triode vs tetrode

[GnG8d]

It goes without saying that if you want more power, you need a bigger tube. But what I don't understand is why people stay with triodes and skip tetrodes, or even pentodes.

I unserstand needing an additional circuit for the screen, but you may possibly be able to completely eliminate the need of a driver box in doing so.

So I guess my question is, why aren't more guys building more tetrodes and less triodes?

 

[Toll_Free]

You want the GODS honest truth?

1. Nobody will pay for them. The tet's, while NOT needing the associated driver, need A lot of protection. Protection = $$$s. $$$s and CBers don't go together.

2. You actually OPERATE a tetrode. You MUST pay attention to the screen voltage and grid current when tuning a tet. Most CBers just look for the maximum swing to the right, which in turn starts eating screen bypasses and the screen itself.

3. Most 'box builders' (notice the lack of the word technician) can't get a 10 meter box to work without a lot of 'trial and error'. When you're dealing with 10 dB gain, that's not so bad. When you're dealing with 20 or more dB of gain, all of a sudden a slight layout issue becomes a 1500 dollar tube being sent back to econco.

4. Not many CBers can actually neutralize a tetrode.



I built a line of tetrode boxes before. It was the BIGGEST FRIGGIN HEADACHE I've EVER had to deal with.

THAT BEING SAID, their ARE a lot of guys GOING to tetrodes. When I talk to someone, and they are doing a layout of their box, I almost ALWAYS try to get them to go the tetrode route, if for nothing else than Gdiss(screen)+ Gdiss(grid) is usually higher on a tet than Gdiss on a triode. THEN, as they get more comfortable (read this as blow up a thousand dollars in surplus 4CX250s lol), they can start playing with the big box.

One thing you can do to play with gain is go what I call "modified grounded grid", where you lift the screen grid a bit above ground (but leave it RF grounded), and the grid directly grounded. IE, take and put about 80 volts on the screen of the 4-400 in a SB220. Use the existing grid chokes to feed it, ground the GRID pin directly, and throw 50-80 volts on the SCREEN GRID of the 4-400s, using the grid chokes you yank out of the SB220. THEN, gain shoots WAY up.

My favorite tube is still the 4-400. A peaked 29 or 2000 will drive it to full Pout.

BUT, the major reasons are expense (most CBers just end up purchasing the NEXT size box. They don't start out looking to run 10 grand, they buy a 250 or 500, then all of a sudden, they are at a pair of 3-500s or a GS35. NEXT, they look at the 3CX3000 or a 4CX5000, as they can be driven to full Pout with the LAST amp they bought.

BUT, if they'd have purchased a SINGLE 4CX3000 in the FIRST place, they'd be running 10 grand for a couple years prior, have a single 40 watt driver, and be money AHEAD, since they only had to purchase ONE power supply!

Of course, this doesn't take into account the CBer factor, where by tuning so all meters swing MAX to the right we burn the screen up 3 times in our tube, rendering it UNREBUILDABLE and then having to purchase another.

Sneaky Pete builds a decent 4CX3000 box, I've seen a pair in his burban about a decade ago.


--Toll_Free

 

[GnG8d]

4. Not many CBers can actually neutralize a tetrode.

That surprises me to hear that a well designed, expensive tetrode would even need neutralized. It's the physical dimensions and construction that create instability at higher frequencies isn't it?

I built a line of tetrode boxes before. It was the BIGGEST FRIGGIN HEADACHE I've EVER had to deal with.

To get the tube to work properly, or from the "operators" who were buying them?

One thing you can do to play with gain is go what I call "modified grounded grid", where you lift the screen grid a bit above ground (but leave it RF grounded), and the grid directly grounded. IE, take and put about 80 volts on the screen of the 4-400 in a SB220. Use the existing grid chokes to feed it, ground the GRID pin directly, and throw 50-80 volts on the SCREEN GRID of the 4-400s, using the grid chokes you yank out of the SB220. THEN, gain shoots WAY up.

My favorite tube is still the 4-400. A peaked 29 or 2000 will drive it to full Pout.

Your talking about converting a 220 to a single 4-400 correct?

BUT, the major reasons are expense (most CBers just end up purchasing the NEXT size box. They don't start out looking to run 10 grand, they buy a 250 or 500, then all of a sudden, they are at a pair of 3-500s or a GS35. NEXT, they look at the 3CX3000 or a 4CX5000, as they can be driven to full Pout with the LAST amp they bought.

