{Collins} Fwd: Re: KWM-2A problems and tubelore question

Forwarded to list as requested.
Note: forwarded message attached.

-- Attached file included as plaintext by Ecartis --

X-Apparently-To: bob@xxxxxxxxxx via; Sun, 24 Oct 2004 13:52:11 -0700
X-Originating-IP: []
Return-Path: <k2qde@xxxxxxxxxxxxx>
Received: from  (EHLO mta4.srv.hcvlny.cv.net) (
  by mta106.biz.mail.re2.yahoo.com with SMTP; Sun, 24 Oct 2004 13:52:11 -0700
Received: from dellgx1 (ool-18bbb8e5.dyn.optonline.net [])
 by mta4.srv.hcvlny.cv.net
 (iPlanet Messaging Server 5.2 HotFix 1.25 (built Mar  3 2004))
 with SMTP id <0I6300DBLWMXLJ@xxxxxxxxxxxxxxxxxxxxxx> for bob@xxxxxxxxxx; Sun,
 24 Oct 2004 16:52:11 -0400 (EDT)
Date: Sun, 24 Oct 2004 16:52:11 -0400
From: Stu Martin <k2qde@xxxxxxxxxxxxx>
Subject: Re: {Collins} KWM-2A problems and tubelore question
To: bob@xxxxxxxxxx
MIME-version: 1.0
X-MIMEOLE: Produced By Microsoft MimeOLE V6.00.2800.1409
X-Mailer: Microsoft Outlook Express 6.00.2800.1437
Content-type: multipart/mixed; boundary="Boundary_(ID_ryC8rw6jBzGk3U42+RMiOA)"
X-Priority: 3
X-MSMail-priority: Normal
References: <20041024180618.74180.qmail@xxxxxxxxxxxxxxxxxxxxxxxxx>
Content-Length: 16287

This is a multi-part message in MIME format.

Content-type: text/plain; charset=iso-8859-1
Content-transfer-encoding: 7BIT

Sorry to hear an old AF Tech is having a problem on one of its FRC-93s. You
certainly have the experience and knowledge to handle any Collins repair
task. I hope my "parts chassis" can assist in locating any components you

Anyway, your KWM-2 oscillation sounds like the classic case of a bad ground
contact under the four compartment shields, or it could be the final amp
feedthrough cap, C123. Attached is a compiled list I've saved through the
years. Best practice is to take a copy of the schematic and note each item
listed on schematic. Yes, that's quite a chore, but well worth the effort.
You be glad you did later. It help me find 28 problems in my two radios
during the last 4 years. Z5, by the way is a "no-touch" item. It's factory
set from the vendor. I may actually have that [adjustment] somewhere but not
sure. Can't look right now but I will later. One other note: Look into the
final-Amp compartment with the cover off. Do you see a white, ceramic
variable cap looking up at you? It would be identical to the caps under the
slug-rack. If not, you may see the plates of a metal air-variable cap facing
downward. If your unit has the white ceramic cap, be sure the two Final
tubes are marked 6146A's. If you have the metal air cap, then you can use
6146B's. Both must be the same. No mixing! [fyi: Collins never spec'd or
used a "matched set." Production did stick with the same
manufacture/date-codes, though. RCA, Sylvania, a few GE].

The screen voltage removal is necessary during the 455-KC IF alignment so
you don't activate the transmitter. Remove the HV tube in the supply as
well! One can't make that adjustment correctly due to the load on the supply
and RF circuit. Follow the manual! These adjustments are quite broad and the
least loading effect allows a more exact setting. I assume you've confirmed
the need to make the adjustment to begin with.

Tube glow (blue) is common in these units (6U8A's, etc). Doesn't mean
they're gassy. Some say, "...if you find a 6AZ8 in a Collins radio, it's
bad...! In this case, just swap tubes around and you'll quickly find the bad
one. Tube testers are worthless with these radios. Always go with another.
Plenty available [eBay].

Lastly, you will need the production date in order to determine the correct
part values/changes and schematic/manual revisions. Collins manuals were
updated every two years to reflect design changes. Dates stamped between the
crystal pins are unreliable as they can be swapped. One trick: Note the
electrolytic "can" behind the meter. Find a dental mirror, or any small
mirror and flashlight, and search for the date-code stamped on that can. Now
check the solder leads no the terminals. Do they look like they were
re-soldered? No one, not even I, can re-solder a lead like the Collins
production staff! If the leads look factory, then the cap is original and
you can trust the date-code. If the cap was replaced, all bets are off, but
there are other ways (components values). For example, the 4, 1N34A balance
mod diodes were changed to a single quad-pack in '64. Does you unit feature
a soldered-in mechanical filter, or a rectanglular plug-in type? Lastly,
NOTE any and all "re-soldered" connections. Although they may be
illegitimate component replacements, they could be out of spec items or more
so, illegal mods.

