[R-390] Resistors for the R-390A and beginner help

[Bob Camp]

Hi,

Well we'll save the PTO stuff for another time. Most of it has been
discussed here or in the magazines.

There the *theory* on the resistors:

First you grab your handy copy of Mil-Hbk-217 and look up reliability of
carbon comp resistors. The interesting little note there mentions that the
entire failure rate listed is for value change. Hmmm so if I can tolerate a
larger value change I don't have to count the failures of the resistors in
my MTBF calculation you say. That's what it says. Official DOD handbook, the
note's been there forever and ever.

Next you dig back into the lore of carbon comp's and find that the values
never did stay put very well.

This all in mind we take a look at some circuits:

I started to do this in detail on the front end of the radio and it quickly
got far more difficult to explain than it's probably worth. Here are a
couple of examples:

The resistors off the AGC line going to the grids of the tubes form an
attenuator with a ratio of say 80%. In other words one resistor is 1/5 the
size of the other one and you get 8 volts on the grid for every 10 volts of
AGC voltage. Let's say that one resistor is 270K and the other is 1.5 meg
ohm. The ratio in this case should be 1500/(270+1500) = 0.85.
If both resistors go up by 50% the ratio stays the same and nothing much
happens. If only one goes up by 50% you get 1500/((270*1.5)+1500) = 0.79.
That's not much of a change in the attenuator. Now you get about 7% less AGC
voltage on the tube. Given that you will see >20% variation between new
tubes 7% isn't going to hurt anything.

Next you have decoupling resistors. They are the first thing off of the B+
line going to the screen or to the plate. Most of them only have a couple of
volts drop under normal conditions. The B+ will vary by 20% or so as the
line voltage swings. In order for any of the decoupling resistors to start
messing up things they would have to increase their drop by say 30 or 40
volts. That's a lot of change for a resistor that starts out with a <10 volt
drop on it.

Next up are the grid bias resistors. Most of them seem to be a meg or so.
Since grid current should be darn near zero ua they should have < 1 volt on
them. A doubling in the resistor *might* get them up to a volt on a good
day. Go down to the tube store and ask for a set of tubes matched to < 1
volt on the grid. Have your American Express card ready :)

Now for the cathode bias resistors. These do set the stage current,
especially on the triode stages. Cut down the current by a factor of two and
the stage gain will drop a bit. Age the tube for a couple of years and the
stage gain will do the same thing. Cutting the current here isn't a good
thing but it happens anyway.  The function of gain to resistor value is a
little complex but at least you can say that the resistor has to more than
double to cut the current in half. A perfectly normal set of tubes out of
the same batch will spread 1.5:1 on idle current. Trying to get things any
closer than this can be a pain. The audio guys go to a lot of trouble on
that sort of thing.

I realize that does not cover every resistor in every circuit in the radio.
It should give you a pretty good idea what is going on though.

Let me know if any of it makes any sense.

    Take Care!

        Bob Camp
        KB8TQ









----- Original Message -----
From: "Tom Warren" <wwarren1@nc.rr.com>
To: "Bob Camp" <ham@cq.nu>; <r-390@mailman.qth.net>
Sent: Tuesday, January 14, 2003 7:04 AM
Subject: Re: [R-390] Resistors for the R-390A and beginner help


> So Bob (actually, I'm serious), how about telling us about PTOs and
Collins'
> troubles in getting them designed.  I've seen one or two things, but
haven't
> found the articles where the Collins boys talk about the drifting
> permeability of the iron oxide (or is it ferrite?) cores and the shrinking
> coil forms.  Where is this all discussed.
>
> It'd be nice to hear a bit more discourse on why the radio can stand a
wider
> tolerance band than is specified.  Except for about 7-10 places (excluding