BUT, if they'd have purchased a SINGLE 4CX3000 in the FIRST place, they'd be running 10 grand for a couple years prior, have a single 40 watt driver, and be money AHEAD, since they only had to purchase ONE power supply!
--Toll_Free

Yeah, that's where I am :headbang

I've talked to a couple of guys recently about a 3cx3000 box to drive with my sb200 (terribly unstable, I know). None had any in the works or on the drawing board. However the one guy had a 3cx10000 and most of the components to complete a box. I actually entertained that idea and figured that I could drive that with my sb220. But I decided that even though I have hardline, clamp type connectors, 220vac etc that this was just too much in my mind.

So, back to thinking about the 3cx3000. But what really appealed to me was getting rid of both Heathkits and getting 10k with a 100w driver. So poking around the net, I got the idea of taking a two (2) hole 3cx3000 rf deck and putting a single 3-500z in one of the holes. Building completely separate rf cages, tuning, circuits etc, but I have no idea what kind of pitfalls might be encountered there.

That's when the tetrode idea popped into my head and I knew that there was a reason that you don't see/hear many. So it's not the tube, it's the builders and I gather that most of them don't have a true grasp on a triode.

 

[Shockwave]

Start asking most 11 meter amp builders questions like how do you handle negative screen current, what impeadance transformation and Q is appropriate for the grid input circuit, how to regulate the screen and bias supplies, or how to safely neutralize a large tetrode and you get that "duhhhh" look! Sure the tetrode has significant advantages over the triode in terms of gain and reduced drive power. The number of large broadcast tetrodes properly working on this band can probably be counted on your fingers.

Simple protection of any grid can be done without significant added cost. All it requires is the use of inexpensive opto-couplers that interrupt your keying relay and are triggered off existing meter shunt resistors. Ameritron has grid protection with an add on board for $30 that could be adapted for screen grid use on a tetrode.

Those pesky screen bypass caps blowing up just when the amp starts making good output has plagued tetrode builders for years. This is a big tetrodes way of insisting you build it a shunt regulator for it's screen grid. It's all about secondary emissions and negative screen current that wants to drive the screen voltage up towards the anode voltage through this second current path within the tube. Strict control over the screen voltage is the way to get the tetrode linear to "take off" without a bang.

No matter how well you design a tetrode amp it almost always requires neutralization because of the high gain. This has to do with the inter-element capacitances within the tube that tends to couple the input stages to the output stages, causing the amplifier to oscillate. Neutralization cancels this problem by taking a small sample of the output, feeding it through a cap or coil to invert it's phase 180 degrees and applying this signal back to the input.

 

[Toll_Free]

Almost all tetrodes need neutralization. It's the nature of the gain. I've also built single sweep tube drivers (grid driven) driving 3-500s before. It was JUST about as much of a PITA as building 4-400s... Actually, the 4-400 as a tetrode is BETTER than an 8950 or 6??6 driving a 3-500. Cheaper, in the long run, as 4-400s are PLENTIFUL as broadcash pulls.

The problems where with operators that where blowing up tubes, or from the guys that wanted to swap out the 4CX250s for 350s or 300s. I even had one come back with F series tubes (they had swapped the sockets out with the coaxials, but forgot they ran on 26 volts). CBers are a CHEAP lot, even the 'big money guys', for the most part.

As Shockwave pointed out, you can nearly count all the broadcash tetrodes on 11 meters with 1 hand.... That's a bit off today, there's been a LOT of tets being built, BUT for the most part, it IS true. If you want to hear a REALLY good one, listen to 766. His is a good tetrode. Another thing most of the tetrode runners WON'T tell you is this: They NEVER go off frequency. They might as well have a crystal controlled exciter on 27.025. If they do, they don't run the biggins when they do... Hence you hear the guys on 11 talkin about runnin their baby drivers (talkin on the OLD pumphouse, before they bought the tet, while the tetrode cools or warms up).

And no, I was talking about converting a SB220 to a PAIR of 4-400s. HOWEVER, I also build SINGLE 4-400 boxes when I get enough parts. They make AWESOME drivers! 20 in, 700 or more out (depending on the emission left in the tube, I get all broadcash pulls).