Basic tube-type troubleshooting is what's next. Swapping in tubes is cool,
so long as you return the original if no change is noted [No realignment
necessary when swapping any tube in a KWM-2/2A Only exception may be the
neutralization of the 6146's if different brands are used]. Once you have a
stable unit, I'll scan you a copy of the production test and alignment
sheets. They're far better than the manual. As you progress, let me know if
you have any questions regarding rev levels or service bulletins. I was
employed by the Collins Data Processing Division in '72 and was involved
with the Collins C-Systems field support group.

Good luck,
Stu Martin, k2qde

----- Original Message ----- 
From: "Robert Stewart" <bob@xxxxxxxxxx>
To: <collins@xxxxxxxxxxxxx>
Sent: Sunday, October 24, 2004 2:06 PM
Subject: {Collins} KWM-2A problems and tubelore question

> Hello to the group.  I have returned to Ham Radio
> after an almost 16 year absence and decided to get the
> radios I wanted the first time: Collins.  Of course, I
> bought them from ebay with the predictable result: one
> works pretty good and the other has problems.
> My background, I used to have an Advanced class
> license, and I taught electronics in the Air Force, so
> I have a fair amount of technical skills.  Plus, I
> built several Heathkit rigs and did some homebrewing
> back in the day.  I've returned to Ham Radio with an
> Extra this time.
> My questions;
> My KWM-2A has a self oscillation somewhere.  When I
> tune Z5, its frequency changes.  Any ideas where to
> start?
> Next: in section of the instruction book, it
> says to disable the screen circuit before proceeding
> with the 455-KC IF alignment.  Is this really
> necessary, or is it sufficient to simply bias the
> finals past cutoff?  I'm guessing that the real
> purpose behind removing the screen voltage is to just
> prevent overheating the finals while messing with T1
> and C54.  What's SOP here?
> Finally, I've forgotten most of my "tubelore".  I also
> have a KWM-2 which works pretty good.  But, both rigs
> have one tube (different on each) that glows blue.  In
> one, it's V13.  In the other, it's V3.  Swapping both
> out doesn't fix the problem.  Is this a sign of some
> voltage or drive problem that I need to address.
> Thanks in advance,
> Bob, AE6RV
> Sponsored by the Collins Collectors Association
> Nets: Tues: 3.805 Mc-2000 Central / Thur: 3.875 Mc-2000 Central
> Fri: 3.895 Mc-2000 Pacific / Sun: 14.263 Mc-2000 UTC
> 1st Wed (of the month) AM Net 3.880 Mc-2000 local (ET, CT, MT, PT)
> Sun AM Net: 29.050 Mc-1200 Central

Content-type: text/plain; name=collins_svc.txt
Content-transfer-encoding: quoted-printable
Content-disposition: attachment; filename=collins_svc.txt

KWM-2 Reconditioning and updating Notes  08-06-01

Maybe you have a crack in a resistor, etc.  Why not take a long plastic =
test tool or a long typewriter
eraser and pock around on the components ( near j23  j24) as  you listen =
to audio and maybe you
can pinpoint the defective component or connection.

Make sure the headphone jack isn't being compressed (and thus breaking =
circuit to the speaker)when the rig's resting flat on the bench.

PA breaks into oscillation on 75/80,  and hot RF on chassis: Clean and =
reseat all shielded "cans," (bandswitch covers) final-amp cage and lower =
chassis shield, and all tube sockets. Loosen and retighten feedthrough =
caps, as well, removing all oxidation between the part and the chassis. =
Chassis to Shield contact is vital with this equipment.

Check C123 ceramic feed through caps . Should be 1000 pf w/5/16th inch =
threads. Part of phase-shift voltage divider in feedback neut. circuit =
(C123 and L23 to 10uh/22uh). There should be no grid peak reading with =
Plate Tuning in TUNE position. C123 may use Mica substitutes. (See =
replacement info in Final Amp section)

6U8A begins to cause HUM after 6 to 8 months of service. Replace with =

Check for heat stress on resistors and electrolytic. Check tube sockets =
for dark color and dull capacitors.

