Power Supply Dilemma

[AB2EZ]

Todd

I checked the voltages on my KW-1 a few months ago... because I was concerned that the screen voltage (nominally 500 volts) was too high. I checked these on the terminal strip on the back of the unit.

What I determined is that my "300 volt" supply and my "500 volt" supply were both reading about 120/110 times their nominal values... which was no surprise. In addition, while my voltmeter showed the screen voltage at around 560 volts, the meter on the KW-1 shows it at 620 volts. I decided that the meter shunt for the screen voltage is in need of replacement. [Right! ... the next time I disassemble the entire KW-1, I will replace that resistor  ]

As you pointed out (above)... the -150 volt bias supply activates the "control circuitry" (relays)... and, according to the manual, the relatively slow turn-on of the bias supply is intended to prevent application of the high voltage to the high voltage rectifier tubes before they have a chance to reach operating temperature. This is an issue to be considered if you were to "solid state" the bias supply.

I agree that it is puzzling that this problem has just come up now. I think that... even with the choke input design... it is possible that the new capacitors are drawing more surge current (at turn-on, the current builds up exponentially in a choke input supply, until the capacitors have charged up to a reasonable portion of the average rectified voltage)... and that, combined with the current drawn by the 6B4G's (before the bias supply turns on) may have pushed you over the edge with respect to blowing the 3A, fast blow fuse. Why now? Maybe your line voltage is a little higher than it was in the winter. Maybe the filament voltage on the 300 volt supply's rectifier tubes is a little higher (less equivalent series resistance?). Perhaps one of the capacitors in the new dual section can was not making good contact (cold solder joint?)... and thus, perhaps, you now have more capacitance to charge than when you first installed it in February '07.

Try unplugging the 6B4G's. If the fuse doesn't blow when you turn the KW-1 on, then you know that you have identified the problem.

If the surge through the 3A fuse at turn-on is too high, and if you don't want to solid state the bias supply (e.g., for the reasons given above) then you may want to consider inserting a small 12.6 volt "bucking" transformer in series with the primary of the 300 volt power supply's transformer. This will reduce the line voltage by about 10%, it will reduce the surge current at startup, and it will also lower your "300 volt" supply to a value closer to nominal.

Stu

[WA1GFZ]

Just a thought but could one of the bias pots be damaged from the tube failure? Maybe you need to tack a wire on a number of places so you can monitor voltages in normal and failure modes. Also I like to go in and rock the pots back and forth to clean any crud off the wiper that could open the connection. Sitting in one spot for years can be a problem. We just blew up a system in the lab when the lab supply pot that had not been adjusted in years threw the voltage from 28 to 55 volts. All it needed was a spin to zero and full scal and a readjust back to 28. Pots are bad news long term if you set them and forget them.

[AB2EZ]

Well, I now see how the center tap of the secondary of the transformer in the 300 volt supply is connected to ground.

The filament of the 12AX7 microphone amplifier and the filament of the 12AU7 phase inverter each have a center tap at pin 9. Although not shown on the schematic, pin 9, in both cases, is connected to ground. Thus, both of these tubes work off of the 6.3 volt filament winding that also supplies several other tubes (including the 6B4Gs). This is a good way to reduce the hum in the low level audio stages...

However, if you were to remove the 12AX7 and the 12AU7 from their sockets... everything that works off of 300 volts (except the 6B4Gs, which have no separate cathode) would stop working... including most of the RF stages.

Just one of the interesting design facets of a transmitter that cost more than three x the price of a family car back in 1952.

Stu

[N3DRB The Derb]

the "new" dual section can might be a shell that has 'modern caps' installed inside it. To my knowledge there is only one source of "real Mallory" twistlok/FP can caps, advertised as being made on the original Mallory manufacturing machinery. As such, they are damned expensive, but in a KW- 1, what the hell?

http://www.tubesandmore.com/  go to capacitors and FP section.

[WD8BIL]

Still might be something as simple as you replacing the sloblo fuses with faster acting ones,Todd !

Did ya go out and get some sloblo 3 amp jobbers ?

[Todd, KA1KAQ]

Hoping to pick some up tonight Bud, and that's exactly what I'll try first. It'll buy me some time until I can dig into it later. Gotta get dad's '50 Lincoln out of storage tonight, planning to stop at RatShack on the way by if time permits. They still had glass fuses last time I was in there.

Derb, those are indeed the caps I used: the Twist-Lok/FP types from AES, along with Sprague Atoms for the axial electrolytics and 600v SBE O.D.s for the wax a paper coupling caps. Always better to spend the extra coin and get the good parts to start with. It's not like I buy a lot of them, or often.

