[R-390] Christmas Wish List (part 3)

Flowertime01 at wmconnect.com Flowertime01 at wmconnect.com
Thu Dec 8 23:58:31 EST 2005


>From Craig C. Heaton AKA wd8kdg Christmas Wish Number Two.

A blow by blow account of what to connect, where to connect, the values are 
we looking for, and the correct order of tubes to swap while measuring for the 
30db goal. Is that clear?? 
------
Warning this diatribe will induce drowsiness.
Warning most satire was left out and this is no fun to read.
But it was ask for and I offer it in the spirit of the Season.
May you all enjoy. If any one wants any thing more, just ask. I'll give it 
another shot. I can fix them my self faster than I can write about it.

Roger KC6TRU

----------------------------------
Ok so you have run all the subassemblies through the dishwasher. Pulled the 
RF deck, cleaned the gear train and lubed it with the magic synthetic Mobile 
oil. Everyone that went this far was sure to visually check the RF band switch 
alignment.  Some have even taken the whole frame out in the summer sun light 
and given it a bath. A few more have removed the front panel, sanded it clean, 
scraped out the engravings an applied non military paint like things to the 
panel. Then all the parts were heaped back inside, the green screws (bolts) were 
snuggled up (yep snuggled), the connectors connected and the RF deck cover 
replaced, many more than the recommended tube shields inserted, all the knobs 
checked and twiddled, some (like me) mounted the micro dial on the BFO shaft, a 
few more installed the "@RARE@ dial bezel with @SPOOK@ cover to arrive at one 
specific incarnation of a good looking receiver. All this effort has been dully 
noted and good credit standing with Santa applied for.

Then the shocking receiver was placed on the "bench" and properly grounded so 
that it could be checked to be a working receiver. After much tube testing, 
tweaking and other maintenance procedures as described in the manuals were 
completed the receiver was found to have a calibration tone on every 100 
kilohertz. The signal generator was attached to the antenna input, the voltmeter and 
load resistor attached to the output, the calculator dusted off and the End to 
End sensitive test was performed. After some button smashing on the calculator, 
review of math and magic thumping of test equipment, the receiver is found to 
clearly pass the military 10:1 signal to noise ratio any where on the dial. 
Some spots almost made 20:1 but the receiver clearly passes the 10:1 with less 
than 4 microvolts in and .4 watts out. All this effort has provided a working 
R390 receiver in much better condition than some things we have seen sell on 
E-bay. At least you know what state your state is in and that's a step ahead of 
some things we have seen sold.

You have solid stated the rectifiers. Inspected the caps and replaced the 
brown and black critters. You have looked under all the subassembly decks and 
fixed any charred resistors, leaky caps and other real problems. You have the RF 
band switch properly adjusted by visual inspection. You have the crystal 
oscillator switch adjusted by visual inspection. You have the dial over run set. 
You have the Zero adjust set. You have the detent load set. You have the 
mechanical alignment set. You checked the slug racks for nice operation. You have 
resolved your ballast tube problem to your satisfaction.  You have the power 
supply caps on the Audio deck under control. You have the PTO end points set to 
your satisfaction. You have been through the RF alignment more than once. You 
have watched the receiver glow in the dark both the top side and the bottom side 
and none of the tubes have that that unhealthy blue glow in them. The 5654's 
tend to get it first. Then the 6C4's. Just using what you have you are doing 
the best you can. But maybe you can get a bit more with what you have if you 
just had some inspiration. You have a good working R390 and you have done the 
real right stuff to get it up to snuff. You have done all the right stuff, But 
you been watching Emeril Live and you want to kick it up a notch. Its Christmas 
and you deserve some additional sensitivity. Besides the sun-spots are not 
all that good this year and you need some additional sensitivity. 

You have been reading the E-bay pages and other things where you have seen 
allusions to receivers that do better than just pass requirements and work. You 
have explored this subject and understand that if you just replaced every tube 
in the receiver it could be better. A quick look at some tube prices and your 
pocket book tells you that that approach is not practical. Asking for a tube 
tester is not a solution to the problem. A more selective scientific selection 
of tubes to place on the Christmas Wish List is in order. You could get every 
tube you ask for no matter what the cost if you approach the request with 
rational. The I want it may fly with the children but these kinds of glow tube 
procurements need a bit more presentation to get through the procurement 
approval process.

