Ceramic Disc Cap Failures

[KD6VXI]

Have an amplifier with a couple tetrodes in it. 4cx250R.

B+ and screen supply are switched TX / RX.  Actually, the plate is 5 x 400 volt supplies in series.  The screen is the first supply, not very regulated, but loaded down w a 5k resistor for negative current.  Resistor checks good.

Grid is - 45 to - 100 variable.

MORE than capable of taking off, once excitation is applied.  Remove excitation, still has a carrier output.  Varying tune and load will cause output to vary and even drop down to nothing.

I'm doing a check in the grid compartment on bypass caps.  Every single one tests 15 Nf, marked .01.  Both on the power supply end of the grid choke, and all screen to ground caps.

Don't have the capability to hi pot them. My hi pot begins at 7kv.

I did hit with a heat gun, the capacitance dropped QUICKLY!.  When removed, it was at .018 uF.

This is with all of them.  10 caps.

Would the concensus be they are bad?

--Shane
KD6VXI

[KD6VXI]

Well:

It didn't fix the feedback problem... But I can now power it up without a gfci blowing......  So obviously the caps weren't passing muster.

Time to dig deeper.

--Shane
KD6VXI

[DMOD]

Have an amplifier with a couple tetrodes in it. 4cx250R.

A schematic would help the rest of us.

MORE than capable of taking off, once excitation is applied.  Remove excitation, still has a carrier output.  Varying tune and load will cause output to vary and even drop down to nothing.


--Shane
KD6VXI

Obviously you have a self oscillation situation.

If this is GG linear, short the input at the cathode via a 0.01 uF, what happens?

If this is grid input linear, short the input at the grid via a 0.01 uF, what happens?


Phil - AC0OB

[Detroit47]

If you got em out change them. Was the box stable before? What are you running for bias when it goes flaky? If you crank it up to -100 is it still squirrley?  Are you using classic Telstar design. Any parasitics in the grid leads? Send me pics if you like. I have had a lot of experience with HF tetrode amps. No need to elaborate.

John

[KD6VXI]

I did change out the capacitors....  I had ten on hand, changed out all jn the grid compartment.  It's a lot //more// stable now.

Unfortunately, not a telstar design, John.  It's a Glen.  Guy made some wonderful transformers!   Schematics are non existent.  Fortunately, it's a 30 year old amp and yes it did work.

If I crank the bias up (more negative) it does get more stable, as I would expect.

One thing I did notice is that the sockets used do not have the screen grid collet on them.  However, this was a 30 year working amp.....  So I doubt that is the issue.

I have a hundred pack of disc caps on the way.  Have to wait for them to show before I can try your idea, Phil. Afraid the puny 35 volt ones on hand wouldn't live :-).  Should I find some more in the junque bawx, I'll give it a shot.

I'm also going to look at PSupp resistors.  The caps looked beautiful, so do the resistors.  The caps where obviously breaking down, so I suspect the suppressor resistors.   Also, this has no parasitic suppressors on the grid.  Something I've always used.

--Shane
KD6VXI

[KF7WWW]

Shane

I have a 4 tube 250b Glen that is burnt also. A buddy of mine asked me to try to fix it. Problem is that the filament / control transformer is toast! And some monkey has had their hands in it. I managed to find all of th obsolete IC’s on the control board, still trying to figure out all of the voltage needed to make it work. I think it has a +-15vdc and 24vdc rail.
My guess with yours. How’s the input circuitry look? All the contacts on the relay clean? I know you know your stuff.. just trying to think of the little stuff. I’ve also heard of 250b tubes going bad in strange ways..

[KD6VXI]

Does your transformer still have the built ID on it, with the voltages listed?

If not, I can take a pic of this one.

Yes, it has a bunch of voltages.  6, 15, 70.....  But man, could he pot an xformer!  Lol.

The gentleman who built them is still at the address listed on the back of the amps, although he is about as sane as the leader of North Korea now.  He's in his late 80s.  Still has a 100 foot tower, but is QRT now.  I contemplated a drive to stockton to try to get any schematics, etc....  But gave up after a phone call.

This one is in //pristine// condition physically.  One burn mark on anything, and that's on the triac they use in the grid overload.  Still functions, however.

Is yours owned by Tom in Oregon by chance?  If so, I know who was inside it before.

--Shane
KD6VXI

[KF7WWW]

Shane

Please send a picture! I’ve been trying to figure it out for awhile. I know it’s not complicated I just wanted to keep it original as possible..

This one was bought from a SK estate in So cal. It has the remote control box. I don’t remember exactly who the guy was, but he was from everyone’s favorite band.. lol..
It’s the brown 2400 model. My buddy tried to buy the 6 tube version also but was unsuccessful. I have a feeling both where fried..

On a side note.. I found your black 4x1 socket..


Sam

[KD6VXI]

Sam,

Here's a few I snapped.  As you can see, this is really clean!

If you need voltages, etc.  Let me know.  I'll take whatever readings can help.

The control board is kind of a clusterflubbit.  It's more an overdrive control, from what I've aeen:  And not a very elegant way of accomplishing it. The other 250 based amplifiers at least used a regulated or semi regulated screen.  This one just taps the lowest portion of the hv stacked supplies.  That 400 plus volts on the screen SURE makes the gain shoot up!

This one is the 1600B model.

If your LV xformer is fried, get Gary to rewind it.  With the nameplate I included, that should be enough to get it redone.

