3-500 amplifier is working

[ka1tdq]

Well, the plate choke didn't explode this time and the 3-500 amplifier is working.  I'm driving it with 23 watts from my solid state transmitter and getting 360 watts output.  I tried using my tube transmitter which puts out 12 watts but was getting very little output (around 20 watts from the amp) and the tubes were getting red in the face.  That's bad because my audio goes to the tube amp and not the solid state driver.  My goal is to modulate this carrier...

It's probably my input circuit tank. No matter what I did with the input variable cap, output doesn't change.  I removed the swamping series 100 ohm resistor/100 pf cap to see if it mattered and it didn't.

So, we're going out for the evening and I'll sleep on this one.  I'm hoping that my Alinco receiver will ship on Monday so that I can get on the air soon.  I just need to figure out the low drive-very low amp output problem. 

Jon
KA1TDQ

[K1JJ]

Jon,

Glad to hear it's working!  I see you added some protective Plexiglas for HV safety.


To test the input tuning, put a scope probe on the fil pin and inject a signal. You should see the signal increase on the scope as you tune the input through resonance.  Since the Q of this input circuit is maybe 3, don't expect a dramatic peak, just enough to know it's resonant.  If there is no peak, then play around with the L and C values until you see a peak.  When the amp is running normally, the lowest input swr may coincide with a peak too.  Anything less than 2:1 is considered OK.

Strange that the solid state rig drives it well but the tube rig does not and causes the finals to blush.  

** Be sure the tube rig is tuned to 3.8 MHz and not some harmonic. Also be sure the carrier looks clean on the scope.  It does sound like either tube exciter tuning error or maybe an exciter parasitic possibly.

BTW, you said you plan to modulate the 3-500Zs with audio - how do you intend to do this?

T

[KB2WIG]

J,

FB OM.

KLC

[WQ9E]

I think you need to take a close look at your tube exciter to see what it is putting out where.  This is a case where a spectrum analyzer is really nice and I guess one of the "spectrum scope" receivers might be of some use.  I still suspect your tube transmitter was putting out spurious RF at the series resonant point of your first plate choke and I would tame the transmitter before it creates more amplifier smoke.

[ka1tdq]

Ah, yes... it isn't possible to modulate a GG amplifier.  I was just thinking too that it is tuning error on the tube transmitter. The loading is fully unmeshed. I used this same transmitter with my last amplifier with no problems.

In a few days I will do some real tweaking with an oscope to make sure everything is kosher before I go live with this thing. I need to finish the plexiglass enclosure too and put the swamping circuit back in the input section.

Thanks for everyone's help in taming this thing. Late at night sometime I will try to holla at my east coast nigs.

Jon
KA1TDQ

[AB2EZ]

Jon

Based on the symptoms, my guess is that the following is happening:

1. The biasing circuit is damaged/destroyed (possibly as a result of the exploding choke episode)... and it is presenting an open circuit from cathode to B- . Therefore, although the filaments are lit (5VAC from one side of each filament to the other side of each filament), there is no path through the biasing circuit for average cathode current to flow. NOTE: a failure of the biasing circuit (short circuit, followed by burnout/open circuit) could have caused the earlier exploding choke episode.

2. The solid state driver provides a DC path to ground at its output (via the output transformer). The vacuum tube driver does not provide a DC path to ground at its output.

3. DANGEROUS: there is no DC blocking capacitor between the RF input of the amplifier and the cathodes of the amplifier tubes. For safety, the RF input side of this blocking capacitor (facing the driver) should include an RF safety choke to ground (providing a path for DC, but blocking RF)

A) When using the solid state driver (with its DC path to ground) in conjunction with the 100 ohm series "swamping" resistor... you are providing a 100 ohm DC path from the 3-500Z cathodes to ground. This results in a "cathode resistor biasing" configuration (acting in place of the damaged intended biasing configuration) that produces enough negative bias on the amplifier to enable it to work... although not the correct bias for proper operation as a linear amplifier.

