8935 tube - good for anything?

[Opcom]

My friend found in the back of some warehouse a big electrical machine of some kind that has an 8935 / YD1185 in it. He does not know what it is. The tube and whatever else is in the RF section is behind shields with all kinds of ominous warnings about dangerous voltages. I tried to ask if it has a pair of copper bus bars or copper pipes coming out (induction or dielectric heater) or if it had an RF connector (RF source for plasma) but he did not know. Has anyone messed with an 8935? From the curves in the datasheet I do not know if it would be a good amplifier. It is intended as an oscillator. an EME's dream anyway. STRAP the moon.

I don't know if we can get it or not. It is too powerful for me I will admit that right now. It also is probably heavy but there is a forklift..

[Steve - WB3HUZ]

Strap the moon and anyone else!

[Opcom]

looking at the plate curves, isn't the tube un-linear in some respects for amplifier use, or does that not matter too much? anyway the grid dissipation is huge on it. be hard to burn it out. I want to get some pictures of the thing..

[K1ZJH]

175 amps @ 7 volts just to light the filament!!! 

[Opcom]

Today I had the priviledge to poke about in another warehouse. This one the lair of the aforementioned electrical machine. The place was mostly full of old lathes and mills and tools and weird 3-wheel scooter carts and out in the yard was a giant forklift bigger than a deuce and a half.

The machine was appropriately in a dark corner beside a pile of diesel fuel cans and a filing cabinet. Anyway, the tube is a 4CX15000. The date code or whatever, was 8935. ahem.. Never send the electroplating power supply repair gentleman to do the radio repair gentlemans' work.. But my friend has an eye for the finest goods.

The first things that jumped out at me were the very large 3-phase transformer with diode stack and what looks like an oversized bleeder stack on top, the decent size oil cap, and the whole deck of electronics, looks like too much just to power the 4CX15000. Also, no plate coil or cap, or other noticeable circuit. Indeed the anode of the tube is connected to the power supply. After inserting head and flashlight into the belly of the beast, the RG-8-like coaxial cable emerging from the unit seems to also be connected to one side of the oil cap.

I do not know what I have here. I see a 4CX15000, air sytem, and more little stuff than a 4CX15000 needs. Is something missing? Nothing seems to have been removed. No control head and no local controls. Is this an old electron beam power supply? Maybe the pictures will give a clue. The only mark on the unit is "Telemark Industries" and that leads here to the correct logo: http://www.tfi-telemark.com/. I don't know what to do about this but it is interesting. The class A series modulator from hades? I have the sense to not buy it, I think.

[K1JJ]

Just fantasizing now...

Your original idea of running it as a class A modulator or series modulated class C screen or grid modulated final would be an interesting idea and could work. Though, not very well for linear amplifier service.

A linear amplifier is probably the best use for a tube of that size, but here's the problem...

That tube will put out 28KW pep with the proper power supply.  It's designed more for class C or modulator service, not high quality linear service. The 4CX-15000(J)  is better for linears, but are almost impossible to find cheap.

The problem is that running as a standard tetrode in grid driven linear RF service will give about -32db 3rd order IMD numbers at best. This is a normal ham amplifier cleanliness figure.  28kw is about 13db stronger than a 1500w amplifier. The result is ON THE AIR, the relative side crud products  will sound like a 1500W amplifier with a -19db 3rd order IMD.  Your ham neighbors will come over and lynch anyone with a signal that nasty... :-)

To clean it up, a tube like this would require a hard-to-tame RF negative feedback system or a feed forward system that is even more complex.

To run QRO, stick with a power grid triode designed for grounded grid linear service. An 8877, 3CX3000A7 and it's relatives are the best choice for cleanest linear amp results.


73,

Tom, K1JJ

[WA1GFZ]

That 4CX15000 thing looks like a regulated HV power supply. That bottom assembly looks like it could be a nice RF amp. You could ground all the grids and drive the cathode.

[W3SLK]

Or...... you could sell it to one of those CB keydown phanatics for some ungodly sum of money and get yourself something along the lines of a 3CX5000(F???) for a nice tidy little linear

[W3RSW]

Man would you look at the blower on that thing ...  Must be one of Dayton's biggest! 

And it comes with an emongus fil. transformer.

Hey, if nothing else it'd be a well regulated screen supply for RKW's 813 amp.
(put a bunch of 813's in series and pick off from a divider for all the screens...   why? dunno, maybe because you can?)