There is a guy named Tom in Oregon that collects these.  He has //every// model made, 1 to 6 tubes, low and high drive, from what I've heard.  His bigger amplifiers are all 80 to 10.  I've also had a couple mono and ones for 20 and 40 come across the bench.

Good deal on the socket.  I'll be in touch, probably make a drive to pick the iron, etc up at the same time. 

--Shane
KD6VXI

[KD6VXI]

FYI.  That's SEVENTY volts and SIX point ZERO listed on the xformer.  Not sixty.

Also, in the middle pic, you can see the toast mark on the (I'm assuming) triac. I've not been able to cross reference the rca part number to anything.

Good thing is, they seem to still be working.....  The black mark appears to be a solder splash or something else.

--Shane
KD6VXI

[K1JJ]

Hi Shane -

A grid driven tetrode amp with 18-22 dB gain is not an easy amplifier to tame, especially on the bands 40M and higher.  I assume yours is grid driven and not GG.

I have a homebrew 160M-10M  4CX-350FJ amplifier (J= super linear version tube) and it took all of my skills to make it clean and stable.  ("Mr. Clean")

It is very important to have an electronic regulated screen and grid supply if you expect to have decent IMD. (if you are drawing screen and/or grid current)   The exception is if you run it with no grid current, (AB1) then a high value series grid resistor can reduce the grid regulation requirement.

Yes, your amp seems to have the normal parasitic instability of the average high gain amp. Is it neutralized?

Go over the layout to be sure no part of the input can see the output. This means common RF T/R relays that can have close proximity for in/out RF flow.  

You should have more than one stage of RF choke / bypass cap filtering to isolate the grid and screen supplies.

If you do all of the best engineering techniques and layout, but still have instability, then place a .001 cap in series with a non-inductive power resistor from the grid to ground right at the tube pin. You will find that somewhere between 5K and 600 ohms will probably kill the instability, but will make it harder to drive.  Use as high a resistor value as possible.  My own rig did not need it, but it took a lot of work to make it that stable in the first place. (neutralization and parasitic suppressor helped)

In the end, my 4CX-350FJ does -55 dB 3rd IMD, which is outstanding. It is part of my class A linear chain driving the 4CX-350FJ  input at 1/10th watt and putting out about 10 watts.  This kind of care is required to see -55 dB 3rd IMD, otherwise -30 dB 3rd is about all a hard-driven, grid driven amp will see.

There's more, but this will get you started as you plow thru this project.

T

[KD6VXI]

Well.  I got some time, took the top cover off.  All LOOKED OK.

Decided to check out R Supp.  Ha ha ha.

Turns out the connections from the strap around the anode to the actual parasitic suppressor where loose enough to almost spin freely! These looked like they where tight.  However, turned out in close inspection that they where only making contact on about a sixteenth of the metal to metal contact.  Tried to take one parasitic suppressor apart, the common end of the 2 suppressors came apart!

Now, would an amp with bad connections to the parasitic suppressors be capable of taking off?  :-)

I'd say, that could cause a problem!  On to check bypass caps on the top side of the rf deck.

--Shane
KD6VXI

[KF7WWW]

Shane

Thanks for the pics! That is a great help.
Man everyone of these amps are different.. Seems like the design of these where changed more then some guys underwear!

[KD6VXI]

SUCCESS!

Replaced EVERY disc capacitor in the rf compartment.  Replaced all the disc caps, matter of fact.

Then, found the base of the plate  hoke was duct taped together. Was able to sour e a slightly longer screw and a ceramic spacer.  Used a diamond grinding bit to get the broken end of the choke flat, and then was able to remount it.

That was found while inspecting PSupp.  All 4 of the suppressor resistors showed 20 pct or higher resistance.  In addition to one not being soldered (cold solder joint) at the choke end, and all 4 not being completely tight to the tube.

Replaced all the parasitic resistors, swept them on my vna (they start becoming reactive about 36 mhz.  Below 29.5, flatness!) and installed everything tightly.

1500 watts carrier continuously, then ran a 'string of dits' into it.  It would take off when unkeyed before.

NOTHING!   Nothing but clean rf output and a clean unkeyed amp when excitation removed.

Thanks for the help everyone.   And 18 db gain, give or take, is what I finally ended up with, completely stable (even removed the dummy load and keyed into no load a couple times momentarily)

Now to build a 13 db attenuator!  Or, maybe Stu's amp attenuator.

--Shane
KD6VXI



Tom, I still want to see a puresignal enabled TX through your amplifier chain.....  You might be the first - 60 db imd guy out there!

[DMOD]

Glad it's working without any internal oscillations.

Just curious:

This one is the 1600B model...

of what?

I still didn't get the manufacturer or whether it is grid driven or cathode driven.

Phil - AC0OB

[KD6VXI]

Sorry, Phil.

It's a 'Glen' 1600.  Uses 4 X 4CX250 (this one has R).  Grid driven, grounded cathode.

This one is 15 meters.  Both marked on the front as well as that's where the output tunes with the capacitors anywhere near mid point.

Glen was a transformer manufacturer here in California.  He used the transformer business to prop up a (mostly) Cb amplifier business.....  But his amps are pretty advanced, have full overload protection, scr based transformer control for TX / rx switching.

They also have some REALLY odd designs of their control board (card edge connector a la Alpha, etc) which uses mostly unobtanium ic and transistors I think he bought on the surplus market...  Gold abounds.

He made them build to order, for whatever band you wanted.  I've seen an 80 meter one similar to this one, legal limit w 100 in.

--Shane
KD6VXI