Check 'IF' feed through at 055 and 155 on the dial. May be stronger than =
0, 100, and 200. This is not a bug, but IF feed through. Install and =
test IF filter.

Frequency count all oscillators: BFO [critical], HFO, Crystal Cal, PTO, =
Tone, etc.

Leaky caps: C100, C101, C103, C259, C1, C3, C4 and C47. Check all tube =
grids for positive bias. Replace proceeding coupling cap.

Possible sidetone coupling into audio circuits. Check bias on AF tubes. =
Should be Cut-off!

Unstable S-Meter reading #1. Check V4 for gas - After warm-up. Adj. =
S-Meter Zero Adj. (across R158. Check R158, R79, R76, R158A, R147, R121, =
and R123. Check R132 Rcvr-Gain pot and its connections. A good rule of =
thumb is it should read about 20 over S-9 on 20 meters using the =

Unstable S-Meter reading #2. [75S?] Add diode 1N458 from XV-7 to ground =
(negative) to clamp it down. Measure turn-on (cold) at XV7-1 & XV6-1, =
then one hour for difference. Select tube V7 for most uniform warm-up =

Check all resistors and caps for correct values, at tube grids, with =
meter. Replace caps with "orange drops" or equivalent.

Replace C102? (V16-1, 6EB8), 100uf cathode bypass cap on 1st audio amp. =
A 0.01 uf 600v cap from XV16B-8 to ground, and a 56 ohm 1 watt from the =
'yellow' transformer lead T6 to ground. This caused sever audio =

AF Gain > 9 O'clock causes sever feedback. Bridge bypass caps with a =
good one (.05uf, 400v). Use gloves!

Check and clean relays with contact cleaner/burnish tool.

VOX Delay & Overheating -- Replace R20/R47, C47, and two 68K 2 watt =
resistors at V4B [6AZ8] with a single 34K, 7 watt resistor, or 2, 68k 10 =
watt. V4b-cathode. Chg. V14 & V4. Should be 34k in cathode.

VOX instability can be reduced by adding a 0.1 uf capacitor across =
contacts 12 and 13 of relay K2. Change the 4 mfd electrolytic. Chang the =
.047, .05, or later .068 mf, as well.

A machine gun effect of the VOX relay can be remedied by adding a 0.01 =
uf capacitor (913-3013-000) from tube socket XV1, pin 7 to ground. This =
effect is caused by relay transient's fed through the ground lead of R3, =
which is connected, to emission switch S9C.

Resistor R175 (47 ohms or 4700-?) and capacitor C262 (0.01 uf) is added =
between tube socket XV3, pin 3 (6AZ8 cathode) and ground to provide bias =
voltage and improve operation of the IF amplifier. R175 - loss of xmtr =

RF gain is not working: Is it not making the sensitivity change, or is =
it not moving the s-meter?  I would recommend cleaning all relays if you =
have not done this, and maybe twice just to make sure. I had n s-meter =
problem with the RF gain control that did not clear up until I cleaned =
the relays twice.

Unusual AGC behavior? Action is quite smooth until the radio warms up, =
and then it seems to speed up and become choppy. RF Gain seems OK.  =
After checking AGC capacitors, suspect grid leakage in one or more of =
the AGC controlled stages, perhaps a grid starting to behave as a diode =
and affecting the AGC action (grids behaving like diodes). Measuring DC =
grid voltage with no signal. Suspect leaking in RF or IF tube could =
affecting AGC action and gain. Measure the DC grid voltage on AGC =
controlled stages with no signal applied using a VTVM.  The Collins =
schematics show the nominal DC grid voltages that should The "idle" grid =
voltage should be about -0.8 volts. RF amplifier it should be about =
-1.2. These voltages could be off (more positive) by 10-20% (i.e. the V6 =
grids reads in the minus .5 - .7 volt range instead of -0.8). After hand =
selecting some 6BA6 and 6CD6 replacements, I was able to achieve idle DC =
grid readings closer to what the schematic indicated. AGC action was =
noticeably smoother, and overall gain increased by 1-2 S-units! A small =
voltage difference reveals a problem tube, but it appears that if a tube =
has a leaky grid, or its grid is behaving like a diode, it either loads =
down the AGC line or otherwise degrades its action, as well as overall =
receiver gain.

V1B runs too hot. Add tube shield [TR6-6020B] to cool, add a 10-ohm 1/2 =
watt (R75) from XV1B-3 to ground. Mount between RCVR-Gain pot R132's =
ungrounded terminal and ground. Note: Military tube shields are IERC =
(International Electronics Research Corp., Burbank, CA). 6U8 size is =
TR6-6020B, 6AZ8 size is TR6-6025B.