B) When using the vacuum tube driver, there is no DC path from cathode to ground, the amplifier tubes are effectively biased in cutoff,  and therefore the amplifier tubes as not producing any cathode current. In a grounded grid configuration, the output power of the amplifier includes the input power... and that corresponds to the small output power that you are measuring at the output of the amplifier when using the vacuum tube driver.

As a separate safety issue: I hope that you have a safety choke connected between the RF output of the amplifier (the output side of the tank circuit where the output connector is) and ground... to provide a DC path to ground. Even a tiny amount of leakage conductance around one or both of the plate blocking capacitors will put the full B+ on the center conductor of the output connector (and one side of the antenna)

Repeating what I and others have posted previously... you are taking big safety risks (possibly lethal) by working with this amplifier.

Stu

[W1RKW]

any chance that the RF choke on the tube exciter is open?

[K1JJ]

The 100 ohm "swamping" resistor, attached to one side of the cathode/filament, is in series with a 100 pf cap to ground. Thus, not effecting DC bias.  It is an RF swamp for potential cathode choke VHF hi-Q resonances.

T

[AB2EZ]

Tom

If the swamping resistor, the series capacitor, and the cathode drive circuit are all configured correctly, then I agree... but the symptoms suggest that the configuration is not correct.

It would be informative to see a complete schematic.

It is also possible that the full B+ will appear (perhaps briefly) across the 100pF capacitor (shorting it out) if the biasing device (Zener and pass transistor) has opened up.

Stu

[K1JJ]

Stu,

Yep, a look at the complete schematic may show some things we have assumed were correct.

Jon, can you draw one as presently wired and post it?


BTW, interesting idea on placing a safety RF choke from the driver side to ground at the linear amp input.  At least the DC component will be quenched during an arc over.  I still worry about the RF component that sometimes destroys drivers. I use vacuum antenna in/out relays these days.


T

[ka1tdq]

I'm switching gears on the transmitter side.  Rather than use my single tube transmitter, I'm going with my original plan and use my homebrew solid state transmitter with the FET intermediate linear going into the 3-500 amp.  

Steve designed the audio portion of the transmitter and I've set the carrier on the collector at 15 vdc from a 50 vdc supply.  This will give around 200% positive peak capability.  The FET amplifier boosts power to 13 watts going into the 3-500 amplifier.  I've measured power out of the 3-500 amp going into a dummy load and I get 160 watts carrier.  Not bad...

I have another problem now though with my antenna.  It is a flat top dipole fed with 450 ohm ladder line and is matched via a homebrew link coupled tuner.  When I connect the amp to it instead of the dummy load, carrier power out drops from 160 watts to 60 watts.  I'm guessing that there's some reactive element in there.

I was thinking about possibly just getting a commercial 4:1 balun and feeding the ladder line directly with that.  It wouldn't be a perfect SWR match, but the amplifier could handle that fine.

What do you think?

Also, I don't have a schematic yet of the 3-500 amp, but I'm working on it.

Jon
KA1TDQ

[KB2WIG]

J,

Stay with the link. You'l have more lattitude to do stuff. Not to start a ballun war, but the (ballanced) link tuna should work out better than winding ferrites.

Check this out,       http://vk1od.net/tx/concept/lctr.htm

and here is da berries.     http://www.k1kbw.com/newrig/link_coupled_tuner.jpg

Watt frequency is the dipole cut for? If your feeding a 50 ohm ant with 450 ohm feedline yer vir wars are.....  
Are you using the CBS* window line? There are a few posts here on da 'phone for using lectric fence insulators and THHN to home brew ladder line.

I've used plexie glass for panels, but they are screwed into the rack and won't fall off withought a fight.

You ar doing FB with your projects; a lot faster then I AM. I've been sitting on my 813 rig fer 5 years.... and I have all the parts I need it.

That is a nice picture you posted. The radio stuff is nice also.... ..

klc



* Crappy brown stuff

[N2DTS]

What a mess.
I really dislike tuning up all sorts of things, many of which can interact.

You have an antenna tuner, that needs RF to tune up, you have the output network of the amp, the input network of the amp, the output network of the driver rig if you use tubes, drive levels, lots of stuff to tune!