[Opcom]

It's a nice piece of equipment, but with what Tom said about the IMD on the tube I would hate to go to the expense and trouble of building an amplifier only to operate on several bands at once. The advice to turn it for the $ for a 3CX3000 is good but them CB nuts are cheaper than hams.

I admit no experience GG-ing this tube family but I have been told by one guy who GG-d a 4CX5000 that great care has to be taken during operation that the loading is enough, or the grids will fry before the operator knows it.

I had thought of getting around this by grid-driving it in the same way as the "triode" connection is used with the 807, where the control grid is driven through a resistor and the screen is where most of the business happens. Much less power would be needed for drive of only the grids.

If anyone has experience with GG-ing and driving the 4CX5000-15000 I would certainly welcome some comments.

Maybe a good use would be to sell it or for a tesla coil; I will refer this to my apprentice but he probably can't afford it.

[W3SLK]

Drop Rich Measures a line at . He would be into it. He built a box with a 4CX5000 and then one with a 4CX10000. Here's his URL: http://www.somis.org/ I bet he would have the low down on that bottle.

[K1JJ]

Yep, Richard Measures does advocate these tetrode tubes in grid driven service.  But again, when you run 10db more power than the average ham amp, you need it to be 10db cleaner than the average ham amp - or they will complain.  A grid driven amp WITHOUT heavy RF negative feedback won't pass the test.  Some grid driven tetrodes in LINEAR service designed for class C operation (many) can't do any better than -25db 3rd order IMD when driven to near full power. Some are even worse. 

Maybe it won't get complaints on AM cuz of the common wider signals and the large variation in bandwidth standards from station to station - but on ssb where everyone demands you be 3kc wide and down 30-40db after that, this is where the problems start.  If the DX signals at sunset (75M DX window) are S7 and your neighbor is S9+60 over, you can imagine that even a "clean"  -35db 3rd order IMD signal is gonna trash you 5-6 kc away and take out the weak DX.   


For linear grounded grid (GG) operation driving the tube in triode config, you will need at good 1500w pep output amplifier for drive. Figure on 10db of gain, more or less for the 4X15 GG stage.  I would think that combo would produce no better than -32db 3rd order too. Still a problem for a QRO amp. 

Adding screen and grid voltage to a GG amplifier (cathode driven) DOES NOT make it easier to drive, contrary to popular opinion.  As you increase screen voltage, you need more grid bias to idle it down, thus making it harder to RF drive the grid back to zero volts.  At least that's W8JI's opinion and based on my own personal experience.


Bottom line, when extreme QRO, you need an overall amplifier chain that yields -40 to -45db 3rd order to blend in.  It is a very difficult thing to do for the average builder.   You can always tell the "-32db IMD" guys running extreme QRO (on ssb) by relativity to other signals - simply tuning off their signals and hearing them up the band 5-6kc.   Once you "normalize" your receiver for the present conditions and the average 1500W signals, these guys stand out like a sore thumb.... especially when angles are high and locals are loud.

When someone pulls off extreme QRO with a super clean signal and no blower noise, now that's really doing something... :-)

I've seen many guys get run off the 75M DX Window due to their splatter. Many come back later with super clean amplifiers once they've paid their dues. I went through it myself before I knew any better.

A single 3CX-3000A7 using a very clean class A solid state Mosfet amplifier (300w) as a driver can run -45db 3rd order right outa the box.  Plenty of power and stability without RF neg feedback required. That tube is possibly the cleanest GG tube out there and available.  Another good choice is the YC-156 (3CX-5000) of  medical magnetic resonancing fame.  No socket required and beautiful specs.  Reasonably priced used pulls.  Loaf these tubes along most of the time at 1/2 peak power and you're golden - clean as a whistle.


BOTTOM LINE?  If you're gonna go through the tremendous effort of building a big amp, give yourself the best chance of success by picking a good, clean tube designed for the particular service,  from the start.

T

[WA1GFZ]

My 4CX3000A is GG and never had IMD problems. Plenty of guys have run tetrodes like 4-400 and 4-1000As in gg. Grids will take plenty of power.
you have to be careful with 4CX1000s as they will fry in gg.
Or you could grid drive it. Either way that is a nice assembly that should not go to a CBer

[K1JJ]

My 4CX3000A is GG and never had IMD problems. Plenty of guys have run tetrodes like 4-400 and 4-1000As in gg. Grids will take plenty of power.
you have to be careful with 4CX1000s as they will fry in gg.
Or you could grid drive it. Either way that is a nice assembly that should not go to a CBer

Frank,

Sure, that's right - plenty of hams run them that way... but at ~ 1500w out.  My point is if someone were to run a LARGE tetrode at extreme QRO in any config without extra precautions to increase its cleanliness, they will have complaints simply cuz of the relative signal increase over the average 1500 watt ham signal on the band. (relative crud increase)

Have you ever measured the IMD for the 4X3 GG at FULL strap?