ALC meter Instability: If reading drifts, replace C157, a 0.01uf 200v =
with a low-leakage Mylar/polypropylene cap between XV17A-1 and 3 (6BN8). =
Check grid emission from ALC tube, and all ALC caps.

If the ALC meter will not zero, resistor R170 (220 ohms) across ALC zero =
potentiometer can be removed. This will extend the range of the =

R4, 68k 1/2 watt, and R7, 47k 1/2 watt cause internal drift due to =

R175 4700 ohm, loss of transmit audio.

Loss of RF gain - Bad micas - Leakage as unit warms up.

Lock Xtal Cal with 3/4 s-meter reading. Monitor for 45 minutes. Reading =
should remain stable and steady. Confirm leakage with VTVM measuring =
grid voltage.

Audio Hum - Check for heater-cathode-grid short in 1st AF Amp, at =
turn-on. Tap tube(s) with fingernail. Check grid voltage de-coupling cap =
(8uf ?). Check AF Gain pot for ground-loop wiring between shielded =
cables-to-chassis and to AF Gain pot.

RF Output drops: Gassy V4, 6AZ8, transmit IF amp., goes dead after a few =

Check value of R46. Should be 2200 ohm, 1 watt (V4B-7). No VOX!

Relay K4 Failure. Check R163, 6k 5 watt on E50-E and F. B+ 275vdc to =
operate K4.

Relay K2 activation: Check for short from K2 coil to armature.

Relay Clip(s) Replacements: Hobby shop's 1/32 inch diameter tensile =
wire, 3-ft. sections. Follow template supplied.

IF alignments: Transformer T2 is properly aligned when the top slug is =
out, or closest to the top of the can. Inductor L4 is correctly peaked =
when the slug is in or closest to the center of the coil. L4 should be =
peaked with the dial at 55.

Capacitor C25 is 3pf in some units and results in poor coupling between =
R2 and L4. Changing C25 to 6pf (961-0122-000).

The 6CL6 plate trimmer capacitor C132 used on 20 meters should be set at =
one-third [2/3 ?]capacitance and L14 adjusted to resonance with C132. It =
is necessary to then realign the 15 meter trimmer C138 and the 10 meter =
trimmer C134.

If the receiver lacks sensitivity, check the coax leads near the relay =
cover on the RF coil shield cans. Dressing the leads away from these =
cans will help.

A low S-meter reading might be caused by R150 being less than 180 ohms. =
Check and replace if necessary.

There are some 6AZ8 tubes, which develop positive voltage at the grid =
causing AVC and ALC trouble. They may also cause excessive S-meter and =
ALC meter zero drift. If the voltage (no signal condition) at V1 or V3 =
grid differs from AVC or ALC line by more than 0.1 volt, the offending =
6AZ8 tube should be replaced. The nominal AVC line voltage is in the =
vicinity of 1.0 volts. The ALC line will be approximately -1.3 to -1.8 =

Insufficient carrier suppression is sometimes caused by power supply =
ripple. It may be reduced by adding a 4.0 uf capacitor C264 =
(183-1763-000) between tube socket XV3, pin 8 and ground.

Changing the value of two capacitors has eliminated the BFO instability =
in some transceivers. Capacitor C53 was changed from 22 pf to 15 pf. =
(912-2760-000) It is located at tube socket XV11 pin2 to ground. =
Capacitor C55 was changed from 47 pf to 100 pf (912-2816-000). It is =
located on tube socket XV11 pin 3 to ground.

Audio squeal in transmit when AF gain control is advanced. This problem =
is caused by a positive feedback path through R164 and the primary to =
secondary capacitance of audio transformer T6.

Correction procedure: Change C104 from 0.0047 uf capacitor to a 0.01 uf =
(913-3013-000). C104 is located between XV16, pin 9 and a tie point. =
Change R164 from 220k ohm to 470k ohm (745-1464-000) R164 is located on =
the tie turret near XV16. Add R182, 68 ohm resistor, (745-1303-000) from =
the 4 ohm tap of the audio output transformer T6 to ground. The 4 ohm =
tap is a yellow lead from the transformer to an adjacent tie turret.