I hope you measure how clean the output is, and find the operating points where it is clean.

All the above drove me nutz when I built a homebrew GG amp.
I was not using an antenna tuner, but the drive levels and tuning and loading stuff, plus bias and plate voltage settings seemed endless.
All that for little power out.
I gave up and scrapped the amp.
And I never bothered to check how clean it was, it was hard enough to tune it up as it was without thinking about the cleanest operating points.

My anti amplifier bias is showing I suppose, but I find it easy to adjust a class C amp for low crud out, then all I have to do is dip and load.
That and the fact that a single 813 will give about 2x the power out that you are getting, at much lower voltages.

[n1uvi]

I wonder if you are seeing some coupling due to the lack of shielding

I know it looks pretty but you'll be blind in a few years without shielding on those triodes

[K1JJ]

What a mess.
I really dislike tuning up all sorts of things, many of which can interact.

heheheh....  I feel your pain.

Unfortunately, it's the only thang we can do for ssb, tube linear operation.  But the plate modulated ease in class C certainly is an advantage.


As for linear, don't feel bad... if I want optimum performance, my class A / AB1 system that gives -50 to -55 dB 3rd order is like tuning up a maze.   The 100 mW FT-1000D low level drives a 1 watt lab amp > to a class 3CX-350FJ with input and output tuning > to a class A 8877 with input and output tuning >  to a class AB1 pair of 8877's with input and output tuning.  And they all have to be perfect to see =55 dB 3rd... or if simply peaked, it drops to -45dB 3rd.

I cringe when it's time to switch bands.  

In contrast, as you say, the 4-1000A plate modulated rig is a breeze and really puts out the soup at lower voltage and less fussy tuning.


* Jon: Could the close proximity of your antenna (RF field) be affecting your power readings?  Try moving the meter around and add some clamp-on ferrite cores to the various leads going in and out to see if thangs change.

T

[ka1tdq]

Attached is a picture of da tuna that I built.  I made a few mods since of not shorting out the ends of the coil and adding another 100pf doorknob on the outpoot.  I guess I just need to play with the setting to match dem dare reactances, ayuh (I'm a getto Maine-ah).  Laahd only knows, I's itchin to get back on da ayuh, ayuh.

Jon
...like it is...

[N2DTS]

Copper tubing works well....

What if the wood catches fire?

[AB2EZ]

This QRO Technologies amplifier (current production, and highly rated on eham.net) uses a pair of 3-500Z tubes. The site provides the complete set of schematics, parts lists, etc.

http://www.qrotec.com/

Of note: the biasing circuitry in this amplifier consists of twenty (20) forward biased 1N5408G diodes in series, plus an additional transistor in series configured with its collector connected to its base (functions as a diode.. but probably acting as some kind of temperature compensating element). There is also a 3A, fast acting fuse in series... but I don't know how much protection it provides for the components (e.g. the plate choke) in the event of a short circuit or other excessive current situation.

The RF output circuit includes a DC path to ground (i.e. the output transformer also acts as a safety choke).


The input circuit does not appear to include a safety choke (DC path to ground) on the external RF input side of the cathode DC blocking capacitors that connect the output of the tuned input board to the cathodes of the 3-500Z's.

The power supply appears to be a voltage doubler design... and the manual states that the B+ is approximately 3900 - 4300 volts (surprisingly high, vs. what I would have expected). I suspect this is a typo. I suspect that under load, this amplifier would produce a B+ of approximately 3000V  [i.e. approximately 1200VAC rms x 2 x 1.25 DC volt per rms AC volt, under load] There does not appear to be a fixed 10 ohm or 20 ohm series resistor in series with the B- lead (to limit the plate/cathode current in the event of an arc, etc.)... although many amplifier designs include one.

I am not suggesting that Jon duplicate this amplifier's design (it is still a very dangerous endeavor to build and/or operate this kind of amplifier as a home brew project)... but it is an interesting point of reference.

Stu

[n1uvi]

Thats not a tuna !!

this is a Tuna !!