[R-390] R-390 > R-725 IF filament conversion

[Gary Gitzen]

Hi R-390 denizens,

Pete forced my hand on this one. I was asleep at the switch.


The following words are musings and ramblings about
replacing the IF deck in an R-390A with the deck from
an R-390; in effect recreating the R-725/URR.
It mostly focuses on the widely held belief that major
IF deck filament rewiring is needed for this change.
If this doesn't interest you, please punt now.



I recently acquired an R-390 IF deck (thanks, Pete), with
the intent of converting and installing it in one of my
two R-390A mainframes. This would give me one Rx with the
mechanical filters, and another with L-C filters.

Upon inspection of the "new" IF deck, I quickly saw
why I'd been warned that "it gets real tight in there"
when rewiring the filament string for 6.3V operation.
Not a project to be undertaken without some careful
thought and planning.

"Why," I asked myself, "does it need to be rewired?"
And I answered myself with "Everybody knows, dummy, 
that the R-390A power transformer doesn't have enough
25.2V current to power the non-A deck in 25.2V mode.
You read Tom Marcotte's excellent writeup on "rolling
your own R-725" with a non-A IF deck on his web page,
and others have written about this issue. Shut up,
get out your soldering iron, and start working!"

But since "everybody knows" has always triggered my
curiosity, I thought I'd take a better look.
I came up with the following 25.2V fil requirements
for the non-A IF deck:
	.15A	4X 6BJ6
	.15A	2X 12AU7
	.15A	2X 6BJ6 + 1X 12AU7
	.3A	3TF7 + PTO & BFO 6BA6s
       =.75A @ 25.2V
Additionally, .15A at 6.3V  1X 6BJ6, needed from somewhere.

Yeah, I thought, that .75A might overload the 25.2V winding
on the xfmr. 
But wait a minute: the 390A also has a 3TF7 and its
.3A load in its IF deck, so we can eliminate that as
an "extra power needed" item.
That leaves us with .75A - .3A = .45A of "extra" 25.2V
current needed for the non-A IF deck.

Then I started looking closer at actual numbers, and
how I configure my R-390As.

First, I never use 3TF7s. I replace the PTO & BFO
6BA6s with 12BA6s, which reduces the 25.2V filament drain
by another .15A. It's now down to .3A "extra" needed.
Second, I don't use 26Z5s. I replace them with soiled
state diodes. 26Z5s have a filament current of .3A each,
or .6A for two.
Hmmmmm......... That's twice the "extra" filament power
I now need for that non-A IF deck. I appear to have .3A
of "extra" 25.2V available. I may not need to
follow the rewiring instructions after all.......

I took a look at the power transformer. The 25.2V winding
is rated 1.2A continuous, plus 4.6A at a 70% duty cycle
for the ovens. But I never use mine in an Alaskan igloo
with the ovens on. I turn the ovens off, as do most owners
of R-390As. This makes that 70% of 4.6A available;
about 3.22A. With that other "extra" .3A, I have over 3.5A
of "excess" 25.2V available. 

Conclusions: Simply replacing the 26Z5s with silicon
diodes frees up enough 25.2V current to power a non-A
IF deck in an R-390A. Even without that change, the total
25.2V load appears to be easily handled, within spec of
the transformer, unless it is operating at minimum rated
temperature with the ovens on. Turning off the ovens
frees up more than enough 25.2V current, even with 26Z5s
and 3TF7/6BA6s.

So where did this "everyone knows" come from?
Tom Marcotte appears not to know. He simply duplicated
the original R-725 mod as closely as he could.
My _suspicion_ is that everyone just assumed it had
to be done that way.

The major benefits of not rewiring filaments for 6.3V
appear to be:
1: The risk of damaging the non-A IF deck is reduced.
2: If ever needed, it will be much easier to restore the
non-A IF deck so it can run in a non-A radio.
All in all, both of the above could have been a major PITA.

Question for the list: have I missed anything here?
Made any incorrect assumptions? If so, please let
me know before I do this. Or stick my foot deeper in
my mouth.

**************
"That was Yesterday"; now fast forward:

I exchanged mail with Pete Williams, David Wise, and
Tom Marcotte a while back, sending them an earlier version
of the above.

Since then, Pete has actually tried it, using an
adaptor/extension cable which flips pins 20 & 8
on the main connector. He reports good/excellent results.
I've taken a cursory look at the connector on the non-A
IF deck, and it looks possible to make the flip there.
I'll be trying that "in the near future". More info
as I think I know it, although Pete may beat me to it.

Pete also forwarded to me a scan of the filament wiring
schematic from hist R-390 manual. Thanks, Pete!

Tom M suggests adding a B+ dropping resistor inside the
non-A IF deck to reduce B+ to R-390 voltage.
Measurements and more research needed.

Thanks to Tom Marcotte for both the excellent R-725
documentation on his web site and for his encouraging
words re the above in private communication.

My compliments and thanks to Pete for once again
taking theoretical musings, doing further research
to identify the pin 8 & 20 flip, then actually doing
it on his bench. More kudos for caring enough to
take the time to report his findings.

Gary Gitzen
San Jose, CA