During the Winter of 1994, I came across an R-390A Receiver with MSR-9 Sideband Adapter. I liked the performance of the equipment. Very well built, straight forward gear. Anyway, I was talking with fellow hams, WA2NPL & KS3K as to just what transmitters were used with this receiver. They both said "well, you probably want a T-368". Roughly seven months later, a Barker & Williamson T-368/C S/N: 018 was purchased and few weeks later, it arrived at my QTH. During the usual D&C it was decided that over all, some aspects of my T-368 would be redesigned/repaired/replaced. Before any of that could even be thought of though I first had to think about safety... and if you want to procede with any of this you need to think about it as well starting with...
HV (High Voltage) safety protocol can not be stressed enough! Just because the green and red "Power" indicator lights are not on does not mean that the T-368 is off and safe.
When the "Filament Power" breaker toggle is up in the "On" position, various other HV supplies are operational in the T-368. This HV is present even with the "Plate Power" breaker in the off position and the drawer interlocks open... It is substantial and can hurt you bad!
Now for the really bad news... With the "HV Power" breaker toggle up in the "On" position, the T-368 Plate HV supply can be energized without the red power indicator light on as the plate transformer relay's contacts can weld shut in the on position. This has happened here and is the very reason the various "interlocks" open the plate transformer primary wiring.
It is well known that all four of the plate supply "interlocks" on the various drawers and back panel of the T-368 can be a source of erratic transmitter operation. Keep in mind that the previous owner may have "fixed" this problem by placing permanent jumpers across the interlocks so that they no longer disconnect plate supply primary power as intended.
If you manage to get across an energized T-368 Plate Supply you will probably be dead or crippled with little to no medical recourse!!! Early last winter (1998), a fellow hams finger was nearly cooked off his hand in an instant! This painful injury was indifferently provided by a tabletop SB-200. As always, familiarity breeds contempt and casual disinterest... a formula for severe injury. Keep this in mind when poking around in the T-368... and keep that left hand in your pocket too!
Also keep in mind, T-368 cabinet must be bonded to your buildings power system ground. At the minimum, a #12 three wire cord (with full size #12 bonding conductor) must be used and connected to a properly functioning 3 wire outlet. A #10 wire would be a better choice but will not fit in the connector properly.
As with almost all things, there are no guarantees but should something fail internally, this may help prevent the T-368 case from being 2700v above ground. Be aware, the shack's RF ground will probably be insufficient for this purpose.
Power requirements. For maximum performance you will need a stiff AC supply. 115v, 30 amp service with bonding conductor is the minimum and a 40 amp service would be better still.
If you have any doubt, stop and get knowledgeable help.
On to the "RF Deck":
The original plate choke and .002 "brick" type bypass cap were removed from the "low impedance" point between the PI and the L in the PI-L circuit. A National R-175A 1 amp plate choke with 500pf 30kv doorknob by-pass was mounted a couple of inches to the right of the 4-400 and the modulated B+ wiring fed there. The previously removed "brick" by-pass was reused as the 4-400 dc plate blocking cap for the tank circuit. This capacitor worked well in this "electrical" position as it is RF current rated. It is allowed to float with the wiring so as not to mechanically stress the 4-400 during thermal changes.
All this plumbing got a great deal of capacity off the modulated B+ and removed 2700 volts from the tank coil and band switch. For HV safety reasons and the fact that it would be a pain to remove, the existing .0039 blocking condenser (a 3" x 3" Sagamo transmitting tub) was left in place as added protection in the event that the 4-400 brick blocking cap developed a leak.
Further Background: Some T-368s will have 3 brick capacitors in parallel in place of the tub at this position. Various military MWOs remove all three and substitute the tub as a service upgrade. If you are planning to keep your T-368 "stock" you might still want remove the three bricks and install a tub as the bricks are prone to catastrophic failure in this position.
The "Plate Current" meter was rewired to read cathode current. This put the meter movement/case at nearly ground potential. Since there were many stories of blown meters due to flash over at some point on the deck, two 200v, 200a diodes were wired across the meter input terminals parallel back to back. Hopefully these will save the meter movement in case of such an event. Since it now monitors cathode current, the T-368 operator has to add about 40ma (screen and grid leak current) to the meter indication and this value will change slightly with operating frequency. This mental addition has not proven to be a problem even after extensive cold 807 consumption.
