Yet *Another* Crapout

[k4kyv]

I wouldn't recommend a fixed regulated bias supply for the 810's, because the resting plate current will change excessively with small changes in a.c. power line voltage.

Better to substitute a heavier duty bias transformer.  An excellent choice would be a 150 v-a "control transformer".  Those have split primary and secondary windings with taps, much like a "multi-match" modulation transformer, so it is easy to set them up for a secondary at 220 volts each side of midtap, good for better than a half amp. Use a full wave rectifier, or a bridge circuit per your preference, several hundred mfd of filter capacitance, and bleed down the power supply to near the full current capacity of the transformer.  Take the bias off a tap on the bleeder. There should be no more than a few hundred ohms from the bias tap to ground.

That will substantially reduce the internal resistance of the bias supply, but when the line voltage decreases, so will the bias voltage, which will tend to stabilise the resting plate current.

Or else design an electronically regulated bias supply so that the output voltage is controlled proportionally by the a.c. line voltage input.

[K1MVP]

But if you got a "Cadillac" thats in the shop all the time,--would it not be better to have a Chevy or
Ford you can depend on.

This from the guy who builds nice stuff but is never on the air. I think Ron needs to come over there and light a fire under you.

It's more like a Rolls Royce with respect to price, Rene. Quality is pretty good, too. I hear plenty of folks 'fixing' their Rangers, Valiants, DX-100s and so on. Keep in mind that I use it regularly and I don't mollycoddle it. It's an old transmitter that wasn't used much for decades, and being impatient to get it back on the air, I didn't go through it completely. No fault of the transmitter, it's higher quality than any other piece of amateur gear I've seen. Blame it on 'operator error'. The ratio of downtime to on air time I've gotten out of it since March 21st '06 is actually pretty good, when you consider how often I'm on. Besides, all the Chevys and Fords need work too.

Todd,
Hey, just cause I am not on phone these days does not mean I have not been on (cw that is).
Right now I am finishing up some unfinished qrp projects,--like my 40 meter cw tranceiver, after
building another regen.
I have just as much "fun" building if not more, than getting on the air,--besides I need a "sabatical"
for a few months.--Man you are one to talk,--you were off for over 10 years, or was it 12?

                                          73, OM, Rene, MVP

P.S,--and after I finish my qrp rigs,--guess what, I got a bunch of parts, and cabinet to start my "big linear" for 160,--gee I might be off the air almost as long as you were,--(ye gads).
 
                   

[Todd, KA1KAQ]

P.S,--and after I finish my qrp rigs,--guess what, I got a bunch of parts, and cabinet to start my "big linear" for 160,--gee I might be off the air almost as long as you were,--(ye gads).                 

QRP? Life is too short for QRP. At least from an AM perspective, on 75-80m, at night. 

Being off the air is like punishing yourself. I don't recommend that for anyone who enjoys operating. I actually brought my logbook to work today to do some QSLing which makes it easy to check back. Looks like my last AM contact before the station went dark was 3/10 of (I think) 1995 with Jerry, K1GUP. Worked W2VJZ the day before. I got back on 3/12/06 (not the 21st), so that makes it eleven years almost to the day. Too long!

Just remember, those rigs are only radios if you use them. Otherwise they're just interesting doorstops. Yes, I've used some for that.

Quick update: Didn't have a lot of time last night, but I checked the bias on the 6B4G drivers and the resting current on the 810s. -60V on the drivers, 50ma on the 810s, right where it should be. I did not check anything under transmit load because of the time needed to do it properly and safely. Maybe tonight. But the drivers both got quite hot just sitting there with no load, which must be normal for 6B4s, 6V6s, etc. The line voltage meter was definitely wandering the evening of the crap out, from around 117 to 122. Usually it hangs right on 120.

Bruce: The mod transformer does indeed have spark gaps on it, set where they are supposed to be set. This was checked last April when W1UJR was visiting. Joe's transmitter was a mess when he got it from Columbia. Parts missing, holes gouged into it, broken glass, blown iron.....as a result, his is the only complete and accurate restoration I'm aware of. Been trying to catch up with him this week to get Deerfield plans set, no luck so far. He's gonna smack me for breaking the transmitter again.

John: check for a pair of what in the audio circuits? 

[WA1GFZ]

Well Todd,
Resting current right sounds like the iron has a shorted turn. Me being a bit of a nut in your shoes. I would find what size wire is inside the transformer then find the thickness of the insulation paper. I would rewind it with kapton insulation and see if I could increase the wire size due to thinner insulation. This would keep the inductance the same but reduce the resistance. Kapton is good for 7000 volts per mil so 2 or 3 mils thick would do it. Paper insulation could be 10 mills thick.
I would also buy a spool of the best wire I could get.
Not being crazy you could send it to Texas Pete D. Joe PJP will not allow you to use different iron in a KW1. (or be tried for treason )

[Todd, KA1KAQ]

Gary WZ1M rewinds and is closer and cheaper than Dahl, so I'll have him rewind it for me. No idea what he uses for materials, but folks are happy with his work and prices. I'll include your notes when I email him.