In the SB220 conversions, you leave the drive configured at the cathode (into the filaments). You ground the CONTROL grid (the 'grid' in the triode) directly to the chassis. THEN, apply voltage through the chokes you REMOVED from the grids in the SB220 to the SCREEN of the 4-400. If you have a variable or variac'ed 150-200 volt supply, all the better, you can use it as dial-a-gain! (This is NOT something I recommend, however.... Just figure out how much voltage you need, and get a supply built for THAT voltage!!!) HOWEVER, make SURE you "AC ground" the screen grid by using the bypass caps from the control grid (remove the caps and the coils from the control grid, move them to the screen grid. Install the cap JUST as you removed it, but instead of going to ground on the choke, you tie the POWER SUPPLY end of the two chokes together (V1 and V2), and then out to a low impedance voltage source).

If your in the 'planning' stages, get a YC156 and a 400 watt driver, that's capable of operating at 225 or so watts. Problem solved. That, currently, is my favorite tube.

OR, you can ALSO go with the 4CX5000. Mine, 140 watts in, 14 thousand out. I can PUSH it to 22 thousand (low bands, about 18 on '10'), but the distortion is pretty bad. At 14 thousand to 14 and a half, it's clean as a whistle. It's run as a TRUE tetrode, though.

When I ran it as a grounded grid/screen triode, it took (IIRC) about 3500 in for the same 22 thousand out. I generally ran it about 12. The 4CX5000 has a 6Kw anode on it, an indestructable screen (especially if you go with the R version), and is socketed (sockets are FRIGGIN expensive!!!).

Another thing you can do is buy the 3CX10,000 box, and unless he's building it with the B series tube, it will have a socket in it... You can trade off the 10,000 and get a 5K for it, and get a few dollars in the process. Drop the tube in, and with a SERIOUS minimal amount of retuning, it should be back up and running. Just make SURE you match the filament voltages!!!! Not all tubes are the same fil V, even if they use the same socket!.

Seriously, though, where I in your position, I'd use the SB200 in LOW, underdrive the HELL out of it with a peaked up radio, and run a YC156. 10Kw +++ capabilities (I've seen 15 out of a single, but the V was pretty high, and we conditioned the tube for DAYS to debarnicle it).

8877, YC156, YC179, 3CX800s. All triodes that have the GAIN of a tetrode. And ALL of them run on (relatively) low plate volts!

--Toll_Free

 

[GnG8d]

...... I need to sleep on this and digest it a bit =)

 

[linearone]

buy a bigger antenna

 

[GnG8d]

I have a direct fed quad, that's not what this is about.

 

[Toll_Free]

359,

Also, the EASIEST way to run a tetrode is to run it grounded screen. Then you return the power supplies to the cathode and associated circuits to ensure the internal elements are STILL seeing the proper ratio of voltages, but as shockwave pointed out, you have problems with the screen grid trying to take off and assume the Plate V.... Which causes screen bypasses to blow up.

Also, as he pointed out, if you're NOT going to run 'grounded screen' (and Orr's handbooks have a great rendition of this circuit for (iirc) the 4CX1000 in them, it can be scaled), then you BETTER run a low impedance voltage source to the screen, and bypass it with enough R to hold the screen voltage constant (Dissipate 2 to 3 times the expected screen current in the screen bleed R). It's not as simple as just running a resistor off the plate supply when your trying to build an actual LINEAR amplifier....

YC-156 and the 8877. Your friends. Want radio drive? Use a 3CX800 to drive anything else.

These are modern triodes. As Reid Brandon has pointed out, they have the gain of a tetrode, but the stability of a triode. The YC156 I don't even use a suppressor in! A few ohms of CARBON R in the drive lead and you have enough NFB to clean it up as well. Simple as pie, as momma used to say.


--Toll_Free


--Toll_Free

 

[GnG8d]

OR, you can ALSO go with the 4CX5000. Mine, 140 watts in, 14 thousand out. I can PUSH it to 22 thousand (low bands, about 18 on '10'), but the distortion is pretty bad. At 14 thousand to 14 and a half, it's clean as a whistle. It's run as a TRUE tetrode, though.

Here is where I'd like to be ultimately. 100w from my Yaesu and 10k out. I think 8-10k is a pretty decent threshold for some reason. It's that diminishing returns thing again. Cost per db seems to skyrocket after this point, but you are making the YC156 sound very attractive.

