Warning: Creating default object from empty value in /homepages/11/d132647312/htdocs/Amfone/mkportal/include/SMF/smf_out.php on line 47
T368 power supply question

The AM Forum
September 17, 2019, 01:25:30 PM *
Welcome, Guest. Please login or register.

Login with username, password and session length
   Home   Help Calendar Links Staff List Gallery Login Register  
Pages: [1]   Go Down
Author Topic: T368 power supply question  (Read 3223 times)
0 Members and 1 Guest are viewing this topic.
Ron NU6F

Offline Offline

Posts: 12

« on: January 23, 2019, 04:42:22 AM »

I just acquired and did repairs on a non workingT368E that had some severe arcing and runaway plate current on unkey that tripped the plate breaker instead of the overload relay .

There were several issues throughout, but I have gone through all shelves, tested all major components, re tubed and tested.

Reset the overload relay pots and tested ok now.

Replaced the 4-400 which ran off with plate current in the operate mode.The clamp tube pulling the screen volts to 86v was still not enough to stop it.

Putting in a new tube, all is fine.

The transmitter is back working and has been on the air.Good audio reports.Fantastic CW rig where most my operation is.

But my abnormality is....Dummy load or antenna, When I load the transmitter up at the 250 milliamp point, I get 600 watts!!!

At 320 ma I get 710 watts out.!
I assumed it was my wattmeter as it was impossible to get this output with 2500v plate volts?
Nope, Bird 43 showed exactly the same.

I thought, well, maybe what I have some hellacious spurs.

No, no grid current, increased plate current or rf out without drive.

So. It had to be plate voltage.It took me a while to rig up a suitable voltage divider to measure things...

But yes, I have 3100 to 3200vdc from the plate supply..The caps, choke and all checked fine when I tore down the power supply.
But I remember noting that my T9 HV secondary resistance reads were 160 and 180 ohms to center tap instead of 110ohms.
Also my transformer is marked 3180-0-3180vac It is the original transformer, same make as all the rest..

My line voltage under load holds at 123 vac instead of 117vac as the transformer nominal spec.

Not enough difference to explain such a high plate voltage supply.

When I back off plate current to 160ma, I can get down to 400 watts and then I can safely operate AM.
But again. To get clean  near100% modulation on my scope, the mod amp peaks can be no higher than 140 ma, not the manuals 230ma.

Normally, this would not be an issue as both the 4-125 and 4-400 plate and derived screen voltages are within limits for the tubes.

But the other day, I tuned up to where my antenna is almost 3,:1.The pi network had no problem loading up to 400w am carrier.

As I was bringing up my mic gain to 140-150ma on voice peaks.....

MAJOR ARC, apparently in RF deck.

Pulled shelf, no noticeable damage.

Over on dummy load,no problem.So, I knew insulation levels were stressed with this 3200 v, especially with 100 watts reflected back into the tank.

One more test on the antenna and on my second transmission..I backed off on my mic gain a bit more.It was ok for a few minutes.....KABOOM again. Reset breaker. Tuned into dummy load ok.

So, I have had the rf deck out.I made a nice 0-10kv DC HiPot tester to test all the wiring, insulators ,feed throughs , coil forms ,spring contacts and caps.

At 6100-to 6400 volts, there are three locations that arc over.

Since I have up to 3200vdc plate current....in AM, I will likely be at over 6000vdc on peaks on the 4-400 plate and the tank circuit, chokes and bypass/coupling caps.

I spent the day and made some very nice lexan plastic guards for the three locations. and eliminated the arcing.

I went to 7100vdc with no leakage or arcs.

But the rf choke bypass transmitting mica is really only rated at 6000 vdc.
It showed no leakage at 7100vdc. But at being 60 years old.This isnt a safe practice for it.

The big Sangamo 0039mfd rf coupling cap also is only 7000vdc rated.

Likewise though, it showed no leakage.

I am now safely operating with no arcing,

I have spare caps for those in the PA tank cct if worse came to worse.

