Replacement for 304TH

[W1EUJ]

I have a linear amp stashed away for when I get time again for radio. It runs a single 304TH in grounded grid, on 2000 V. Perhaps I look in the wrong places, but I'm not seeing alot of NOS 304THs. Is there another triode that will give me about the same power gain, and run on the same plate voltage?

Thanks

David Goncalves

[WA1GFZ]

304 is a crappy linear go with a 3-500z or one of the Russian tubes and really strap

[W1EUJ]

Additional limitation is 2kV, 500mA transformer.

[WA1GFZ]

The 304 makes a nice audio tube but its long internal leads make for trouble at RF. 4CX1500 would be another good tube if you can find one.

[Opcom]

beware 304THs on ebay. 4 out of 5 I bought were no good when 3KV was applied. The flashover was magnificent.

Mine were modulators, so I put 3-500Z's in and turned the bias way down.

[W1EUJ]

Those 3-500Zs look close - how's the availability of those?

[KB2WIG]

RF Parts  prices....    http://www.rfparts.com/index.html


3-500Z       use 3-500ZG
   3-500Z / 8802 – AMPER Amperex NOS 319.95
   3-500ZG –NOS Eimac NOS Call
Eimac 3-500ZG tubes are limited to stock on hand. 
   3-500ZG –GEN Greenstone Generic Tube  154.95
   3-500ZG –AMP Amperex    269.00
   3-500ZG –RFP RFP Best Buy 179.95
   3-500ZG –RFP–MP RFP Matched Pair 359.90
   3-500ZG –RFP–M3 RFP Matched Set of 3 539.85
   3-500ZG –RFP–MQ RFP Matched Quad 719.80
   3-500ZG –PR SELECTED Pulse Rated
HYPOT 14KV@10UA G-P  235.00
   3-500ZG –TAY TAYLOR    164.90
   3-500ZG –TAY–MP TAYLOR Matched Pair 329.90
   3-500ZG –TAY–M3 TAYLOR Matched Set of 3 494.85
   3-500ZG –TAY–MQ TAYLOR Matched Quad 659.80
 

[K1JJ]

Hi Dave,

Many of us have been thru the same thing with big tubes.... we have a certain set hanging around and become partial to using them. Many times there are better tube choices out there for the job. We paint ourselves into a corner and then wish we had sprung for a few bux more for a better finished product in the end. I've done it many times...

I would agree that for a linear, the 3-500Z is a perfect choice. It is designed for this service and will give the most stable performance and cleanest operation.  In LINEAR service, the difference between the 3-500Z and 304TH might be as much as 10-15 db of poor 3rd IMD. This can mean the difference between splatter reports and a clean signal. That is something some builders do not consider when thinking of the choices.

Anyway, I've had great results with the Taylor carbon plate 833C's from RF Parts. I would take a chance on their Taylor 3-500Z's at $164 each, almost the cheapest in their listing.  They have a guarantee, so check it out.

Hope to see you again up at Nearfest in a coupla weeks.

Good luck.

Tom, K1JJ

[K1JJ]

beware 304THs on ebay. 4 out of 5 I bought were no good when 3KV was applied. The flashover was magnificent.

Mine were modulators, so I put 3-500Z's in and turned the bias way down.

Hi Pat,

I've noticed your posts and it looks like you've played with a lot of interesting rigs and configurations.  I'm curious what rigs you have running now and some that you've built?  Any pictures?   It's always fun to see what others are doing.

Thanks.

T

[Opcom]

Hi Tom,

I prefer to restore, rebuild, or re-engineer. With everything from BC transmitters that worked at one time to the worst of what some folks have named "Hammy Hambone" jobs, I get enough variety. I build from scratch when necessary or when what I want is not available or is unaffordable. I am sure my doings are relatively common, since I don't have alot of time to physically craft truly fine instruments.

Here's what is going on now, divided into sections:

/////////////////////////////

What Radio gear is basically in running condition now is:

*USA and Military:
RCA BTA-250 that needs moved to 160M
Sunair GSB-900DX HF XCVR
National NCL2000 amplifier
SDP 1000L RF Power Amplifier
two GRC-106A HF tranceivers (one CARC 8-) )
Racal RA-6217 receiver
two Logimetrics 925 signal generators covering 50KHz to 80MHZ (for VFOs)
ITT FA-8190 UHF AM aircraft tranceiver (military band)
ITT FA-8191 VHF AM aircraft tranceiver (civil band)
two Rycom R1307 receivers
RT524
RT246
two PRC-77
PRC-25
two Racal TRA-967
(also various audio items, I like the sound of vacuum tube gear)

*Riceboxes:
Kenwood TS-430S (with FM, narrow CW, and wide AM filters)
two Icom IC706
Alinco DX-70
Alinco DR-135
Alinco DR-435


/////////////////////////////

Things that I am working on now -need fixed or finished up:
Link 250-UFS 250 watt 6M base station
"Col. Tucker" KW AM rig covering 80-10M
two R-390's
two R-220's
Harris RF-130 (needs about 200 tantalum caps replaced..)
also adding LSB to one of the GRC-106A's