With regard to blown meters, what probably was happening was that when some "event" caused RF to arc over some place on the tank, the ionized air in the RF arc provided a low impedance path to ground allowing the power supply capacitors to wholesale dump their stored power taking the meter with them. The plumbing change in the first paragraph should prevent this sort of thing from ever happening again. Still, the diodes are cheap protection for a increasingly difficult to find part.
The 6000 IPA tube protection circuit was not working properly so the poorly functioning 6AH6 & OA2W combination were removed and a new ones installed.
The "Loading" adjustment stop washers were found to be damaged and were replaced. These washers are used as adjusting knob turn counters and prevent the vacuum variable capacitors from running to their internal adjustment limits.
Replaced the exciter and fsk cables. These were cut out during a previous owners "de-milling" process.
Replaced all missing Type C to BNC adapters used to port the PTO and 4-400 grid to the front panel.
Replaced a missing rack spring on exciter rack.
Cleaned and lubed the 4-400 blower motor.
Some background for the next few paragraphs: The T-368 4-400 is fed from two B+ supply points... one that is modulated and one that is not. The modulated feed goes to the 4-400 plate and is fed modulated B+ from the mod iron. The un-modulated feed goes to the 4-400 screen and is fed B+ from a point above the mod iron.
A previous owner mod had the 4-400 screen being fed from the modulated B+ supply rather than the stiff B+ supply. This change was implemented by that owner to improve the modulation percentage. Transmitters without clamp tubes are wired this way and perform very well.
After some thought I decided to abandon this mod so as to remove the additional dissipation current demands of the 5893 clamp tube and the 4-400 screen current from the mod-iron secondary. Why add to mod-iron stress and besides, limited but useful self-modulation of a screen can be had with the addition of a large inductor in the screen supply lead between the screen and the screen loading resistor.
I found on my shelf a Stancor C-2308 HV filter choke. Its good for 8hy at 3kv at 300ma. Mechanically, the choke was installed on the large perpendicular center shield above the exciter unit and as close to the rear of the RF Deck as practical. Electrically, the choke was installed in series between the 5893 plate connection and the 4-400 screen connection. A 10hy or more choke would have been better but 8 was all I had available at the time and at least partial modulation recovery was better than none.
Replaced the tank coil as the old one was broken where it bolted in on the left. This is a somewhat common problem as the ceramic form is supposed to be lightly captivated by two interference fit nylon screw/nut combinations. This way they can be adjusted to allow some freedom at tank form mounting point. If too tight the form will break when the "RF Deck" flexes such as when lifting it.
It is interesting to note that a previous owner insulated the left side (side closest to antenna x-fer switch) of the tank coil chassis mounting tangs with heavy rubber. I think arc over was only possible at this point if you suddenly lost you antenna or the SWR went very bad. It was probably quite necessary when B+ was on the tank coil but made pretty much unnecessary with B+ removal. I was told this was an "ER" recommended mod and the mod was done in a professional manner that looked kind of factory like so it was left in place.
Remount RF parasitic choke to its proper location on plate of 4-400.
Refit the RF Deck's RF output transfer spring for better contact and alignment with RF pickup shoe mounted on antenna x-fer relay. A fair amount of pressure is required as is careful alignment. You want the spring to mate fully across its width with the antenna x-fer relay pickup shoe. These are silver plated contacts and wiping with Caig Contact Cleaner just prior to closing the deck will prolong your next visit. Poor contact here will cause tuning/loading problems during tune-up and will sometimes hiss and squeal at full power! It will also cause unusual/unsteady IPA (6000 tube) plate current readings such as high 6000 plate current, low 4-400 grid current.
On the "Modulator Deck":
A few bad caps were changed but the low level speech amp got something of a physic.
Since the low level amplifier had the been modified many times, a decision was made to gut it. The various clippers and filters and extra stages of amplification were removed and the leftover chassis openings blanked off. Then a standard two stage 12AT7 mike pre-amp/volume control circuits were coupled to the 6C4 audio circuit. An additional BNC connector was installed on one of the blanks so as to permit plug in of the front panel volume control for use between stages of the 12AT7. The existing negative feedback loop was kept, had part additions and made adjustable. Additional input volume control cabling was performed.