But I am thinking that I can probably add a 20 volt  16 amp "buck" transformer to the T9 plate transformer 125 vac primary and get the output 20%lower and closer to the nominal 2500vdc.

Has anyone else had this issue of excessively high plate volts.
Sorry for the long post.

My unresolved issue

* 20190122_224554.jpg (3794.34 KB, 4032x3024 - viewed 178 times.)

* 20190122_224712.jpg (3669.15 KB, 4032x3024 - viewed 158 times.)

* 20190122_224426.jpg (3597.83 KB, 4032x3024 - viewed 137 times.)

Offline Offline

Posts: 2096

« Reply #1 on: January 23, 2019, 10:49:08 AM »

If some turns on the primary winding of the HV transformer are shorted, it could cause the problem you are describing...If the output voltage continues to climb, additional primary windings may be shorting.. 
  Wouldn't you want to get the voltage down to where it is supposed to be rather than beef up the tank to handle the additional voltage?  Good Luck   Steve KL7OF
Ron NU6F

Offline Offline

Posts: 12

« Reply #2 on: January 23, 2019, 11:46:28 AM »

Hi Steve, thanks.
I hadnt thought of that, but its a possibility.

I never noticed any abnormal in rush "thump" or heating of the transformer.But with everything warm after operation for a few hours, I might not notice it.

Well, ,I have no experience with the T3 and thought that others had experienced seeing a wide variability in the plate voltage.

Before I purchased it and did resistance reads and noted that the HV secondary resistance reads on each half were 150 to 160ohms, instead of the manuals 110 ohms.
Primary resistance reads were normal.
But with such relatively low resistance, the few shorted turns would not be visible.

I have thought of re installing the plate transformer primary opening jumper,J13 on the cabinet.

It would be a VERY convenient point to add a "buck" transformer of 15-20 volts rated at 15 amps which would drop that secondary voltage accordingly.

* 15482622909051984179968490446000.jpg (3448.62 KB, 4032x3024 - viewed 175 times.)

Offline Offline

Posts: 1765


« Reply #3 on: January 23, 2019, 12:52:25 PM »

It would be a VERY convenient point to add a "buck" transformer of 15-20 volts rated at 15 amps which would drop that secondary voltage accordingly.

Yes, but if you get the phase wrong (50-50 chance), then you add instead of buck...KABOOM!!

If there were a shorted primary turn, the primary current would be very high even if the secondary were unloaded. A shorted turn would do that.

The higher secondary resistance might suggest the turns ratio is off (too high). Make it safe, and apply something like 12VAC to the primary, and measure the secondary voltage. The voltage ratio is the same as the turns ratio.

Keep in mind there are the HLR Mods for a T368. As I recall. this includes getting the B+ off the output tuning network.

"By removing the DC from the tank circuit, a possible source or arcing is eliminated. Further, the total shunt capacity is now about 1000 pF compared to 6000 pF in the original circuit. (This improved to the frequency audio response drastically, Ed., HUZ)."

Would be nice to be able to "let her rip" at full plate voltage without concern of flash over. An extra DB or two in RF power output is negligible, and IMHO not a worry. Could easily run 600 watts carrier and not exceed 1500w PEP...just depends on how hard you drive them modulators. An occasional peak beyond 1500w is a "who cares" so long as it is clean.  Wink


Offline Offline

Posts: 494

« Reply #4 on: January 23, 2019, 01:45:26 PM »

VSWR on big transmitters can create big problems! That power got to go somewhere. I never used SWR myself but always looked at forward vs. reflected power and try to keep the reflected power as low as possible.
Are you running the solid state replacements in the power supply? If so try going back to the tubes and that will reduce total plate voltage.
A lot of the stuff built back then was designed around 115/120 volts and modern 125 volt distribution can be iffy. How are your filament voltages?

Ron NU6F

Offline Offline

Posts: 12

« Reply #5 on: January 23, 2019, 02:13:06 PM »

 Yes, it was obvious the peak voltages built up quite a bit.
But all the crazy power output reads led me to make up a voltage divider to measure that B+
No, solid state. Original 3B28's or I would have had an even worse situation.