Audio squeal the instant the transceiver returns to receive from the =
transmit function. A feedback loop from the microphone through the =
balanced modulator and into the receiver first IF amplifier and then to =
the speaker. To check this condition, operate in the receive mode. =
Disconnect the antenna and set the MIC and AF GAIN controls at 12 =
o'clock and speak into the microphone. If the audio heard from the =
speaker is quite loud, the following change will eliminate the effect. =
Install R122, a 47k ohm 2 watt resistor (745-5722-000), from vector =
turret E40 terminal D, to tube socket XV3 pin 7. This will put positive =
voltage on the cathode of the microphone amplifier cathode follower tube =
when in the receive mode, thus cutting off the audio feed through. This =
change will have no effect the audio voltage to the VOX circuitry.

To reduce the 3.5 MHz transmitted spurious, move the ground end of C19, =
0.01 uf capacitor, located at socket XV5 pin 8, to the same ground lug =
as C18, a 0.01 uf capacitor, the other end of which is connected to =
socket XV5-3.

A transient occurs in the RF output of some KWM-2 transceivers when the =
VOX drops out. This may be eliminated by installing a 100k ohm 1/4 watt =
resistor from relay K@ terminal 4, -70V line, to relay K2 terminal 5, =
Receive -70V / transmit-ground line.

It has been found that in some units the crystal calibrator trimmer =
capacitor does not have sufficient range to allow the crystal to be =
properly calibrated. For transceivers exhibiting this problem it is =
suggested that the 5pf capacitor, C267, be replaced with a 10pf =

In some cases a "yelp" is emitted from the speaker immediately after the =
transceiver returns from transmit to receive. This undesirable sound =
usually occurs when the RF gain is at maximum and the audio gain is set =
high. The mike gain has no apparent effect on the problem. For those =
units experiencing this problem, it is suggested that the following =
changes be made: (refer to schematic section F). Disconnect the end of =
the 4.7k ohm resistor R176, connected to L9-4, and reconnect it to L9-3. =
Install an IN1490 diode, CR10, with the anode to L9-4 and cathode to =

ALC problems? Receive is OK; transmit: TUNE/LOCK, OK, 100w out; All =
other modes LSB, USB, CW), zero RF output, ALC meter reading at full =
scale+. To disable the ALC just short the ALC jack to ground. This =
disables it and saves a tube. Operate but do not over drive the final.

A low frequency spurious oscillation caused by the mutual inductance of =
coils L3 and L33 can be reduced by relocating capacitor C238, 0.01uf, to =
the junction of coil L3 and resistor R169.

To keep any 60 Hz ripples from affecting the zero on the ALC meter, add =
a 20uf electrolytic across the -70 volt bias line at the input of R189 =
to ground. R189 is located close to the back of the emission switch.

Examine R86, a 2500 ohm 7 watt resistor for burning and color-code =
stripe loss.

R148 820 ohms 2W - The value had drifted up to 1400 ohms. Replaced with =
820 ohm 5-7 watt wirewound. ) This resistor and a diode were installed =
in the PA screen circuit as a part of a service bulletin (This resistor =
heats during transmit).

R4 (68K) and R7 (47K) both 1/2 watt - Both had internal value drift as a =
result of over heating. Replaced with 1 watt units (caused mushy =
transmit audio).

R175 (4700 ohm) - One day the transmit audio just quit. This resistor =
had broken down internally, no external evidence, however.

R50 (47K 1/2 W) (in the tone osc) - Signs of aging. Replaced with a 1W =
unit (also recommended by Collins mod).

IF feed through from BFO (confirm with scope). Copper shield plate =
across XV4. See photo parts and KWM-2 Video. attenuates 455kHz from the =
balance modulator into the outside shield of some coax cables at XV4. =
Ungrounding may help. Also, Add R207, a 680 ohm in series with L2, 10 =
mh. L2 and C216 are located on a bracket just to the left of the PTO on =
the top chassis. R207 connects to the end of L2 and the two red/white =
stripe wires feeding T+275 to the transmit IF Amp.

I got 13dB S/N with a 0.5uV input from 7dB. I found that the 6DC6 was =
noisy! I gained 6dB better noise performance just by replacing it.

Because of the variation in tube and circuit tolerances, The transmitter =
second mixer, V6, 12AT7, sometimes runs above the rated plate =
dissipation. This may result in reduced tube life. To promote longer =
tube life, changing the plate resistor, R143, from 1500 ohms 1/2 watt, =
to 2200 ohm, 1 watt has reduced the plate dissipation (745-3366-000).

To provide improved ALC zeroing, resistor R38, 150 ohm 1/2 watt, has =
been changed to 220 ohm 1/2 watt (745-1324-000).