-32 db 3rd specs is probably being too kind for any non-RF feedback tetrode, even in GG, full power.  

But running it at 1/2 - 3/4 of full pep does wonders.

Bottom line is we can pick up another -10db or more better IMD numbers simply by running a tube designed for the job, like the 3CX-3000A7 running reduced power, but still QRO.

But we all run what's available and cheapest - and that's quite OK... :-)

T

[KD6VXI]

It's a nice piece of equipment, but with what Tom said about the IMD on the tube I would hate to go to the expense and trouble of building an amplifier only to operate on several bands at once. The advice to turn it for the $ for a 3CX3000 is good but them CB nuts are cheaper than hams.

I admit no experience GG-ing this tube family but I have been told by one guy who GG-d a 4CX5000 that great care has to be taken during operation that the loading is enough, or the grids will fry before the operator knows it.

I had thought of getting around this by grid-driving it in the same way as the "triode" connection is used with the 807, where the control grid is driven through a resistor and the screen is where most of the business happens. Much less power would be needed for drive of only the grids.

If anyone has experience with GG-ing and driving the 4CX5000-15000 I would certainly welcome some comments.

Maybe a good use would be to sell it or for a tesla coil; I will refer this to my apprentice but he probably can't afford it.

I've run grid driven, as well as strapped to triode, 4CX5000 A and R tubes.  By strapping them both to ground, you have a huge triode.  Problem being, equalizing currents between screen and grid.  It can, and is done, but isn't the "best" way.

You can split the currents, Bill Orr, as others, have designs in his and other radio style handbooks.

The 4CX5K is good for about 4 kilowatts of RF carrier, AM modulated.  Multiply that by the ratio of plate dissipation, and you can figure out what you can get out of each tube.  Figgir, 3Xs the dissipation is what the amplifier will do, PEP.  Of course, carrier is a quarter to a fifth.

Measures never finished the 10,000 tube.  The 5000 tube never worked above 15 meters.

Lots of people using the tetrodes (and triodes) with handles are going to using a small amount of L in front of the Tune cap.  This makes tuneup easier, the Pi a bit more broadbanded, as well as lowers operating Q.

If you can purchase it for a low enough price, the power supply can be ran from single phase, albeit at a lower Pout.  The tube, etc. can be worked on all bands we use it on, as well as others.

I've never measured IMD out of the 15,000, but it can't be all that bad, as it is used as a broadcast tube.

Easiest would be strap it all over as a triode, if the price is right.

CBers see about 35 kilowatts out of that tube, on 11 meters.

--Shane

[Bill, KD0HG]

I've never measured IMD out of the 15,000, but it can't be all that bad, as it is used as a broadcast tube.

We use 4CX15000As in Class C FM rigs. They'll do 20-25 KW output on VHF with around 250-400 watts of drive. (Or a pair of 4CX250Bs).  VERY rugged grids. Plate voltage typically ~ 9 KV; current around 3 amps. At a more reasonable 4 to 5 KV they'll make about 9 to 10 KW out in Class C. There's piles of used ones out there at almost every high power FM broadcast station. I've got the Eimac spec sheet on them at the office, if anyone wants to see it I'll post it here.

You could probably run one in Class A linear if you wanted clean output...And to run up the electric bill..

Tom, would the IMD specs likely be any better in Class A rather than AB?

That was THE final in a lot of Harris FM transmitters.

If I was going to use one on AM, I'd run it Class C, screen modulated.

[K1JJ]

Tom, would the IMD specs likely be any better in Class A rather than AB?
If I was going to use one on AM, I'd run it Class C, screen modulated.

Hi Bill,

You know all this, but here goes... :-)

Yes, I'll bet a 4X15 wud do nicely in Class C screen modulated service since it was designed for that - more or less.  I know the Tron had one for commercial service BC and it worked out just FB at 5KW carrier with big audio.

Yes, idling the tube as high as possible makes a tremendous difference in IMD. (In linear, at cut-off  being the minimum and full power idle for class A.) Most run it somewhere in between.  Pure class A is such a power hog and heat producer, it wud have to be for a very good reason.  If for pure linearity, it might be best to consider the 4CX-15000J or the 3CX-15000(J?) designed for the service.

I like using pure class A for lower level driver stages.