I never go to solid state unless it's a supply I am building for a rig.

The :soft" current limiting from  the rectifier tubes is a necessity.

Well, the only filament voltage I've looked at I'd obviously the 4-400.
The more I think about it, I suspect the problem is likely shorted primary turns on that plate xfmr :-(
Once I rewire in that primary jumper plug in the cabinet,I will pull the 3b-28 caps and have easy access with a clamp on ammeter to see what idle excitation current that beefy transformer has.

I'm afraid I may be looking for a plate transformer.

My "bargain" of a T3 may come back to bite me.

Well, at least I saved this one from the scrap pile.

I may be posting a "want"  for a new one.

Perhaps you regular users of the T3 could keep an eye out for me.

I have no idea whatsoever  how hard they are to find, and what the down payment and monthly payments would be.


Offline Offline

Posts: 494

« Reply #6 on: January 23, 2019, 04:48:34 PM »

Are you using a digital or an analog meter? 20K per volt was what all the readings in the book were determined with. I never use digital meter in high RF environments because they can get flakey.
On T9 what taps are you using? Thatís the primary side of the HV transformer. Are you getting the same resistance between resistance between 1 and 2 and 4 and 5? They should be equal both being 120 volt windings. You will have to brake the strap between 2 and 5 to check this. If one side has an issue they wonít be equal. Somewhat unlikely that both sets of windings have an issue.
Think they strap 1 and 4 together along with 2 and 5 for 120 volt operation. If 1 and 5 and 3 and 6 think thatís for a lower voltage input and will give high voltage when connected to 120, somehow just canít see a shorted primary but maybe that happened.

Ron NU6F

Offline Offline

Posts: 12

« Reply #7 on: January 23, 2019, 09:46:18 PM »

Thanks for the response.This one is the E model, and as such,it has the single 117vac primary. Not even a tap for 100v or 117 like a few of the other transformers.

Sadly, no options like on the original non lettered version.
I was wondering if anyone had t as Ken actual measurements on their B+.

This T3 looks as if it was made with a mix of modules of different erasers.

But ,I did spend considerable time checking all relays, transformers and caps throughout.
Including building a HIpot tester that gives me 0-10kv DC,  current limited to 10 ma .

It was non operational when I purchased it.

The last issue he had was runaway plate current in key up position during the delayed drop state that holds in B+ and the antenna relay from .5 to 9 seconds after you release PTT.

The other issue he had was several times, a large BANG from an arc under various conditions.

I attributed the arc to the fact that the plate tank tap to the bandswitch 3-6mc position was missing its locknut.
The tap was just resting on the bandswitch lug.

Undoubtedly ,an open tap at full power would have built up very high voltages.

One of the 3B28 filament leads were unsoldered in its socket.

Per the owner, it had never been changed.

At any rate, I had no luck at the Electrical Suplly house here in Reno in finding a female 3 prong 120 v, 15 amp twist lock female ,cable end socket to make the shorting plug for J13 on the cabinet.

The previous owner just jumpered it out.I probably would have done the same.

But NOW I could really use J13 returned to it's normal state.

It is a very convenient place to insert a variac to adjust the B+ in the radio.

I dont want to have to drill any holes or "hamsterize" anything to replace J13 with some other connector.

Modern 3 prong twistlock flanged plugs are much larger than J13, so I cant go that route.

So, I will keep scrounging for what I need.

I am including a photo of the transformer I have in mine.
Ot is rated as 3180 volts each side of CT as opposed to 3080 v in the manual.

There might be the slim possibility this supply was not a standard supply. And for some other variant that used a higher plate voltage.

* 15482979285423948855789308965404.jpg (3704.83 KB, 4032x3024 - viewed 171 times.)

Offline Offline

Posts: 366

« Reply #8 on: January 24, 2019, 02:27:42 AM »

Running 123 vac on a 117 vac primary, depending on the volts per turn, will drive up the secondary by quite a bit. Another thing is that the current draw, even at no load, will increase alot.
Ron NU6F

Offline Offline

Posts: 12

« Reply #9 on: January 24, 2019, 04:21:25 AM »

 So, have you seen any variation from the standard 117 v to 6160 volt ct transformer in any variant of T368?
The manual states, at that value , the DC output at the cathode of the 3B28rectifiers should be 92470vdc.