Phasing was adjusted for max. modulation. If you intend to use different microphones a phasing switch should be mounted on the front of the low level amp for easy access with the modulator drawer open a few inches.
Since you will probably not have matched pairs of 4-125 modulator tubes one will almost certainly be "hotter" than the other. You can swap them between sockets for best modulation percentage... a kind of poor man's negative peak clipping. Its not a lot but every bit counts. Broadcasters been doing this sort of thing for years.
Replaced "Keying" switch. Unbelievably, the installed switch was a 30amp 3PDT. This massive switch was used to toggle a bias line! Reminds me of that old fart saying that an elephant is actually an ant ordered to military specifications.
Poorly regulating VR tubes for the 4-125 tubes were replaced.
Bad cap in the antenna transfer relay timing circuit was replaced.
Non-military wiring was removed.
Non-military Amphenol mike connector was removed and rewired to the existing military mike connector on the front panel. A military pilot light was installed in the connector's place and wired to show when the rig is in transmit mode at a glance. This has actually been very useful during break-in operation and should be useful during RTTY as well. Yeah right! Well, I keep telling myself its useful...
On the "Power Supply Deck":
The power supply deck received a couple of non PCB 65mfd 2000v caps in wired in series with two .5meg compensating resistors wired in series across each capacitor. Theses caps were the exact same size and dropped right in! This additional capacity created problems with the 3B28 pair and the 20 second time delay. Sometimes when going into transmit, the 20 second delay would restart due to severe line drop. I found out quickly that there is a big difference between charging 8mfd and 32mfd at 2700 volts.
A suitable 30amp relay was obtained and fitted with an adjustable R/C timing combination. This and the "Tune" resistance already in place were the basis of a "step start" system. The relay timing circuit supply was connected to the main filament transformer feed. This feed is controlled by the 4-400 filament voltage rheostat. Since the filament voltage control is always set for 5 volts on the front panel, it is a relatively accurate voltage source. Using this controlled power source and once set, the step start timing will never vary.
At WA2NPL's insistence, the HV choke and an additional 100 ohm, 100w resistor were inserted in the plate transformer return lead. The resistor acts as an additional "light transient" cushion for the B+ circuit. I know that there are ops out there that will say this resistor addition will result in less than "ideal" HV regulation. I would say try it and see for yourself.
With the HV choke in the return lead it is at ground potential and should now last forever.
Since all this was done, no more dead 3B28s and no more 20 second delay restarts. Oh yeah, no more blinking lights at my neighbors house either!
Removed a non-military 50 ohm 100 watt sliding resistor in the AM B+ protection circuit and replaced with a 25 ohm, 25 watt wire-wound rheostat. This brought the AM B+ protection circuit control back out to the front panel where it belonged.
The "Cabinet":
The antenna transfer relay needed repair as the receiver signal would go up and down due to contact resistance. The fan was shock mounted to isolate it from the back panel and a powerstat was installed (by the previous owner). This 'stat was used to eliminate some annoying fan speed related cabinet resonate noise problems.
Replaced the worn-out "Thermal Reset" reset button with a "like new" red anodized one.
Replaced the 4-400 thermal protector... it was literally burnt up.
Installed a molex connector in fan wiring for quick rear panel fan motor disconnect.
A further note on "Interlocks":
As received, all four plate primary HV interlocks and one control voltage interlock on my T-368 were disconnected. A decision was made to reconnect two of the interlocks.
The first to be reconnected was "S14". It is the one on mounted on the upper right front corner the "Enclosure" for monitoring the "RF Deck" position. It prevents "Modulator Deck" control wiring from coming up with the RF Deck in the out position... even a tad.
The second interlock to be reconnected was for the power supply deck. This light pressure interlock is one sorry design. To help with the connections conductivity, I used cu/al grease designed for coupling copper and aluminum house wiring. It contains very sharp, microscopic, sintered zinc particles which bite through contact oxidation... another good thing.
With this interlock working, there is no danger of the HV supply accidentally coming up with the deck out and the HV output shorting bar down in the shorting position... a very bad thing.
Final Assembly:
The rig went back together and was working shortly thereafter. The exciter needed realignment and replacement of leaky caps causing blocking bias problems.
Some fooling around with the 12AT7 equalization was necessary for use with my EV-664 mike and my particular voice.
Installed IREC shields on all tubes.