So you put out an E-mail request for some help to the R390 reflector Fellows 
and start asking for the inside real spook tips on how to get this done. All 
right enough ducking and dodging. Enough introduction. Enough procrastination. 
Its time to fish as I am not cutting no stinking bait. I did the RF end to end 
sensitivity test in DB milliwatt and too much detail. So I will do this tube 
time in AC volts across a 600-ohm resistor. All in plain text as tables do not 
go through my mail tools. That's My choice and I'm sticking to it. Beat the 
600-ohm problem with a pair of 1.2K ½ watt resistors. A pair of 1.2K ¼ watt 
resistors will work. Pick you AC meter with the most readable scale. Go digital 
if you must. Warm up the signal generator.

OK so you have this receiver on the bench. Its semi time. You did a good face 
wash, mechanical check and electrical alignment. You do the RF end to end 
sensitivity test described in part 2 of Christmas wish list and the receiver 
fails to get the 20:1 you need to have the trick chief hack off on the paper work 
so you get down the hall and on to important things. Where and how do you 
start trouble shooting this receiver problem. The receiver is aligned. Its OK. It 
works (sort of). It has no specific problem. It just does not pass the shops 
20:1 signal to noise test setup. I seen this twice a day six days in a row took 
a trick shift (2 days more or less) off and did it again for years. You need 
to know how to deal with this problem. It is not in the manuals. It was taught 
in school.

You need a minimum tube test set of each type tube in the receiver. (More is 
good but not required) For every tube type you need N-1 tubes. There is one 
6DC6. N-1 is zero 6DC6 in the tube test set. There are 3 6C4's one is in the 
first mixer. Above 8 MHz the receiver uses only two of the 6C4's So N-1 is 1 6C4 
in the test set. The 26Z5 are sand state and need zero. The 0A2 either lights 
or got replaced. N-1 is zero 0A2. The pair of 5654 AKA 6AK5 needs 1 The three 
6AK6's need 1 because the line and local are in parallel and you only meter 
one branch. You need 4 each 5749 AKA 6BA6. You can grab the 5814's out of the 
calibration area and V509 for the theses test, but 3 each 5814's help.

So you need 
1 signal generator (and wire to antenna input)
1 AC volt meter (and test leads)
1 600 ohm resistor
1 6AK5
1 6C4
2 6AK6
2 6BA6 /5749

That's the ingredients

A blow by blow account of what to connect, where to connect, the values are 
we looking for, and the correct order of tubes to swap while measuring for the 
30db goal. Is that clear?? 

For the 60's Vets 

                                   EXACT

On the back of the receiver is J116. On the outside this is a BNC connector. 
On the inside this is a mini BNC connector. If your receiver is missing this 
item ask Santa for one this Christmas. 

On the inside of your receiver is a coax with two mini BNC numbered  P116 and 
P114. The P114 connector should be attached to J514 on the IF subassembly. 
The R390 did this sanely in BNC cable and all this mini stuff is avoided. If 
your receiver is missing this small assembly or it does not have continuity ask 
Santa for one this Christmas.

Set the signal generator up for 455 kHz, 150 microvolts un-modulated. For 
those with a copy of the TM. this is paragraph 73 Adjustment of GAIN ADJ 
Potentiometer. Right out of the TM and to follow here in detail. But for reference and 
history just so every one knows where this stuff is being drug in from. So 
people with keyboard say I miss the truth some days. Its close to Christmas, My 
wish list is in and I'm not about to jinx my chances with any fibs here. 
Understand.

If you were doing the electrical alignment of the receiver and got to 
paragraph 73 to set the IF gain R519, and knowing that proper prior planing prevents 
poor performance, you would make a could switch flips while in the conduct of 
this test and ensure the IF and audio deck was going to make the grade when 
you get to the end to end sensitivity test. Paragraph 73 has no clues on how to 
do a signal to noise test on the IF and Audio decks. This is where the inside 
information comes from. Believe you need a 30:1 ratio here. Exact reference to 
a voltage or power level is not needed. Exact calibration is not needed. The 
signal generator should be in the 2KC-band pass as best you can rock the 
generator into the band pass. The mechanical filter is not tunable, so you have to 
rock the generator into the filter. The 150 microvolts is not exact. The 150 
number is what got published in the book.  Crank your generators output to 
150-microvolts and accept it. 

If I have no frequency counter how do you get my signal generator set to 
exactly 455? And the follow on question is how do I get the BFO set to exactly 455 
and the knob pointing at zero? Glad you ask and it offers a nice transition 
into the next paragraph.