My transformer is marked 115 vac in for 6360vac ct output.

So, even being so, with my 124 v nominal power, that's only

We are applying an input that is 7.8% higher.Therefore, that would only account for the secondary to be at 6860vct, or 11% higher than a stock original transformer operating at 115vac in.

So, the original 2500vdc should come put to 2780 CDC.

Not the 3200vdc I see.

Perhaps someone here is familiar with any variations in transformers for the T3?

Ron NU6F

Offline Offline

Posts: 12

« Reply #10 on: January 25, 2019, 01:45:45 AM »

Thanks for the suggestions from all.

Problem resolved with a 12volt 16 amp buck transformer on the final plate xfmr primary.

I spent the better part of the day testing T9 with a few different setups.

My first concern was the suggestion that I possibly had shorted primary turns ,thus increasing the turns ratio primary to secondary

I pulled the J13 shorting plug and tested with my variac.Neutral from the variac to the J13 blade that has wires 13 and 17 tied down.

Then the adjustable output of the variac through a 25watt light bulb for current limiting tied to the J13 blade with wire 14 on it (goes to tune/operate switch) .

I could bring up the variac output to my full line voltage here (124volts) and. Thank goodness, the 25 watt bulb was showing only 1/2 brilliance at the most.My clamp on ammeter is scaled too high to read such low current. But it did display .1 amps of primary current.

With only its primary connected and all secondaries lifted, a typical good transformer should only absorb 1-2% of its full load power from  magnetizing its core along with and eddy currents in the laminations.

If my primary had even one turn shorted, I would have expected more like 60-70 watts.

I had the plate caps of 3B28's lifted and used a 30k 200 watt wirewound resistor to ground from one of the  plate caps lifted from the 3b28's.

This 30k  resistor has 6 -5k ohm taps from top to bottom.
It's the only way I have to measure AC or DC over 1000v.
It works very well.

With my variac and no light bulb for current limiting. I got the following results for the AC output from the T9 secondary to ground. I got pretty much identical results from either leg of the transformer to ground.

100vac on primary gave 2808 vac
112vac on primary  gave 3180 vac
120vac on primary gave 3360 vac
124vac on primary gave 3460 vac (that's my local line volts)

So, I decided that I would try to reduce the T9 primary volts  by 12 volts .I had a 12 volt,16 amp transformer that would be perfect as an autoformer to buck the excess 12 volts of my house AC.Since this transformer had a 16 amp secondary rating, it would work well with typical primary currents in T9.

J13 makes this very convenient.My J13 was bypassed and I had no luck in finding the matching cable end socket for it. I did find that Hubbel makes a newer version of this recessed plug with a flange.The new version is rated as a 15 amp 277vac .It has the right angle hook on the wide blade like modern twistlocks.

For now,I just ran my three wires out the removed J13 hole and mounted my buck transforner in a box I had.

When the Hubnel connectors come, I can make a nice plug in unit for J13 with no T3 wiring altered.

I have not heard of anyone else with this issue on a different transformer in T9. But hopefully this will help someone in the future.

It appears to have been either original or Depot maintenance that did it.Lacing all in place and all wire markers intact.

If this T3 was used as only RTTY. the transmitter could have run at much higher power, like 650 watts without an issue, no AM mod peaks to spike things.

But in AM, this thing was  overstressing insulation and caps throughout the radio.As  I initially mentioned. Voice peaks were causing major arcing in the rf deck if the plate loading wasnt perfect

Now. I am looking at 2550 vdc on the 4-400 plate instead of 3100v.

A much safer feeling .Hopefully my first T3 will last as long as me ,the ART13, BC348"s, ARC5's and the R390a.


Reno NV

* 15483987154011289571318927755445.jpg (3643.91 KB, 4032x3024 - viewed 152 times.)