Additional Repairs Remaining:
(1) Complete the installation of MWO-3. Some T-368s from the beginning through the "C" model have PTO "leakage". That is, when tuned to the same frequency as the receiver, you can hear the T-3 PTO leak a small but annoying "signal" which will sound like a "birdie" in the background of a desired signal. The military devised a repair/upgrade whereby the PTO 150v regulated line is switched with the antenna changeover relay.
Some T-368s either have all of MWO-3 installed, a partial installation or none at all. I am somewhat lucky as my T-368 has MWO-3 partially installed. The modulator deck additions and wiring harness additions were completed during an earlier military sponsored trip to Tobyhanna. For reasons unknown, the RF Deck "Function Switch" wiring part of MWO-3 was not included on mine. When completed, MWO-3 should do what its supposed to do.
MWO-3 Update: My RF Deck does indeed include MWO-3 modifications. For some unknown reason, pins 16 and 17 on the deck's "Cannon" connector were connected together with a very short run of blue wire. A jumper at this point will allow the RF Deck to operate with an unmodified T-368 cabinet or an unmodified modulator or both but without MWO-3's benefits. The jumper was cut and the rig works as anticipated... no more birdies on receive!!!
Click HERE to view a scan of the schematic from TM 11-809-35 Change Order C-5 dated 5 October 1965.
Click HERE to download a larger scan suitable for whatever with Paintshop, ect...
Click HERE for a further concept of MWO-3.
Be aware, the Change Order is not the MWO. It does not walk you through a modification. I have yet to locate the actual MWOs for the T-368 series. I can only figure that they were carried out and promptly tossed in the round file.
Note: Change Orders are long winded errata sheets intended to modify the -10, -20 and -35 manual set usually upgrading all manuals at once. The idea was that a Tech was "supposed" to write into a manual set by hand, all of the various informational changes contained in each Change Order. There are multiple Change Orders from multiple years. If all corrections from all of the Change Orders were entered in the manual set well... I just don't think anyone's penmanship is that good... Must say that I have yet to see an upgraded manual with all the B/S penciled in...
(2) Install new bearings in both fan motors.
(3) Install "protected thread" weight bearing screws on upper rear case ledge as per "E" and "F" units. Drill rear access panel per protected thread screws. This will allow for retention of the rear panel in it's installed position with all rear panel locking screws to be in their unlocked position. With this addition, the back panel is still supported and can then be removed/installed without difficulty. Especially when connecting/disconnecting the fan.
(4) While not an "operational" issue, the T-368/C model was manufactured with "drawn" or "stretched" steel front panels containing rounded, drawn corners with control/function lettering stamped into the metal surface. My particular T-368 has an assortment of different panels including one with folded, welded corners. All panels contain different stamped lettering fonts and one panel is silk screened. The welded, silk-screened panel is from an "F" model. I was able to get three "correct" panels, all with drawn, rounded corners and all with the same stamped lettering. They panels will be going out for painting in early summer.
Also of note, the existing military plate choke was placed between the PI and the L in the tank circuit to allow the broad-band continuous tuning from 1.5 to 20 MHz without any kind of switching or extraordinary choke design with inherent mass/ballistic problems. I haven't tried as it is unnecessary (and illegal) but this rig is probably a ham bands only rig now as the amateur plate choke will almost certainly have out of band problems.
The end result? A rehabed T-368 is the "Harley-Davidson" of non-broadcast, transmitters and certainly one of amateur radios better kept secrets. Operation is totally simplistic. Audio reports from people I know say "sounds just like you". You can "swing that monkey" all day everyday and have no trouble at all. Initial run up was done on a 2kw dummy. 1 hr, key down, no problems. Old buzzards, this rigs for you! Maintenance has been little more than a yearly air-filter change and tube test. An added bonus! The T-368 has a few safety features which can prevent damage to the both of you. It is the rig of choice here...
And, if all this good stuff was just not enough, in the winter, your T-368 can double for a shack heater. -Grin-
Future:
Mechanical RTTY is coming to OTT HQ. An MD-239 RTTY Modulator has been obtained and will be coupled with the T-368 this summer. Should be interesting.
Bruce, KG2IC has a webpage devoted to the T-368... a very good read indeed.
KG2IC's T-368 Page...
April 12, 1999
(c) Ott Kazott