Un hook P218 from J518 and P213 from J513. Tube pullers worked good for this.
Un kook P114 from J514.
Hang a 600-ohm resistor across terminals 6 and 7 of TB 102.
Hang the AC voltmeter across this load resistor.
Set the local gain to max
Set the RF gain to max
Set the BFO off
Set the limiter OFF
Set the function switch to MGC

To set the signal generator on frequency do the following :
Back the signal generator output down very low.
Turn the BFO off.
Set the band width switch to 1KC.
Hook P114 to J513.
Turn the signal generator modulation on.
Set modulation to 30% and 400 Hertz as published in TM paragraph cited above.
Hook the signal generator to J116 using a suitable length of coax and 
connectors as require too make the setup.
Hook P114 to J513

Rock the signal generator into the 1KC band pass while watching for a AC volt 
meter peak.
Set the band width switch to .1KC.
Rock the signal generator into the .1KC band pass while watching for a AC 
volt meter peak.

This get the signal generator peaked into the 455 crystal filter of the .1KC 
band pass.

To set the BFO to zero do the following;
Turn the modulation off on the generator.
Turn the BFO on.
Tweak the BFO Pitch knob to zero beat with the signal generator, while 
watching for an AC voltmeter null.
If null is not with the Know pointed to zero, perform a shaft clamp 
adjustment as required to correct problem.
Do not adjust knob to shaft. 
Shaft at knob is burred and will not adjust easy. 
Spline bolt in knob is inaccessible at BFO zero. 
Set clamp inside front panel so Spline bolt is easily accessible while making 
this adjustment.
Just some inside info not in the TM.
Turn the BFO off.
Set the generator output to 150 microvolts turn the modulation back on and 
peak it into the 455 crystal as best you can.

Un hook P114 from J513
Hook P114 to J518
Set the band switch to 2KC
You likely pop the top off the mechanical filters here and trim each cap up 
to the best you can.
This is a subject for another mail.
You may neutralize the BFO at this point
This is a subject for another mail.

Set the gain adjust as follows.
I can name five ways to do this "properly" For this test do the following and 
readjust later if necessary.
Hang a DC voltmeter on the diode load to ground.
Set the signal generator modulation off.
Set the signal generator output to 150 microvolts.
Set the BFO off
Set the bandwidth to 2KC
Set the function switch to MGC
Adjust the IF gin R519 for -7 volts on the DC load.
Remove the DC meter.

As long as you are here, you may as well check to see if this receiver IF and 
Audio are going to make the signal to noise test. You can do it now as part 
of the alignment of come back as part of your trouble shooting.
As long as you are here. 400-milliwatts across a 600-ohm load is Volts = 
square root ( P * R) = 15.4919 Volts = square root (0.4 * 600 )

Verify the local gain as follows
Set the signal generator modulation on.
Set the signal generator output to 150 microvolts.
Set the RF gain to max
Set the local gain to max
Set the BFO off
The AC volt meter on the local gain must exceed 15.50 volts. (It could even 
be twice this voltage)

Verify the line gain as follows
Set the signal generator modulation on.
Set the signal generator output to 150 microvolts.
Set the RF gain to max
Set the line gain to max
Set the BFO off
Move the AC volt meter on the line gain with the load resistor
Output must exceed 15.50 volts. (It could even be twice this voltage)
The line level meter will flat peg out on the +10 meter setting.
100-milliwatts across a 600-ohm load is Volts = square root ( P * R) = 7.745 
Volts = square root (0.1 * 600 )
Reduce the line gain from max until the AC voltmeter reads 7.745 volts.
The line meter should read 20 DB +10 on the switch and 10 on the meter scale.
I know this is tricky math but follow along.

Ok a 150 microvolts modulated 30 % with 400-hertz tone should produce 
.4-watts of output into a 600-ohm load on both the local and line outputs. The power 
gain is there. Both channels work. The line meter works. The BFO is set. Run 
through all the 2, 4, 8, and 16 bandwidth switch settings and check that the 
power is up. All the mechanical filters are at least passing signal. You checked 
the .1 and 1 when you rocked the generator to 455. Depending on what 
modifications you have made to the caps in the audio section, the power output may be 
over 1 watt. At least a ½ watt in a stock deck. 

If you have -7 volts on the diode load and less than ½ watt out of either 
audio channel you know you are looking for poor tubes between the diode load and 
the output. Grab your schematic and see which tubes are in the line.

If you have trouble getting -7 volts on the diode load, look at the 5749's 
and 6AK6 in the IF deck. The diode load should crank to -10 or 15 on the end of 
the resistor range.