To eliminate a parasitic oscillation tendency in the tone oscillator =
circuit, capacitor C261 has been added from tube socket XV2 pin 2 to =
ground. This is a 100 pf capacitor (912-2817-000)

To eliminate the possibility of ALC coupling capacitor C157 breaking =
down, a capacitor rated at 500 vdc should be used (913-3152-000).

Check the following components for failure where 40 meters - No loading =
problem: R8-S14, R139, R179, C128/C129.

Flat topping on one polarity only once the volume got high. At pin 3, =
1st stage plate, I see a slight break in the positive rising part of the =
waveform. Everything looks ok at low volume and on the negative part of =
the wave. But, it's at the second stage grid, pin 7 where I see real =
clipping. Not subtle, the waveform is ok below 5V but at 5 V and above, =
it's flat topped just like a conducting diode. On the negative side, =
there's no clipping and at full volume it goes to about -15 V.
Check/replace the cathode bypass on the driver (half) of the output =
tube. Substitute a known good tube, that's about all you can do, again, =
I'm assuming that the audio is clean and nice on average signals with =
the RF gain at 4 o'clock and the audio at about 10 o'clock, if not, you =
may have other problems.

Xmtr/Rcvr IF Alignment baseline. Slug depth on all IF transformers (Top =
slugs only) in inches: T1-9/16, T2-5/16, T5-5/16, L4-1/4, L9-11/32, =
Z5-1/4. Bottom Slugs" T2- , T5-(top surface: 7.0mm, bottom Horz black =
surface: 13.0mm). Note: L9 (8.5mm) is factory adjusted. Do not touch. =
Note, T5's cores should be at top & bottom-most position. The more =
separated the less mutual coupling; the greater the selectivity at =
455kHz. Peak L9 for 2kHz Bandwidth rather than 8kHz.=20

AGC/AVC release time (M-2 or 75S ?). Install 0.25mfd 200 vdc cap from =
E60L to XNB-4. XNB-4 is GND with NB on. Adjust the release time via cap =

Try tack-soldering a 68K in parallel with R180, a 680K, to prevent AGC =
voltage from blocking the receiver during fast noise peaks. This is a =
M-2 version of a 75S-3B trick.

The time constant is 300 to 600mSec then, I assume that if the AGC meter =
is at let's say s-9 +40 it would take 500 to 600mSec to go to 0. Collins =
never got into the real slow time constants in the S-Line... 600 Msec. =
You have the option of putting in a bigger cap if you like. Change C256 =
from 0.01 to 0.068uF and R180 from 680k to 47k. Makes a real difference; =
no pumping of SSB signal, and Less distortion on the audio. There are =
two time constants in the AGC circuit: One fast (680k and the 0.01uF), =
and one slow (1.5M and the 0.47uF). The fast one is the problem. Of the =
AGC voltage produced 30% of it is as a result of the fast circuit. This =
circuit causes the pumping action on SSB signals. It rises and falls =
with the voice peaks. To me this very annoying! If you bypass the fast =
circuit (clip a jumper lead across the 680k resistor) you will hear the =
difference. The audio clears right up. But you will still need some fast =
AGC to cover any overshoot.

The original circuit is better for coping SSB with a lot of QRN. You can =
hear a weak signal behind a stronger SSB signal/spatter because of the =
faster decay of the fast time constant circuit. But, for casual rag =
chewing it's very fatiguing. I prefer the long time constant circuit. =
Try SB8 without the 680K R180 resistor.

Dennis Brothers videotape places a 100 ohm resistor in series with the =
top end of C235 (a .01 which is connected to the plate of V16B), and =
adds a .05 from the screen (pin 8) of that same tube to ground.

KWM-2 AGC time constants use a dual or stacked time constants consisting =
of a fast section and slower section in series. The slow time constant =
charges to the average AGC value for the incoming signals, the fast time =
constant then acts on short noise pulses as well as voice peaks so that =
average AGC voltage is not overcharged by a pulse minimizing the amount =
of information to be missed. Similar to the ALC. The fast time constant =
should be 10% of the slow one, if you look at the values on the S-line =
and KWM-2 you will find that the fast constant in much faster than that. =
I think 3 or 4 %. Increased the time constant values to this ratio and =
they do work better. The M2's AGC is superior when copying weak signals =
with strong off frequency QRM and when lots of thunder static is =
present. It's not for round table QSO with strong signals present. The =
AGC is compromised towards weak signal performance in the presence of =
interference. (fast - 680k & .01uf, and slow - 1.5M & .47uf. Fast s/b =
10% of the slow value). Change C256 from .01 to .068uf and R180 from =
680k to 47k. - No pumping on SSB sigs., less audio distortion. Test: =
Jump across 680k and hear difference.