The second important factor is the design of the tube - what is it best suited for?

And third, how hard is it driven? If I remember correctly, simply by reducing a linear to 1/2 power from max pep will improve the IMD by -9db. That's quite a lot.

And lastly, the configuration the tube is in matters a lot for cleanliness.  For example a guy running a grid driven pair of 4X1's with 7kv driving it with an SB-200 is gonna have the lynch mob as his house - or at least people yelling every time he unkeys the mike... :-)  The same tubes running GG, 5kv and running 1/2 pep max power would be as clean as a whistle.


It's all relative to the current "standards" out there. Years ago you could run a spark gap and get away with it. Try that in 1935 when they had DC CW and you were toast. Later, the first ssb rigs drifted and had poor suppression but no one cared. Eventually the Collins rigs set the new standard and you were on the outside looking in with the older ssb rigs.

Today the standard is the modern ricebox and the GG linear using tubes designed for ssb linear:  3-500Z finals, 3CX-800's, 3CX-1200's, 8877's etc.  Get on the air with this rig and you blend in nicely. But get on the air with something of lesser clean design and you may find you start to stick out, even though it was quite acceptable only a few decades ago..

It's all relative.

BTW, we called for you a few times last night. There was a void in the south west - Colorado area... :-)


73,
Tom

[k4kyv]

When someone pulls off extreme QRO with a super clean signal and no blower noise, now that's really doing something... :-)   

Slopbucketeers got just as screwed with Johnny Johnston's P.E.P. bullshit as did AM'ers. Since the peak-to-average ratio of the human voice waveform averages something on the order of 7 to 10 dB, if the legal power limit were defined in terms of average power, you could run a clean signal to about 750 real watts output while letting the p.e.p. go where it may, which could be as high as 3750 watts or more on instantaneous peaks.

But still, it is average, or mean power that determines the loudness and interference potential of a signal, not the instantaneous power of occasional voice peaks.

Limited to 1500 watts pep, a clean SSB signal with a typical voice and no processing would average no more than about 300 real watts, as measured by an rf ammeter working into a known resistive load or by an accurately calibrated average-reading wattmeter.  To get the average power up to something on the order of 750 real watts requires heavy processing or else driving the leen-yar well into saturation and thus flat-topping on voice peaks, both of which make for a crummy distorted signal, and most probably, spurious sideband products that make the signal wide and splattery.

So this makes a shitty sounding over-processed SSB signal "legal", while a clean undistorted signal of equal loudness would violate the limit.

[KM1H]

Put the tube on Ebay and grab a YC-156 which are readily available as pulls for $300-500; less if you get one from a ham who never finished his amp.

That tube will run 1500W with 100W drive and 3500V for the next 100 years or drive it at around 300W at 6-7 kV for serious smoke. Probably the most linear tube available these days at a reasonable cost. The 8-10 minute warm up annoys some   but the filament power is pretty modest so once on, leave it on for your operating time. No socket is a big plus.

It has huge output C so you need some L before the Tune cap if you want 10/12M. Many have even got them working fine on 6M.

Carl
KM1H

[K1JJ]

Put the tube on Ebay and grab a YC-156 which are readily available as pulls for $300-500; less if you get one from a ham who never finished his amp.

That tube will run 1500W with 100W drive and 3500V for the next 100 years or drive it at around 300W at 6-7 kV for serious smoke. Probably the most linear tube available these days at a reasonable cost. The 8-10 minute warm up annoys some  but the filament power is pretty modest so once on, leave it on for your operating time. No socket is a big plus.

It has huge output C so you need some L before the Tune cap if you want 10/12M. Many have even got them working fine on 6M.
Carl
KM1H

Yep, wonderful advice, Carl!

That wud be a great transaction.

The YC-156 (3CPX-5000) is the perfect ham tube.  You could go thru 3 sets of the common linear amp tubes before a YC-156 would die, if run conservatively.  I figger if they're good enuff for the demands of high-resolution linear magnetic imaging in the medical whirl, they'd good enuff for RF linear amps.

It has the same filament structure as a 3CX-15,000B7, and is indirectly heated so it takes relatively little power to light up -  but with it's smaller plate external anode, there is a greatly reduced limit to its dissapation, of course.

I once built a 6M amp with one, but ended up using an 8877 instead. Yes, the output C wuda been a challenge on 6M.


I wish the used price wud come down so I can buy a spare.. :-)  It will - wait a year or so.