Ok a 150 microvolts modulated 30 % with 400-hertz tone should produce 
.4-watts of output into a 600-ohm load on both the local and line outputs. You can do 
the signal to noise test on either the line or local. You can do them both at 
the same time. If they are not equal you now the 5814 and 6AK6 in the audio 
deck is the difference between the two outputs. Ok a 150 microvolts modulated 
30 % with 400-hertz tone should produce .4-watts of output into a 600-ohm load 
on both the local and line outputs.


To get the receiver to pass a 20:1 one signal to noise test this much of the 
receiver must be able to demonstrate a noise level that is 30 DB down from 
this output. 30 DB in AC volts across a 600 ohm resistor is 16.67 volts.

Verify the IF and audio noise range as follows.
Set the signal generator modulation on.
Set the signal generator output to 150 microvolts.
Set the RF gain to max
Set the local gain to max
Set the BFO off
Set the band width to 2KC
Set the limiter off
Set the function switch to MGC.
Set the audio gain for the channel under test so the AC voltmeter reads. 
17.32 volts 
This is .5 watts into a 600 ohm load and equivalent to 27 DB.
Set the signal generator modulation off.
The AC voltmeter must read less than 0.65 volts.

This test is not exact. Set the Audio as high as it will go. 
Turn the modulation on.
Read the AC volts.
Turn the modulation off
Read the AC volts
Is the difference more than 16.5 volts.
Yes you are OK
No you have work to do.

Seeing all these silly AC voltages you understand why a nice Analog AC 
voltmeter with a DB scale and range switch you understand how to use is nice here. 
With a DB scale on the meter you just read the max DB level with the modulation 
on. Switch the modulation off and read the meter in DB again if the 
difference is greater than 30 you pass go and collect $200.00 if not you have work to 
do.

What do you do when you do not have the 16-volt or 30DB difference?

Round up those tubes. We know from life that front to back makes the most 
difference. In this test setup V501 is first and that 6AK6 in the audio channel 
is last. Some AGC tubes are out of the circuit. The exact order is as follows
V501, V502, V503, V504, V506, V801, V602 and V603 or V604. That all there is 
in the test string, eight tubes.

Pull the BFO and PTO 5749s for test critters. Pull V508 for a test critter.
V502 and V503 are under control for this test. V501 is the test socket. 5749 
/ 6BA6 is the test subject.
If you have two extra 5759's install them into V502 and V503. If you have 
been here and done this before you have the two poorest 5749 you own marked and 
ready for this test.

Run the test.

Verify the IF and audio noise range as follows.
Set the signal generator modulation on.
Set the signal generator output to 150 microvolts.
Set the RF gain to max
Set the local gain to max
Set the BFO off
Set the band width to 2KC
Set the limiter off
Set the function switch to MGC.
Set the audio gain for the channel under test so the AC voltmeter reads. 
17.32 volts 
This is .5 watts into a 600 ohm load and equivalent to 27 DB.
Set the signal generator modulation off.
The AC voltmeter must read less than 0.65 volts.

Swap the 5749 out of  V501 and repeat the test.
Did this tube provide a higher maximum voltage?
Did this tube provide a wider range between modulation on and modulation off 
than the first tube?
Check all the 5749 you can round up and rank them more range is best.
If you do not have spares, swap the poor ones into V502 and V503 and run them 
all again.
If you do not have spares put the poor one in the BFO
Put the second poorest one in V508 AGC IF AMP. 
Run the receiver in MGC do not listen to SSB or CW and wish for 5749 for 
Christmas.

First put the very best 5749 into the PTO.
Second put the next best one into V501.
If this set up will not pass the 30:1 ratio test here, then put the best one 
in V501 and wish.
Third put the next best into V502,
Fourth  is V503.
Fifth is V505 (BFO)
Sixth is V508 AGC.

Save at least two off the 5749's for test.
Any tube that will not get you the 30 after 2 spares is a discard.
Save the rest if you have them.
The next time you run this test put your two weakest but (hopefully) passing 
5749's into V502 and V503 and grade all the other 5749 you can round up. 
Install the best of the 5749 and re grade all the spares so you know which will at 
least pass and which bottom end ones to save to start the test with the next 
time.

You next test the 6AK6's in V504. You need to leave one in the audio channel 
you have the meter on. You can swap the other audio channel with V504 and pick 
the best one to go into the audio channel under test. Then check that one 
against the third one again for comparison. Put the best one in V504 and the 
second best in the local channel V603. If you have one spare 6AK6 to start with 
you can rank all three and place them in V504, V603 and V604. 