/////////////////////////////

Things that are just lying around forlornly:
six T-282 AM transmitters 200-400MHz (no power MD-141 supplies/modulators)
--note -  these have been known to be able to do 100W ATV as amplifiers
three MD-129 PSU/modulators (for the T-217 'autotune', probably will work with the T-262's)
FM BC exciter, RCA? 10 watts -needs help. (heck it needs a home)
three Collins 618S-1 transceivers
three 1500W CCS RF generators on 13.5 MHz using 3-1000Z's


/////////////////////////////

I have built a scant few from the ground up. The major things I have built from nothing but parts and more or less my own designs were:
3-1000Z SSB-rated linear amp
100W PEP AM-durable linear with two 6146's
50-950MHz spectrum monitor
2000V 260mA power supply
400V 500mA power supply
800V regulated screen supply for 4CX5000
200W 400Hz inverters running from 10-15VDC (to power a Collins 618S-1 servo/bias circuit)
MOPA CW rig with a 45 (before I was licensed -it tore up the TV pretty good)
100W audio amp using 8 6L6G's in push pull parrrrallel
"demonstration platform" version of a microwave directed energy weapon

--The only things of these still around today are the 400Hz inverters, 6146 amp, and the spectrum monitor.


/////////////////////////////

I've done alot more of taking something that exists and making it do what I want or what others want. One of those is a class-A series modulator and power supply for a low power AM transmitter (807, 5 watts carrier). It was made from one of those electronically regulated tube type HV power supplies.

/////////////////////////////

I am between antennas at this time, except for a 40FT piece of wire. Soon, I keep telling myself. Pictures of much of this stuff can be seen at www.bunkerofdoom.com

best regards,
Patrick

[K1JJ]

Hi Patrick,

That's quite the list of gear.  Good quality stuff. It all adds up after awhile.

I noticed the screen supply for the 4X5. That is always a challenge when getting into higher voltages. Rich Measures has a good MOSFET floating electronic regulator circuit for positive 0-2000V.

Have fun and thanks for sharing.

T

[Bill, KD0HG]

I have a linear amp stashed away for when I get time again for radio. It runs a single 304TH in grounded grid, on 2000 V. Perhaps I look in the wrong places, but I'm not seeing alot of NOS 304THs. Is there another triode that will give me about the same power gain, and run on the same plate voltage?

Thanks

David Goncalves

David:

I'm assuming that you don't want to do a complete rebuild on that amp or change out the plate supply.

If your amp is in otherwise good working order, a more common 304-TL will work just fine in it if you change the bias voltage accordingly. A few years ago I experimented with both the 304-TH and -TL triodes in the same amp in both linear and class C modes, and found little difference in performance, output power, or efficiency.

Of course, the sockets, filament voltage, neutralization and required plate voltage are identical between the TL and TH variants. A big problem with other transmitting tubes is the required plate voltage. They won't play on your 2 KV plate supply, most of them need a healthy 3 KV minimum. Same for other options like a triode connected 4-400. This is your limiting factor. The 833-C that Tom suggests might be worthwhile to check out, if you have the physical room for one in that amp. I found them to be bulletproof in broadcast service, and you can't burn holes in the carbon plates- LOL. They might play allright on 2 KV. A parallel pair of 810s might also play well on 2 KV.

By the book, a grid-driven 304-TL in Class AB2 linear mode at 2 KV needs about 50 watts PEP of drive, 500 mA of plate current and will put out around 700 watts of RF. I found that to be a conservative rating, you can easily load one up to 700-800 mA at 2 KV and get a KW and more of RF out in SSB mode. Darn near the legal limit out with a bright straw yellow color on the plate. Don't plan to run the thing key down like that, and blow plenty of air on the seals. If you want to be an animal, 900 mils seems to be the limit at 2 KV, do the math...

At 2 KV, the TH needs -90 volts of bias, idling plate current is around 100 mils, the TL needs -170 volts of bias and the idling current is the same.

You should be able to still find NOS or good used 304-TLs at hamfests, or put the word out. They're stilll somewhat common. Maybe Fair Radio has some lying around.

Single 304-TL meter demos:

[k7yoo]

There is indeed another "triode"--a triode connected 813. Run 2, 3, or 4 of 'em. (I assume you are talking 20M on down). They play nice at 2KV and are cheap. You can get 4 of 'em for the price of a new 3-500Z.
Skip

[W1EUJ]

I want to thank you all for your suggestions.

Bill, you are right - I'm not looking to do huge rebuilding of the amplifer. Given that the -TL variant is of the same construction as the 304TH, do you see the same problems with gas/arcing in the used and NOS tubes you find?

Availability concerns me very much. I held onto the DX-60 because of its spacious chassis and 6146B finals (I expect will be available for some time). The matching amp using a harder-to-find and often gassy tube was concerning. This 3-500Z is very attractive, where late manufacture tubes are available. Any tube that is 304TH-like in resistance, operating plate voltage and current, and heater - AND will remain available 5 years from now - makes the top of my list.

Tom, thanks for all the info, and I do hope to meet up at NEARFest.