Chatted with Joe 'PJP on the LL last night about this, he things there may be a fault up in the RF section since the 300v ps tranny also has its mitts in there as Stu pointed out. Little work has been done in that section beyond replacing a shorted 4-250, new 807W, and replacing a bad plate blocking doorknob. Didn't get a chance to dig into it last night as I had a load of furniture to move into the storage unit (which meant moving a bunch of stuff out, then back in). Got home late, so maybe tonight.

Initial reports on the rewound mod iron were good, at least.

[AB2EZ]

Todd

From the schematic... the 300 volt supply is not applied to the RF stages unless:

The send-standby-calibrate switch is in either a) calibrate position or b) send position.

If the 300 volt supply were applied to the RF stages... you should hear the VFO in your receiver. You would also see the 300 volt power supply on the multi-function meter. [The meter senses the 300 volts from the same bus as the RF deck]

So... I doubt that a problem in the RF deck is causing the 3A fuse to blow.

Separately, from the cabling diagram... the 300 volt supply appears on terminal 13 of the terminal strip on the back of the modulator deck. From that point, it goes to the send-standby-calibrate switch via a wire in the wiring harness on the back of the transmitter.

Thus

1. You can measure the 300 volt supply with your multimeter, from terminal 13 of the terminal strip to ground. [Terminal 12 is the -150 volt bias supply, terminal 14 is the 500 volt supply, and terminal 16 is ground.]

2. You can disconnect the 300 volt line going to the send-standby-calibrate switch by loosening the screw on terminal 13, and temporarily removing the wire from the wiring harness that is connected to terminal 13. That way, you could double check to see if the send-standby-calibrate switch is properly disconnecting the 300 volt supply from the rf deck when it is in the standby position. With the wire disconnected from terminal 13, and the power completely off... and assuming the send-standby-calibrate switch is functioning properly, and in the standby position... the resistance from the wire to ground should be infinity. If the switch is not functioning properly, then you should see about 112k ohms from the wire to ground (from a voltage divider on pin 6 of the 6AK6 buffer tube, that is connected from the 300 volt line to ground on the rf deck). This (112k ohms to ground) is also what you would expect to see if the switch is in the calibrate position.

Stu

[N3DRB The Derb]

Todd,

the Gonset 201's use those things at 34.95 a pop. I just couldn't afford new ones, needed 5 of them. 175 bucks to recap it. Just couldn't afford it.

I never use NOS lytics. They gotta be fresh.

Wound up dead bugging it under the chassis, not my best work, but I didn't even have any terminal strips. These next hamfests I really have to load up on repair supplies. Term-strips, caps, resistors, etc.

[Todd, KA1KAQ]

Well, it's not a 'fix' in the true sense of the word, but re-installing a 3A Slo Blo fuse did the trick. For the time being at least, it's good to go. I still want to track down the offending part(s) for my own peace of mind.

Stu, thank you for your tips on measuring the voltages and where else to look. I glanced at the schematic and it appeared that the 300v was fed to the RF section through a resistor near the audio driver circuit. I probably didn't look closely enough or follow it far enough. The other issue of course, was the additional tube filament winding on the same transformer, but that seems to have checked out o.k.

Derb, whatever it takes to get that Gonset screaming, do it. You can dead-bug it for now and upgrade the caps later with the chassis mounts when funding allows and toss the others in the junkbox for spares. Or just run it. 

And Buddly - yes! In a strange sorta way the old saying applies 'if it ain't broke, don't fix it'. Or maybe in this case 'if it was kinda broke and kinda fixed, don't go looking for more trouble until you have the time to wrassle with it'. That Slo Blo was obviously in there for a reason, put there years ago by someone who knew the rig better than me!

BTW, Bud - does it sound a bit...familiar? Getting a rig and using it without ever checking the fuses?  It's good to know someone who can appreciate such things.

[WA1GFZ]

My V2 CDC went through two sets of hands before the welding rod size fuse was discovered.

[W4PQW]

I would try a SLOBLO 3 amp fuse..That will probably take care of the problem..If you are already using a slo fuse,Maybe a soft start relay would be in order..If I can help ,be glad to..  73  W4PQW

[Ralph W3GL]

Hey Todd, what did I tell/suggest to  you back a week or so before you started this thread? Remember, when you checked into the "whats for dinner" net the last time...

PUT A 3 AMP SLOBLOW IN THERE...

Of course, look for the real problem, if there actually is one (other than old age on the KW-1...).