If your in the 'planning' stages, get a YC156 and a 400 watt driver, that's capable of operating at 225 or so watts. Problem solved. That, currently, is my favorite tube.

Seriously, though, where I in your position, I'd use the SB200 in LOW, underdrive the HELL out of it with a peaked up radio, and run a YC156. 10Kw +++ capabilities (I've seen 15 out of a single, but the V was pretty high, and we conditioned the tube for DAYS to debarnicle it).

I have a 2517 that drives the sb200 to ~375w, it doesn't have a hi/lo but it could have. That puts me back to a driver however. The 2517 is currently boxed up and will probably be sold eventually. My 2970n2 would fit the bill with no driver, but that's my mobile, I don't even own a power supply, lol.

I guess it's a tube thing. For some people it's a hassle, but I like the additional operational element that tubes add. But a tube radio, driver and final get to be a bit too much, hence my wanting to eliminate the driver. I guess we all have a limit to the "fiddling". Not to mention I'll have to liquidate some pieces to get where I want to be limiting driver options.

I'll be back, still mulling this over.

 

[Shockwave]

359, out of all the advice you've received for a choice of tube, the YC-156 is probably going to produce the most power with the least headaches and cost. It beats the 3CX3000A7 in dissipation and gain. The gain comes at the expense of warm up time on the oxide coated cathode. Tubes using this type of cathode take several minuets to warm up.

They produce stage gains in excess of 20 times the drive power where the average directly heated triode usually produces just over 10 times the drive power. I'm not sure I would agree they act like a tetrode because unless it's control grid is heavily swamped with resistive loading, the tetrode can easily exceed a 100 to 1 drive ratio with their gain.

 

[GnG8d]

No matter how well you design a tetrode amp it almost always requires neutralization because of the high gain. This has to do with the inter-element capacitances within the tube that tends to couple the input stages to the output stages, causing the amplifier to oscillate. Neutralization cancels this problem by taking a small sample of the output, feeding it through a cap or coil to invert it's phase 180 degrees and applying this signal back to the input.

Okay, I see now. I was under the impression that it simply had to do with the lead lengths, grid structures, grid connections, anode to cathode shielding etc, but I see exactly what you are saying now.

 

[GnG8d]

And no, I was talking about converting a SB220 to a PAIR of 4-400s. HOWEVER, I also build SINGLE 4-400 boxes when I get enough parts. They make AWESOME drivers! 20 in, 700 or more out (depending on the emission left in the tube, I get all broadcash pulls).

Assuming linear gain, that's 100 in and 7k out of a pair, correct?

Another thing you can do is buy the 3CX10,000 box, and unless he's building it with the B series tube, it will have a socket in it... You can trade off the 10,000 and get a 5K for it, and get a few dollars in the process. Drop the tube in, and with a SERIOUS minimal amount of retuning, it should be back up and running. Just make SURE you match the filament voltages!!!! Not all tubes are the same fil V, even if they use the same socket!.

That deal is off the table, lol. I called the guy about a 3000 box and he offered up the 10000. When I didn't immediately commit and instead told him that I needed some time to think about it, he shot me an email the next morning saying to hell with it, he was tired of messing with CB'er and he'd just keep it and build a multibander out of it *shrugs*

 

[GnG8d]

359, out of all the advice you've received for a choice of tube, the YC-156 is probably going to produce the most power with the least headaches and cost. It beats the 3CX3000A7 in dissipation and gain. The gain comes at the expense of warm up time on the oxide coated cathode. Tubes using this type of cathode take several minuets to warm up.

I'm thinking that you guys are narrowing it down to this tube fairly quickly and I appreciate both of you guys taking the time with a relative noob.

 

[Shockwave]

Some more things I forgot to mention about the YC-156. Where directly heated thorium cathodes can be run at slightly reduced filament voltage for increased life, don't try this with an oxide coated cathode. It has the exact opposite effect on these tubes and shortens the life.

The YC-156 was also made for one specific purpose that is now becoming obsolete. It was custom designed for use in many MRI machines and may even be out of production as we speak. Apparently the RF generators in these devices have progressed to the point where this tube is no longer used.

That being said, the tube is fragile and may become difficult to replace in the future. Buy a spare now and I strongly recommend using some type of grid over load protection. I believe the tube has a 5KW anode dissipation but the grid is no stronger then the 8877 at only 25 watts. This makes for some very low IMD when used as a linear but easy to melt the grid.