[The Slab Bacon]

Just for what it is worth, the problem with the 304ths (and other old power tubes) flashing over may be caused by small amounts of gas developing in those tubes from sitting. This can be a problem with any and all of the big transmitting tubes because they dont have a "getter" in them to flash out the last of the air and other gasses left in them from manufacturing.

If you have new power tubes, or used ones that have been sitting for a long time, it is a good thing to give them a break in period before hitting them with full strap. Fire thm up with just the filament burning for a day or so first, then hit them with just enough plate voltage to get some color on the plates for short repeated cycles for a while. Then hit them with full strap.

Since these tubes dont have a "getter" the filament and the tantalum plates act as the getter. We have had many discussions on here in the past about this subject. Even if they dont need it, it is still a good insurance policy to do this just for good luck.    Just my $.02 worth.

                                                   The Slab Bacon

[K1JJ]

Hi Dave,

I wanted to add a few more points for my reasoning of tube selection in linear amps....  It revolves around how clean the amplifier output will be.

Generally, most of us don't really get picky about this and go running around the band giving out bandwidth reports - so consider this just an info discussion. The following tube talk does not apply to class C finals, just linears.  Old buzzard triodes are perfect for class C, plate modulated operation.


Generally, a conventional old buzzard triode running in grid driven, linear service is good for maybe -18 to -23db 3rd order IMD at best. This is considered rather poor linearity by today's standards.  Yes, it will run OK and produce good power and efficiency, but will add some more bandwidth to our already "wider than other modes" AM signal.

When we try to drive them hard for maximum power out, this poorer IMD really begins to show up.  Run these tubes in pure class A, grid driven and they will get better specs, almost approaching the better linear designed tubes, but the efficiency will be horrendous. In contrast, I've seen  GG triodes designed for linear service (running class B for efficiency) slammed hard and the bandwidth is still tight.

But there is a solution. Running an old triode tube (or some tetrodes)  in grounded-grid "GG" linear service can pick up 10db or so of improved IMD numbers. This makes a huge difference in cleanliness. Tubes like 813's, 4-1000A's tetrodes in GG (and others) can be used and still have good gain for easy driving.  (Every db of negative feedback created by GG operation = 1 db of improved 3rd order IMD = 1db harder to drive the amp)


But the low mu triodes in GG have so little gain and are harder to drive in GG. You almost need a linear to drive a linear in extreme cases... :-)  I've been there.

A tube designed for linear service will do -32 to -38 db 3rd order IMD right out of the box and be easy to drive, and stable as all heck cuz it is run grounded grid.  Another good choice designed for clean linear servce is the 3-500Z, 8877, the 3cx-1200, 3CX-3000F7/A7 triodes (all GG). The 3CX-3000 tube is probably the best choice being it shows -40db 3rd order in older Eimac specs. This is unbelieveably clean, and even more so when run at reduced amateur power levels..

I guess my point is if we are going to be running wider AM bandwidth from the start, might as well go with the cleanest amplifier final we can. No sense adding on unnecessary bandwidth.

Of couse, all of this assumes that the DRIVER is at least ~10 db CLEANER than the amplifier to have a minimal effect on the mixed output signal.. It is cummulative.

As a compromise to KEEP COSTS DOWN  and great availability, I'd choose a 4-1000A in GROUNDED GRID with 5KV or maybe four 813's in GG with 3500V. 

I had a buddy who ran four 813's in grounded grid at 4KV and could deliver a clean 2KW out.   It's a great amp cuz the air is minimal - quiet and effective -   cheap tubes and EZ to drive with a regular 100W rig..

BTW, a good test for ANY linear is to run it on SSB full throttle and have a buddy tune around and take measurements of the crackles and trash generated after the normal signal is out of the bandpass. Use 2.1 kc bandwidth or something similar. It's all relative, so don't worry.  Listen up 3kc, 3.5, 4kc,5kc, 6kc etc. Or put in some programmed taped material into a dummy load and tune yourself in and take the same readings.  Use a receiver away from the rig that uses a 6" piece of wire out the back. Taking baseline relative readings like this will show improvements as you do them.

Compare your amplifier to the readings taken when the driver is barefoot. Your barefoot driver (ricebox) should drop of dramatically after 3kc up the band and be clean. However, the linear will always show poorer products when on line - that is the game... :-)

And remember, most riceboxes are barely better than -32db 3rd order IMD themselves.  Check out the specs and see. They really should be -40db or better to have little effect on the average linear's performance. So when you hear a big difference between your ricebox and linear, you can be assured the linear is probably operating below -28db 3rd IMD.  Those of us with software spectrum analyzers can do some fantastic measurements these days.


For what it's worth...

T

 

[Opcom]

If the transformer is 2KVAC to the rectifier and has good insulation (most important!), one way to cheat it up is to put in series with the cold end of the secondary, a bridge rectifier powered by a lower voltage/same current transformer. Boosts the high voltage. Assuming it will fit.. This will give more than adding a variac to the primary, although that is always a good idea too.

Maybe a nutty idea but I don't think it's dangerous as long as the HV transformer can take the 500V-1KV DC on its cold side.