent regulation. The value is not stated, but whenever there is a choke input filter and HV transformer primary is keyed, there is a transient voltage pulse and accompanying current pulse. The higher the choke inductance, the larger these pulses become. Adding more filter capacitance after the choke lengthens the pulse. The transient can approach or exceed voltage ratings of other components.

Since you have 18 positions available on each board, it would be better to populate them all when going from 400V to 1KV diodes. The increase in PIV from 7200V per board to 18KV per board is an excellent safety margin not only for help against HV keying transients but also surges, lighting, power cuts, etc.

The max lead diameter 1.32mm on the 1N5408s may be larger than the originals. Drillijng-reaming out the holes may or may not be advisable, you have to look at the backside of the boards. It could be as well to use 1N4007s (max 0.864mm) if they fit the holes since they are 1A instead of 750mA. If the originals are good, why not leave them and see how they do?

I don't know that transmitter and only looked at the manual, but it was intended to be turned on and left on instead of HV-keyed, so a simple inrush resistor & delay of 200-400ms would be a fine addition whether upgrading the rectifiers or not.

[KA3EKH]

I ran a BTA-1M that I converted for 160 on 1885 for years and many contacts, all I did for keying was to keep the original filament system in place for standby and use the external interlock circuit that interrupts the plat transformer primary to control the push to talk circuit.
Did end up using a time delay relay so the antenna would change over a fraction of a second before HV was applied and another time delay to keep the antenna applied for a split second after the PTT was released because with keying the HV some energy still stays in the filters and tank so if you don’t include delays bad things may happen.
Another thing is you may want to consider taking out all of the series resistance junk in the HV feed that was used for low power operation and throwing it away, if you want to run reduced power you can always feed the HV transformer on 120 vols instead of 240 or what I did was install a 240 volt autotransformer ahead of the HV transformer and used that to control power. Think I had my transmitter set for around 375 watts of carrier.
As you have already seen there is no protective bias and that transmitter was never designed to be operated with HV applied and no drive, unless you want to cook your tubes.
Also check to see if the modulation transformer is on wooden blocks, an old broadcasting trick on how to keep using that transformer after flash over to the core. Don’t think I have ever seen a R that came out of service that was not like that. Been told by old timers that back in the day with simple processing the transformer worked just fin but by the seventies and eighties all the management wanted real aggressive “LOUD” sound and processing was such that is was not uncommon to run 125% positive peaks and those old transformers were the first thing to start failing by flashing over to the core and throwing a plat trip. So, the answer was to place it on a couple two by fours and that allowed you to continue to flog the modulation.
I never saw a BTA in service for anything except a backup when I was working but did remove and repair several that were used in that function but the majority of all that stuff is now long gone from the industry. My M used 833 tubes in the modulator and PA and there were enough RCA and Gates transmitters still in service that 833 were still in production as late as the nineties and may still be being produced somewhere although do not recall seeing any of those transmitters as backups this century.

AA link to a web page that I did for my BTA-1M:

http://staff.salisbury.edu/~rafantini/RCABTA.htm

[K8DI]

Thanks for the replies. 

This model depends on grid drive to bias the finals.  Creating a bias supply would not be an issue, though.  Question— for the purpose of keeping the tubes under control, how much bias?  The mods run -400 and are in AB1. The finals are in C, and with drive, run -115 to -150 depending on output power/configuration.  More bias would reduce standby plate current, but would it need to be switched off under drive?

Has anybody built a soft start or T/R switching setup that uses a chopper/dimmer type arrangement to fade to plates up and down over a short (~300mS) time? There’s plenty of high powered SCRs one could use, the hard part (to me) would be the drive circuit. However, this would seem a very slick way of “switching” plates on and off…

I’m also thinking, with the various plate transformer taps, I should be able to use some combo to get the HV in between the nominal 250 and 500 watt levels. 1928v would be the calculated B+ for 375w carrier. Not sure how close I can get to that, but it’s a goal.

Ed

[KA3EKH]

Its your transmitter and I am not telling you what to do but for me I cannot see ant advantage to keeping the HV applied and building in some form of bias and all that when all you have to do is just use the external Plate ON/OFF control for a PTT line? That way you turn the transmitter on and after the time delay for the plate circuit any time you close the plate control line it transmits, when you open it its off. There is no delay or anything like that, I used this method for years and have had no issues as where if you want to start building up bias and all that sort of stuff that will work but you will have an issue where the MO will be running and that will be picked up by your receiver and like I said, I cant see any advantage to keeping the HV up when you are not transmitting. Except maybe increasing your electric bill.
Maybe if you are planning on using an external radio as a MO so you can do things like SSB but you will find that the tank for the driver and PA don’t like to move much without doing a lot of retuning, if you are going to use the original MO and a crystal what advantage are you getting by keeping the HV applied when not transmitting?
On that transmitter they had a set of plate contactors for Day and Night power so when the transmitter is in the night mode it can be strapped for 250 watts without much issue. That was one of the big improvements of the R over the M where in the M they just inserted lots of resistance after the HV supply to drop the power so the transmitter just wasted a lot of energy making heat. They fixed that problem in the R using two different plate contactors one for Day and high power and one for Night with different taps on the plate transformer for low power.

[K8DI]

I haven’t decided to NOT just use plate switching, just making sure I’ve thought it all through. Once I’ve powered it up I’ll be inspecting the contractor contact condition, too. If the thing is worn down then using it for PTT, it may not last long…

And, yes, the osc will have to be killed during receive. 

Ed