You next test the 5814's in V602 because this socket tests both side of the 
tube at once in series. You can pull almost all the 5814's in the receiver and 
rank them. You need to have a tube in V506 the detector and V602. When you do 
this in the R390 you have to watch the series filaments to keep the tubes you 
need lit up. Rank them all and put the best as follows.
First one in V506 the detector
Second one in V601 audio.
Third one in V602 audio
Fourth one in V507 limiter
Fifth one in V509 AGC
Sixth one in V205 Calibration
Seventh one in V206 Cal buffer

Save at least two for the next test event. 

Once you do this one time you start to see how tubes are making a difference. 
You swap the order of V501 and V502 a couple of your best 5749's and see a 6 
or more DB change in the span of the modulated to un modulated signals and you 
become a believer. Once you get enough good tubes in the receiver to get the 
range in the IF and audio sections you can hear the difference in the receiver 
when you put it back on the antenna.

V501, V502, V504 are the first three to receive new tubes if you can not get 
the 30:1 range. This is 2 5749's and a 6AK6. Three 5749's and two 6AK6 will 
get you there as V501, V502, V503, V504 and V603 or V604. You almost can always 
find a few good 5814's to fill the string.  Trade a 6AK6 or 5759 for the 6DC6 
on your Christmas list and have a 5 pack shipped in time for the Holidays.

Once you get the IF and Audio deck aligned and tested you put the receiver 
back together and start on the end to end sensitive test as detailed in part 2. 
You go into the RF deck alignment knowing the IF and Audio are good those 
tubes are good and the signal to noise ratio is good. What ever you need to do to 
the receiver now is in the RF deck.

If you went through this mess of testing for the um-tenth time it takes about 
a ½ hour to just do it. First time can take a week of you hobby time. If you 
go through all of this and you have a meter needle that just wants to bounce 
big time, and none of the tube juggle seams to help, you may need to be 
rethinking caps. But you should still be able to rank your tubes and know the better 
from the poor. You can look into the IF deck and judge the cap problem. An 
assembly full of brown or black things is a cause to heat a solder iron. Been 
there done that is cause to consider some new tubes.

Once upon a time a brought 5 new Raytheon 6BA6's. They were the hottest tubes 
I have ever had. Power out over 1 watt. I can not get any of them to pass the 
noise test. I'm burning them in the BFO and AGC rectifier. I know its BFO 
noise but I can live with it. Moral new tubes may be worse than what you have. 
There are lots of good brands.
So all this work may not cure all ills.

In service we just knew if you did not get past this test you were not going 
to make end to end test. It was a place we learned to divide and conquer. We 
had to set this test up and do the BFO and IF gain here anyway. It was just a 
few switch steps to do the test. If you had to go for tube jockey status, this 
test at the half way point  sorted the tube types. Only the 6C4, 6AK5 and 6DC6 
were left for the RF section.

These tubes were done the same way with the receiver set up for the end to 
end sensitivity test.  You set up the RF alignment and could get the 20:1 ratio 
when you did an adjustment or you swapped some tubes. If the receiver was 
looking poor after getting the IF to pass, you just started with a new 6DC6. If it 
did not make a lot of difference you put the old one back in and went through 
the adjustments. You swap the 6AK5 into the crystal deck above 8 meg and rank 
the 6AK5's. You swap all the 6C4's you have into V603 and rank them. Put the 
best one forward. Do you put the best 6C4 into V202 or V203? Depends on what 
you are going to do under 8 meg. 

You keep at this long enough, buying tubes and grading them you get over 20 
end to end and have a wonderful receiver. You stay at it and you find you have 
25 or more end to end. Then you find the meter is just not laying real still. 
So now you can go at it back to front one stage at a time with signal 
injection and see where that meter bounce is coming from. Then you can consider nosey 
caps. But until you get a real good set of tubes in the less noisy caps are 
not your major problem. I have found just getting good low noise tubes to be a 
problem or buy and try.  This cap subject has driven most of us to just acquire 
a complement of good quality caps and re work the whole subassembly one-week 
end. Again the IF deck gets it firs and then the Audio deck gets it and last 
we do the RF deck.

So I hope this covers the subject clearly and allows everyone to test their 
R390 so as to generate a concise wish list of parts that will make a difference 
in what they hear with out exceeding the allowance.

Merry Christmas to All
 
From 

Roger KC6TRU





More information about the R-390 mailing list