The 8 minute warmup doesn't bother me much. Just have to miss a few of those impulsive DX contacts, I guess...
I run mine at much reduced power and the IMD is great. Have never had a complaint once in several years using it... can't say that about other tubes. I'm now building up a class A solid state driver to give it an even better chance. I'm coming off the low level 10mw ricebox output (FT-1000D...clean -70db 3rd order) and going from there into a class A chain.   This way, the IMD can never be worse than the final itself.

I was surprised to learn that most 100w (class B) riceboxes have trouble doing better than -30 to -32db 3rd themselves. That would contaminate the signal before it even gets to the final.

By idling the YC-156 high, running it at greatly reduced power, and driving it with a super clean class A signal, it should approach -40db 3rd or probably even better ... at least that's my goal.


73,

T

[Bill, KD0HG]

Now that I think of it, the 3CX2500 series is also easy to find.

In spite of the tube number, they've got ~4 KW of plate dissipation. The 2500A7 and 2500H3 has leads and doesn't need a socket. The 2500A3 has no attached leads, but it's easy enough to attach connections to the base with hose clamps. It's filament voltage is a common 7.5 volts (at 50 amps).

This jug was used in the Collins 21E and Harris MW-5.

Many years ago I build a grounded grid linear amp using one. Power gain in that mode was around 6. I floated the filament transformer and put the RF chokes in the primary so only around 3 amps needed to be handled by it.

Grounded cathode AB2 the thing would do 6 KW out with 150 watts of drive.

IIRC, there was also a 3CX2500 variant specifically designed for low-IMD linear service (I don't remember the suffix), but they're gonna be scarce if you can find one. The A3s, F3s, and H3s are all currently made by CPI-Eimac, in China and by Svetlana.

[K1JJ]

Now that I think of it, the 3CX2500 series is also easy to find.

Grounded cathode AB2 the thing would do 6 KW out with 150 watts of drive.

How 'bout that... I did the same thing, Bill.

Back in the mid 70's I first tried one in GG linear and found I needed a 4X1 to drive it... :-)

So then went to grid driven linear config and even neutralized it. (It sure needed it) That worked better, as you found, but wasn't too clean.  I worked at it and found it cleaned up somewhat when I ran it in full A1 dissipation...  :-)  Gawd what a power hog.  So went on to something else.

It sure is a common tube.  For AM, I'll bet running it in class C, neutralized, grid modulated would  be just fine. 

Of course, the old standby, one modulated by a pair is the obvious choice of the commercial guys.

Hope to catch ya on, maybe this weekend, OM.

T

[Bill, KD0HG]

Looking forward to it.

Most likely I'll be out of the Ghetto around 3700, + or -. I've got my antenna lowered some 50 feet because of recent winds, I'll hoist it by this weekend.

That linear with the 3CX2500 was the first one I ever built. Yer right, definitely a drive-hungry beast. I could have best used something like a 30L-1 to drive it. It did make enough juice to melt my RG-8 feedline one wet night, the shorted output caused it to blow the 30A breakers in my parents' house...The current surge made the metal conduit *BANG* like it was hit by a hammer..My dad almost ended my budding ham career right then and there. That, on top of my locking the thing key down and running around the front yard waving a lit four-foot fluorescent bulb like a star wars light saber..

b

[Opcom]

Well let's see how much the old boatanchor is going to cost. Better than having it rot forlorn in the warehouse. Yall have fun with yer radios! I hope to have a real antenna in a frew months. Tree service comes Tuesday.

[K1JJ]

It did make enough juice to melt my RG-8 feedline one wet night, the shorted output caused it to blow the 30A breakers in my parents' house...The current surge made the metal conduit *BANG* like it was hit by a hammer..My dad almost ended my budding ham career right then and there. b

heheheh -  Yep the parents just didn't understand it was part of the shakedown... not necessarrily a bad or permanant thang.

With that same 3cx-2500, I too, had a short circuit, but in the amplifier.  I had used #6 wire run around the cellar to the mains. I didn't use a fuse or breaker cuz I didn't have one.  That #6 wire melted and smoked so bad it filled the cellar with a gagging black smoke before I cud turn it off.  It happened so fast. I opened the hatchway, but it was too late - the OM smelled it thru the floor and came down stairs and freaked out... :-)   

He didn't shut me down, just yelled. I guess he understood since he helped me from the Novice days. In fact, he took Ed Hare (W1RFI) and me up to Boston for our General tests in June, 1966, so he knew a little about the insanity.  I had failed the Conditional test a month before (bummer) but we both passed the General this time.


Patrick: Well, you've sure got some good parts to work with until that antenna goes up... :-)

T