[R-390] BFO 455 Crystal and Z501 C520 C525 L501

[Flowertime01 at wmconnect.com]

Fellows,

I felt a need to break this mail down. I hate when it bounces against the 
size limit and need dispensation to get it forwarded.

Remember the first lamp read out counters hit the Army Field stations in the 
late 50's In 70 we had one nixie readout counter in Phu Bia. In 75 we had two 
counters in Okinawa. We did receiver after receiver every day around the world 
pushing a AN/URM 25 into those R390 crystals and called it good. The TM 
procedure was all we had. As bad as it reads, we read it until we understood what 
had to be done.

If you have a counter use it.

When you was doing two receivers a day every day you learn a lot quicker that 
when you only own one and do this once a year. 

After you get the IF deck aligned at 455 and before you jump over to the RF 
deck you need to finish up on the IF deck.

When you inject 455 into the IF deck through one of the RF to IF bnc 
connectors (J513) you are upsetting Z501 in the R390/A and its mate in the R390. You 
also need to set the BFO at 455Khz. If you are blessed with a frequency counter 
you do it that way. For every one else you push 455 through the crystal 
filter as a filter and zero beat against it. You cannot get 455 through the filter 
without some work. 

As you finish the IF deck and have it all cabled back to gather, reach into 
the RF deck and remove the third mixer. This will quite down the RF out put 
into the IF deck. 

There were comments last week about an adapter that had a 0.047 600 volt cap 
in it of use to the R390 community. And this is where it be used. 

Use the cap to stuff 455 into the plate pin of the third mixer (pin 1 of a 
6C4).

Again set the IF band width to 0.1 and peak the generator through the 455 
crystal in side Z501. If you have a frequency counter use it.

You can zero the BFO against the generator at this point.
 
Z501 is properly cabled and you can tweak the slug L501 for 1.0 band width.

You can adjust C520 to go with the crystal and you can adjust T208 in the RF 
deck.

If you do not know if your 0.1 crystal filter is centered and you do not have 
a counter, then you have to do the TM procedure to find the skirt points on 
the crystal. Then knowing the skirts (marks on the can) you set the crystal 
C520 center on the marks. This then becomes your reference for 455.

The TM says inject a small signal into the test point on the grid of the 
third mixer. This has all the RF deck noise mixing in with the generator and you 
are had put to get good reading on the diode load and audio meter for doing 
signal to noise ratios. You cannot determine the noisy tube in this TM setup. 
Pull the tube. Get the RF deck noise out of the mix.  The TM setup also makes it 
hard to find the peaks on the caps and crystals. The RF deck is working along 
just fine. It may have a noisy tube or tubes that you have not yet gotten to. 
So you do not know it is putting crud into the mix. You can go crazy because 
you have no idea what the RF deck is adding to your signal mix. The generator 
and the RF output are close about 455 you have no idea if you are peaking a RF 
deck noise or a signal generator frequency. You are adding two low level (RF 
deck and signal generator) signals and mixing them through a non linear mixing 
tube circuit. You wonder why you can make no sense of what is going on at the 
other end of the circuit in the IF deck. 

The TM says over drive the crystal by 60 DB. On a URM25 we just went up one 
switch setting. As long as you did both sides (above and below 455 ) with the 
same generator output level you find the marks for C520 (the 455 crystal trim 
cap)

So you find the best 455 pass through the crystal as you can get by rocking 
the generator. You bring the signal down for a readable meter (-5 volts on the 
diode load). Then you kick the generator up 60 DB. (what ever works on your 
equipment) one switch range or some meter level. Then you move the frequency up 
on the signal generator until it falls out of the crystal bandpass and the 
level through the bandpass is again at -7 volts. Trim C520 to dip the meter. mark 
the slot on the can.

Adjust the signal generator frequency under 455 until the meter reads -7 
volts. trim C520 to dip the meter. Mark the slot of C520 on the can. 

Set C520 to between the two marks.

If you have a frequency counter skip all the above.
Set the generator to 455. set the bandwidth to 0.1 Trim C520 to minimum. 
Never move C520 again in your life.

Setting C520 up to dip at 455 is a one time event. As you change band width 
over to 1, 2, 4, 8, and 16 you do not adjust C520. You do adjust L501 at the 
1Khz band width for peak like you adjust the mechanical filter or the IF cans in 
a R390 for their respective bandwidths. The peak of any bandwidth is not 
relative to other bandwidths.

You want the generator set to 455. How you get your test equipment set is 
dependent on your bench population. Back when R390's were new, the best reference 
to 455 was that crystal in the IF deck. The best way to get that test 
equipment item (the crystal) calibrated was to follow the TM setup procedure to find 
the skirts and set C520 to center. Later in Life we learned to not worry about 
the shock hazard (we was being shot at, a shock was just a wakeup call) and 
pulled the third mixer tube and inserted the signal into the plate pin of the 
RF deck. One can unplug the RF deck connector and then the third mixer plate 
pin does not have B+ on it. This works very well but it took time to undo and 
redo that connector. Works on any day you do not happen to have a cap to go on 
the end of your signal injector coax.


The TM then says to adjust L501 at 1. bandwidth for maximum at 455.

If you have a counter set the signal generator for 455. Set the band width to 
1 set the generator output for about -7 volts and adjust L501 for maximum DC 
load.

If you do not have a counter set the generator to peak through the crystal 
filter at 0.1 bandwidth. move the switch over to 1.0 bandwidth and adjust L501 
for peak. This is a one time pass. The C sets 0.1 while L sets 1.0 These two 
different bandwidths act very different. You tune them up separately.  