* 15483988342282116380457985703799.jpg (3719.72 KB, 4032x3024 - viewed 139 times.)

* 15483988783659045313819193594426.jpg (3913.26 KB, 4032x3024 - viewed 135 times.)

Offline Offline

Posts: 366

« Reply #11 on: January 25, 2019, 01:56:12 AM »

Not knowing what the volts per turn is, you bucking the primary confirms my previous post.

Offline Offline

Posts: 494

« Reply #12 on: January 25, 2019, 09:02:52 AM »

I had a RCA BTA1R that I was running on 160 and I installed a autotransformer on the primary side of the plate transformer and ran it like that for a long time. It was great being able to control the output power using that transformer so I would propose that you keep the autotransformer in the primary. Would be real helpful running the transmitter in reduce power for tube and component life.
In military service and in National Guard service no one ever used AM that was almost always used as a RTTY transmitter. Least I never saw or recall hearing of anyone using it in AM except maybe to set up a channel but AM was dead in military use long before the AN/GRC-26 systems were retired.
There were cases where the modulator deck was removed and just the power supply and RF deck were used to save space but that was never an official modification.
Ron NU6F

Offline Offline

Posts: 12

« Reply #13 on: January 25, 2019, 11:20:06 AM »

I agree that ,with this plate xfmr, it wasnt used much on AM or someone would have addressed the crazy high plate voltage.

But in RTTY, it would have been no problem whatsoever.
3100 volts on a 4-400 with 320 ma is nothing.

But for all I know, it may have been deployed somewhere with much lower line voltage. Pr, it may have been used with a variac plugged into J13. It is set up perfectly for that with a neutral, hot and load side to ONLY T9, so mo other voltages are affected.

But I dont have a 1500w variac available.

And actually the small 12 volt 16 amp transformer is perfect to get the primary volts down.

I really love this rig on cw.It is the nicest multi cw rig I have ever had.Beautifully shaped keying in the grid block scheme.And a bonus to having the clamp tube still allow 120ma of plate current in key up state, is that key clicks are virtually eliminated.

I spent a month making a good 410v regulated keyed supply for th ART13 ,and it's much nicer than stock method now...but the T3 is perfect.

I'm 95% CW and usually only AM once or twice a week for mil nets.

I think this is a keeper now.



Online Online

Posts: 193

« Reply #14 on: January 26, 2019, 06:48:04 PM »

I looked at the salvaged T-368 transformer in my storage shed but it turns out different than yours by quite a bit.  This one has dual primaries marked at 0-117-141 V each so it must be an early model.  I measured the secondary resistance from center tap to ends at 50 and 60 Ohms, which doesn't help in terms of evaluating your transformer either.

Friends who went by Fair Radio Sales a couple of years ago said they had T-368 plate tramsformers outside in the back, but the shipping on one of those would probably cost a fortune.  It sounds like yout worked out a solution anyway. 

Geoff Fors
Monterey, California
Ron NU6F

Offline Offline

Posts: 12

« Reply #15 on: January 26, 2019, 10:27:21 PM »

Thanks for checking!

With my 125v line voltage, the old BASIC model transformer would probably would put out the correct voltage for the lettered models.

And, your result since reads match with the book.

The lettered model secondaries having the higher voltage..

I'm guessing they didnt update the resistance reads.

And of course, in those days, techs were using Simpson 260's or aTripletts.

So 110 ohms or 140 ohms, not much difference on the meter.

Yes. It's working great now and I can concentrate on improving my audio.Picking a better mike element from my spares.

The FET mike that came with the TUG8 style amplifier board doesnt sound that good.

I have good Astatic original crystal mike.

I'll have to remove the FET mike bias and jumper out the coupling cap and give it a try.

Much nicer experimenting than the Tesla Coil I had in the cabinet!!

Ron NU6F

Offline Offline

Posts: 12

« Reply #16 on: January 30, 2019, 11:22:59 PM »


The buck transformer plugged into J13 has worked perfectly on the 2500 volts.