Joe won't beat me up too badly as long as I don't blast holes in the rig. I'm of pretty much the same mindset (use it as a 1950s transmitter, don't try to make it a 2007 model), but don't mind doing the mild mods that are reversible if they won't let me be buried in it and someone else ends up getting it. At the current rate, it may deserve to be retired and admired after another few decades. Meantime, I fully intend to run the crap out of it. Looks like I'm succeeding?

[W2XR]

Frank,

I think Joe did send the bad mod xfmr from his KW-1 to Peter Dahl for a rewind. But if I recall correctly, although the rewind worked OK, the xfmr talked so much it was unusable. I think he had to send it back to Dahl one or two more times before they assembled that thing tight enough so that it stopped singing.

73,

Bruce

[WA1GFZ]

Yes, I seem to remember the same story. I designed transformers years ago and it is all about potting. A mod transformer I would pot with a vacuum pulled on it so all the gaps get filled. I would also make sure the last center E was well clamped. Today wire insulation is much better and insulation is thinner. I've used wire you can melt the coating off with a soldering iron. It took a solder pot set pretty high. 

[AB2EZ]

Todd et. al.
I was thinking about this last night, while cooling my heels in a bed and breakfast here in South Jersey... like trying to solve a mystery.

It seems to me that if there is a problem in the modulation transformer, the observed symptoms (fuse doesn't blow, resting current ok, higher than normal modulator plate current under modulation, very low audio) would not be caused by a moderate number of shorted turns in the modulation transformer. A moderate number of shorted turns in the primary would result in higher than normal modulator plate current under modulation, probably some imbalance of the class AB modulator... but not such a dramatic drop in the modulation at the output of the KW-1.

Check the following (with appropriate safety precautions)

1. Is there a short between the primary and the secondary of the modulation transformer?

2. Is the blocking capacitor between the plates of the 4-250's and the tank circuit shorted out?  [I don't have a schematic with me, but depending upon the details of how the tank circuit is designed, a shorted blocking capacitor could lead to a very low audio frequency load between the plates of the 4-250's and ground, without producing a DC short to ground.]

Separately, since the secondary of the driver transformer normally looks into the very high impedance associated with the grids of the 810's, you could remove the 810's, leave the plate supply off, and measure the output of the driver transformer with audio supplied... to see if it looks reasonable.
 

[AB2EZ]

When I got to office I use here in South Jersey, I downloaded the KW-1 schematic from the CCA web site:

Updates on my last post:

I looks like a short in the 4-250 (x2) plate blocking capacitor would also cause a DC short to ground, via L117, if L117 is still in place and not open.

There are four (4) ways to place a short across the modulation transformer secondary:

1. The rear shorting switch S502
2. Either of the 6000 uF capacitors that are part of the high-level splatter filter (if they haven't been removed... which would normally be done if the splatter choke has been bypassed)
3. The AM/CW relay. NOTE: this relay is normally in the CW position, and is moved to the AM position (removing the short across the secondary of the modulation transfromer) when current flows through its coil. If the relay contacts were sticking, or it the relay coil was open, or if the AM/CW switch contacts weren't working properly, then the short across the secondary of the modulation transformer would remain in place.

Best regards
Stu

[WA1GFZ]

Stu,
A shorted turn in the mod transformer would not effect the final except to reduce the modulation swing. It would reflect a low Z back to the modulator tubes if the short was on either side. A shorted turn would suck up the AC swing reducing modulation. Primary to secondary short would cause some weird distortion I would think.
This would be the least common fault because usually there are a numbers of layers of insulating material between windings.
Anything is possible at that age though.
Might be time to take it out and connect it to a variac and series bulb as Steve suggested.

[Todd, KA1KAQ]

FB Stu -

I actually had a bad blocking cap years ago, it would blow out one of the 15a screw-in fuses anytime I tried to apply HV. Once it was replaced, I could get on to the other issues, like swapping out the 872s for 4B32s.

The AM/CW relay also lights the 810s, and only switches in upon power up, unless you change to CW and back. If time permits, I plan to give it a thorough cleaning this weekend anyhow, while I'm in there. But it has no problem latching and holding, so any issue would have to be related to a specific contact.

Since recapping and repairing the audio deck, I've had to keep the audio gain down even lower, so I'm still a bit surprised at the likelihood of a shorted transformer. It's what happens when you run break-in up in the ghetto.

[Todd, KA1KAQ]

Primary to secondary short would cause some weird distortion I would think.
This would be the least common fault because usually there are a numbers of layers of insulating material between windings.
Anything is possible at that age though.
Might be time to take it out and connect it to a variac and series bulb as Steve suggested.