When aligning the KWM-2, monitor output with a Scope or a VTVM/RF Probe. =
Do not rely on the panel meter. Bad trimmer caps will not peek and =
require opening and cleaning.

When aligning the 80 meter slug-rack, be sure the slugs are not tuned to =
the Varible IF frequency. If necessary, connect a signal generator tuned =
to 3.7Mhz to XV5-2 in order to peak the slugs properly.

To detect RF feedback (audio distortion), monitor SSB quality with =
separate receiver.

ALC problem. Transmit: TUNE/LOCK, OK, 100w out; All other modes (LSB, =
USB, CW), zero RF output, ALC meter reading at full scale+. Short the =
ALC jack to ground. This will allow 100w output in all modes. Measure =
the ALC voltage when unshorted (around -12v). should be -1.5v. C157 is =
the most likely culprit if DC voltage at V17-1 is -45v; not zero [as =
specified...]. Replacing C157 cured the problem. Another case, in the =
final analysis, electrolytic capacitor (C6B) in the cathode of the Tone =
Osc circuit (V11-gassy grid). Disconnecting that portion of the dual =
electrolytic can (25mfd at 25vdc) and substituting a similar value cap =
solved the problem.

Collins reports high failure rate with 6U8 in XV-13. Tube goes soft =
(gassy) in short time. Tube can be switched into another position and =
operate normally.

ALC problem. Transmit: TUNE/LOCK, OK, 100w out; All other modes (LSB, =
USB, CW), zero RF output, ALC meter reading at full scale+. Short the =
ALC jack to ground. This will allow 100w output in all modes. Measure =
the ALC voltage when unshorted (around -12v). should be -1.5v. C157 is =
the most likely culprit if DC voltage at V17-1 is -45v; not zero [as =
specified...]. Replacing C157 cured the problem. Another case, in the =
final analysis, electrolytic capacitor (C6B) in the cathode of the Tone =
Osc circuit (V11). Disconnecting that portion of the dual electrolytic =
can (25mfd at 25vdc) and substituting a similar value cap solved the =

Can't turn the AF gain control up past 9 o'clock without getting severe =
feedback through the speaker? Most often caused by an open bypass =
capacitor. Start with the receiver squealing and bridge each bypass cap =
with a known good one, Use a .05 @ 400 V starting in the audio section =
and working my way back through the unit. Note this is a dangerous =
process. Be sure to take adequate precautions, rubber mat or gloves and =
one hand only in the gear, the other safely in your pocket.

Low crystal injection on 10 meters: Replace 6U8A, V13.

Low power output: Check screen resistors, R146, R69 and R70.

Scope indicates sidetone present in receiver's CW mode. See SB-10. If =
tone is on TR275 and 200/R200 volt line, pull big relay to reduce tone =
level. If it reduces level, but is still found on 275volts at R50, check =
C49 at V2B.

Final Amp Compartment:

Check PA grid drive for more than just voice signals. If mixer balance =
is not good, and the preselector stages have been aligned more towards =
the oscillator frequency than the signal frequency, more "oscillator" =
signal will appear on the PA. Check: Does detuning the preselector in =
both directions from resonance have the same beneficial results? If so, =
oscillator leakage is NOT the problem, but bad ALC is the problem. Check =
all caps for leakage. You should be able to drive the ALC meter to peg =
without  clipping the PA.

Try: Gould Allied Control [100 Relay Road, Plantsville, CT 06479] - K2: =
TP163-4C-115vdc, 15k, $6.80, and k4: TP163-6C-115vdc, 9k, $6.85.

If plate tuning at 28.6 MHz does not dip to approximately 9 on plate =
tuning scale, close the 10-meter air coil to bring it to this point. All =
tuning and neutralization should have been accomplished prior to any =
adjustments on this coil. If the dip is shallow spread the coil =
slightly. If the dip is high, compress it slightly.

Low output on 10 meters is sometimes caused by the parasitic suppressor =
coils L15 and L16 having their turns too closely spaced. Separating them =
too far can result in parasitic oscillations.

Secure tube socket screws and Driver, and check V8 tube shield is secure =
[Before Neutralizing].

(SB-6?) Check PA screens resistor R148. S/b 820 ohm 2 watt drifts [from =
heat] to 1400 ohms. Replace with 820 ohm 5-7 watt.