I can tune throughout whatever mode ,power and audio level with no arcing anywhere.Plate voltage at 400watts is 2550.,not 3200volts.

There was an intermittent very noticeable ac line hum that appeared frequently on the carrier in any mode..Especially running the exciter only, and to a much lesser degree when the final was operating at full load.

Here again, it was my high line voltage.

T3, the exciter plate transformer comes standard using the 117vac primary tap. I measured my secondary ac volts at 470 volts each side of c.t. instead of 440 vac as per the manual.

My corresponding exciter B+ was a bit over 500volts .
I moved the tap to 125v (my line voltage is 124v)

My exciter B+ was now 460volts.The hum on carrier was gone from the exciter and from the 500 watts out.

The bias supply transformer has only the 117vac input, but the -500vdc full bias did not cause a problem anywhere and no components are stressed out of limits.
As this is an E model, the PTO osc and buffer filaments are fed by the ferro resonant voltage regulator transformer and the high line volts is corrected.
 All other filament transformers are adjusted when you set the 4-400 to its nominal 5volts.

So, I believe all supply voltage issues for the T3 have been resolved.

A faint hum in AM mode from the D104.In CW clean.
Terminating the 600 ohm audio input on the end of my long shielded accy cable in AM with normal mic gain shows no hum on carrier, so the cabinet audio wiring and the J12 remote cable are not a problem.

My audio sounds crisp. No distortion, but lacking in lows.

From what was suggested somewhere, I increased the 6C4 cathode bypass cap C2 and the negative feedback cap C40 .

I increased the cathode bypass cap from .1 to .56mfd.
I increased the negative feedback cap from .0027 to .004 mfd.
Not very radical, but enough to show an improvement.

So, gain was reduced a bit, but another 1/8th of a turn on the 600 ohm audio input gain pot brought me back up to 100% modulation with a noticeable improvement in my lows.
Not dramatic , but enough to be quite acceptable.

Thanks again for everyone's help.

Ron NU6F
Lemmon Valley Nevada

Offline Offline

Posts: 13104

« Reply #17 on: April 12, 2019, 08:45:26 PM »

100vac on primary gave 2808 vac
112vac on primary  gave 3180 vac
120vac on primary gave 3360 vac
124vac on primary gave 3460 vac (that's my local line volts

I know the thread is getting somewhat old but I am curious if by chance someone wired it for cap input rather than choke input on the HV supply? An old trick that I have played and others as well, when using 3B28 xenon capacitors to get higher voltage on a given PS. Doing this using 866 mercury vapor tubes would kill your plate transformer. It's not normally done with 3B28's either except those in the fast lane  Tongue Tongue

The secrecy of my job prevents me from knowing what I am doing.
Ron NU6F

Offline Offline

Posts: 12

« Reply #18 on: April 12, 2019, 09:56:22 PM »

Thanks for the reply..

Well, that's a thought on a possibility.

But, no, checking, it was all stock wiring  in the filter.

I ended up wiring a buck transformer onto a three prong plug that goes into that shorting Jack for the plate primary. So my primary volts is now 110v, loaded plate volts is around 2800 now.
Everything runs fine.

Oddly enough. Two weeks ago at the Ham and Hifi surplus store....All by itself was a monster boat anchor of a transformer.

I couldn't believe it...
It was a T368 plate transformer!! (non letter version with the dual 115vac primaries)

He just wanted to clear out the last few items in the spare building,,he was going to stop renting.

My brother spotted a very clean Dx100 there also.
$20 for both boat anchors!!!

I checked out the transformer...its fine and the resistance reads of the HV secondary is the 110 ohms specified in the manual!
So,I have a spare plate transformer...cheap!!

Mine measures 90 ohms either side of CT.

The T3 is racking up lots of daily contacts and plenty of DX on 80 and 40cw!


Pages: [1]   Go Up
Jump to:  

AMfone - Dedicated to Amplitude Modulation on the Amateur Radio Bands
 AMfone © 2001-2015
Powered by SMF 1.1.21 | SMF © 2015, Simple Machines
Page created in 0.094 seconds with 18 queries.