I'm going to do some testing this weekend if I get all the house chores done as planned. Several stations did report that my audio was weak and distorted when I tested two night ago. Al VE3AJM compared it to the sound from close in stations when the band starts to go long, that warbly, underwater sound. The meter jumps around a lot more than normal too, as if the gain is cranked waaaay up. If the transformer was shorted to ground, I'd expect no audio and fuses to blow.

[W9GT]

Hi Todd, 

Very sorry to hear about the rig crash!  I know how it feels.  :-(
When the  mod transformer in my 30K-4 failed......the audio level went down and became very distorted.  The mod current, on average, appeared to be a little high, but the mod idle current was near normal.

I tested the transformer by comparing the winding resistance measurements to what I had recorded when I originally got the rig and had been doing some troubleshooting.  I measured the primary resistance and found that one side of the winding had a reduced resistance reading (went down about 20 ohms or so).  I then tested the transformer with a small AC (filament) voltage on the secondary and looked at the voltage across the primary.  I found quite a voltage imbalance in the readings between one side of the winding and the CT and the other side and CT.  Probably a 30% difference.  This led me to believe that one side of the primary had shorted turns...probably between layers of the winding.  I believe this is a common failure mode in those old mod xfmrs and it is hard to be sure of, without the transformer actually being tested under working conditions (with HV applied).

A megger is also useful in looking for insulation breakdown between the windings or from the windings to the core.  This does allow you to look at it with a higher potential applied, but is not very useful for looking at the winding itself.  Using an RCL meter to look at the inductance of the windings, might also be a feasible approach, but would probably require some before and after failure readings to be certain of what you are really looking at.

I highly recommend Peter Dahl for replacement mod iron, but as you know, it is a rather expensive solution.  My 30K-4 now sports a new Peter Dahl "Heavy Duty" mod iron and it works quite nicely. 

73,  Jack, W9GT

[WQ9E]

Todd,

Another approach is to hook an audio signal generator to one side and a scope to the other.  If you have shorted turns, the AF generator is going to have difficulty building up voltage across the transformer and of course you can hook the generator to the secondary side and see if you have equal amplitude on both halves of the primary.  I made a copy of a great little article on this process, of course now I cannot find it.  I should digitize everything and put on the computer so it is harder to lose one thing (but easier to lose everything at once if you don't have good backups!).

Rodger WQ9E

[AB1GX]

Double Crap - I just nuked a pair of 6LQ6 output tubes or I think I did.  I know less than zippo about tubes.  They're not drawing any cathode current, but the tubes are lit up with 900 Volts on the caps, 300V on the what-ever-it's-called grid and about -60V on the control grid.  Even with 0V bias on the control grid I get zippo current. They're toast, right?  Or, how can I be sure?

How does a tube fail? I just looks like a bunch of glass and iron to me.

[Bacon, WA3WDR]

I would expect some current with 900V plate and 300V screen and 0V control grid, even from very badly abused tubes - but zero current says something else failed, maybe the metering shunt and the meter, or the line to the meter, etc.  Check for continuity from the cathodes to ground, then try removing the tubes and see if you can get some current reading by placing a power resistor of maybe 10K from the plate circuit to the cathode circuit.  Then make sure the voltages are applied to the right pins of the tubes, etc. 

[Todd, KA1KAQ]

Rodger, I'd sure like to see that article if you come across it again. Unfortunately a lot of my test equipment is already in storage (some never came out from the last move), so I'm limited in what I can do.

Alas, no time this weekend to do any radio work beyond getting the HRO-5C assembled in its rack and out of the way.

Did exchange email with Gary, WZ1M. He is standing by to receive at least one zorched transformer, maybe two if the spare in the Viking KW is also bad.

More to come, as time permits. Many thanks for the great suggestions and commiseration. 

[WQ9E]

Hi Todd,

I haven't found the article yet but here is a link to a website that provides a simple test method using a scope:  http://www.repairfaq.org/REPAIR/F_flytest.html#FLYTEST_016  Although the article in general is structured towards flyback testing the "ring" test will work great for AF and power transformers.  If your scope has a "gate out"  you can probably use this for the pulse source.

Second thought, if you have "zorched" the transformers in the KW-1 AND the Desk KW maybe you need a priest and some holy water in addition to a transformer rewinder.  Or do you by chance have some CB genes buried deep somewhere in your family tree???

I spent the weekend turning $600 worth of wood into a swing and slide set for my daughter instead of finishing my Viking 500 restoration.   I hope next weekend I don't turn several hundred dollars worth of Viking 500 into smoke!

73, Rodger WQ9E

[WA1GFZ]

Be careful runing a flyback test. You and or the signal source can get belted.