The messy TM process about finding the crystal skirts is just for 
"calibrating" the crystal to the best center on 455. Once you set that crystal it is good 
until you open the Z501 can and pet the crystal container. Once you get to 
where you believe the signal generator is set to 455 you just go forward with 
alignment. The 0.1 crystal was just the best way to get a signal generator set 
to 455 until the arrival of frequency counters. The messy TM procedure was just 
the best way to "calibrate" the crystal for use as a frequency reference. 
Still works good today once you understand what the outcome is supposed to be. 
Getting the RF deck noise out of the mix also helps thing go cleanly. 

You can get some higher peaks some where else than at exactly 455. That is 
not the objective. The objective is to get the 0.1 crystal balanced at 455. and 
1.0 inductor into the centered on 455.

Get the signal generator as close to 455 as you can get. If you have a 
counter use it. If not you peak the generator into the bandpass of the 455 crystal 
and know your within a 100 hertz of 455. If you do the procedure above you have 
a 455 crystal filter that is within a 100 hertz. You likely only diddle the 
signal generator into the counter down to within 20 or 30 hertz just because 
you can.

Pull the third mixer tube and inject into the plate pin of the socket. Then 
you can set the bandpass filters at peak without changing frequency. Most days 
the R390/A get their mechanical filters peaked up while we are doing the 
signal to noise testing of the IF deck. The book says 150 UV at 455 into J513. We 
skip the matching stuff because 50 years has shown it does not change the 
outcome. A lot of work goes into getting a good 30:1 signal to noise ratio out of 
the IF and audio deck. The R390 IF deck gets its transformers peaked and the 
R390/A gets its mechanical filters peaked.

Once we move over to putting 455 into the plate of the third mixer to set the 
crystal filter, L 501 peak and the last 455 output transformer of the RF deck 
we do not go back and reset the mechanical filter caps. We done been there 
and done that. 
We are trying to get out of here on time today. We are not going back to trim 
that stuff up once more. There is a long ways to go in the RF deck before we 
are done.

But if you have been working without a frequency counter, You are now pushing 
the best 455 frequency you can set up on your bench at this point. A peak is 
a peak. You could see enough improvement to make it worth your while to go 
back and trim every thing in the IF deck up again with this test setup. Just 
leave the IF Gain Adjust alone. Besides you likely will use Chuck Ripple's 
procedure to set the IF Gain Adjust as you place the receiver back in service (hook 
it to an antenna). 

If I did not spell Chuck's name correctly, please let me know. At least put 
his web page address up for everyone. Thank You.

But while you are here and have every thing peaked. See how much signal it 
takes to get-7 volts on the diode load and write it down. Then next time you 
need to do your receiver you can just run the generator into the plate pin of the 
third mixer and do everything in the IF with this setup.

We could not do this for show and tell in the shop because we had no way to 
validate that what we were driving the receiver with was equal. The show and 
tell signal to noise process had to be repeatable and by the book. The book does 
not say 150 UV into the IF and 30:1 signal to noise if you want to get the 
whole receiver to pass muster. But you start fixing receivers that do not muster 
and you find some things at some points that clue you in as to which way to 
go next.

The book does not say pull the third mixer and inject a signal into the plate 
pin. But you learn it removes the RF deck noise from the mix and makes it 
easier to set up the 0.1 and 1.0 bandwidth adjustments as well as that last RF 
deck transformer. You also learn to use a counter if you have one. The TM never 
got updated to doing things with a counter on the bench.

Pulling the tube and injecting on the plate pin of the socket, gives you a 
point to get 455 into the IF and set your BFO against if you do not have a 
counter to count the BFO directly.

Also at this point you can neutralize the BFO. Instead of using a cal tone we 
used the signal input. You disconnect P213 (the crystal side of the balanced 
coupling between the RF and IF deck) and stick the long spline or Philips 
screwdriver into J513 and short the center pin to the chassis while you trim C525 
for minimum meter reading on the Audio power meter You have to kick the diode 
load meter up a range or two when you turn the BFO on. You can find a null on 
the diode load meter here as well.

The TM says hang a RF meter on the 455 output so you cannot use your little 
jumper from the IF deck to the back panel to short P213. In life we metered the 
audio out with a power meter (TS585) [ and AC volt meter with a DB scale and 
a 600 ohm resistor] (try two 1,200 ohm in parallel a couple 1/4 watts or 
better a couple 1/2 watts will handle the 1/2 watt output of the receiver. 

On the RF meter you are reading total signal and noise out of the BFO and 
receiver at 455Khz. You are adjusting the BFO for minimum noise. Looking at the 
Audio output of the Diode load reading -30 volts with the BFO on you will not 
see a large dip when you hit minimum noise. As you need to do this with the BFO 
and a signal you are hard put to see the dip on the over 1/2 watt audio 
output.

Plug in your head set and listen to the audio. You can hear the change in 
noise. Adjust for least noise.

Try no input to the IF deck. Turn the BFO on. With no signal to mix with the 
diode load and audio noise should be low. Trim C525 for a dip in the audio 
noise.

Make a BNC short with connector and come coax. Use this on your R390. Use the 
mini BNC cable from J513 to the back panel and place the shorting plug there.

Do this setup and trim C525 to minimum output. Draw a line on the chassis 
side in line with the slot on C525. Now put J513 and P213 back together. Readjust 
C525 and see if its goes to minimum at the same point.

A good BFO is a quiet BFO and it just gets adjusted for minimum output 
against a signal.

For the SSB guys, remember a quiet BFO and a strong BFO are not the same 
thing. C525 changes the noise of the oscillator circuit not the output level of 
the oscillator. Max output from adjusting C525 is not a better SSB signal 
catcher.

Remember all this is done with 455 going into the plate pin of the third 
mixer.
Remember to use a cap between that B+ plate pin and you signal generator.
Exact generator output level is not specified.
If you have a counter use it.

Roger L. Ruszkowski