Match 6146's via resting/static plate current measurement, mike keyed. =
Disconnect one plate cap at a time after shorting to ground. Match to =
within 10 mills or less, with the top cover secured.

Grid current, but no power output and one can dip plate. ALC is OK. =
Check 10 meter load trimmer, C146 for short. This cap affects all bands.

Inspection hole: Lower right hand side. For contact Burnishing: 3/4" =
diam, 2-1/2" from the top cover, 1 1/2" front, exposing the relay =

If you are at min capacitance and cannot load, then loosen the trimmer =
cap that is coupled to it on the band in question. In "Lock Mode" then =
raised the Mic gain just until its starts to produce grid current. Then =
they adjust the trimmers for 230ma plate current at dip.

But the problem is that it wont fit because the diameter of this cap is =
1 3/4" whereas the diameter of the original cap is only 1". A =
Multi-section, Twist-lock electrolytic Capacitor... Capacitor, Fixed, =
Dry, Electrolytic: triple section capacitor; 40/40/40mF, 250VDC 3 =
twist-lock gnd lugs, 1" Dia., #138-1764-00 Antique Radio, 480 820-4643, =
they carry the can type filter caps. As long as it fits in the 3 chassis =
slots holes you can use a larger value cap and/or one of higher voltage. =
AR has part number C-EC40X3-525, a 40/40/40 at 525dcv, at 1 3/4 x 3 in. =
The other solution is to bore out the hole (do this neatly) ugh! and add =
a 'twist-mount can mounting plate' PN: S-H120M to accommodate the wider =
1 3/8" can. Because the plate will cover the hole you will still have a =
neat looking repair.

Neut Circuit [32S-1], C62 is the main RF return for this section, if its =
missing you wont be able to tune L16 or for that matter L11 shouldn't =
tune either, if you look at the pictorial you will find it at the last =
section of the band switch next to L11, its not a feed through, looks =
like silver mica to me in any case without it you wont get much output.

Neutralization: In the KWM-2 there should always be 2 nulls (or peaks) =
when adjusting those Erie ceramic trimmers.  If you look at the =
construction of the trimmers, you'll see that they turn thru their =
adjustment range twice in 360 degs.  They'll go from max to min and then =
back to max.  If the value that is required is somewhere in between =
(where it should be), then you will see 2 nulls (or peaks, whatever the =
case may be); one between the max and the min and once more between the =
min and the max. So, if you DON'T see the two peaks or dips that =
particular trimmer has to be repaired or replaced. Hopefully, you are =
seeing two identical (as well as distinctive) nulls.  If they are close =
together, that means that the value required for the null is closer to =
either the max or the min.; than in the center of the adjustment.  =
Choose one or the other.

C123 Replacment: Remove orignal C123. through the orignal hole, install =
and center a bare wire connected to a terminal post nearby. Connect a =
standard 0.001 mf 1KV disk ceramic from that post to ground to produce =
the same effect as the orignal feedthrough.

Add a 12vdc 4" muffin fan over the PA cage. Run them both from the =
516F2's filament supply and an external bridge rectifier and a 100mf =
across the output of the bridge. The lower DC voltage (8VDC) lets the =
fans run very quietly, yet they do move allot of air. 12vdc fans is a =
great idea. I suggest using a single diode and 1000uf cap. This gives =
about 9vdc which slows down the fan slightly so that it is totally =
silent. AES has small 12VDC fans for $4 each! (S-FAN02).

New Tune-up Procedure...

	Peak the receive preselector in receive
	Put the emission switch in CW and the meter to plate
	Close the cw key
	Peak the drive control for maximum reading on the meter
	Dip the meter with the loading control
	Advance the mic gain until the plate current no longer rises
	Dip the plate with the loading control until 230 ma


While still in the tune position, peak the exciter tuning for max grid =
drive switch the meter to read plate current dip with PA tuning knob Go =
to "lock key" and you will find the PA is already dipped.


Sponsored by the Collins Collectors Association http://www.collinsradio.org
Nets: Tues: 3.805 Mc-2000 Central / Thur: 3.875 Mc-2000 Central
Fri: 3.895 Mc-2000 Pacific / Sun: 14.263 Mc-2000 UTC
1st Wed (of the month) AM Net 3.880 Mc-2000 local (ET, CT, MT, PT)
Sun AM Net: 29.050 Mc-1200 Central

This archive was generated by a fusion of Pipermail (Mailman edition) and MHonArc.