Are 3CX800A7 a good tube for longevity/SSB?

Discussion in 'Amateur Radio Amplifiers' started by KK4YDR, Feb 1, 2016.

ad: L-HROutlet
ad: l-rl
ad: abrind-2
ad: L-MFJ
ad: Subscribe
ad: Left-3
ad: Left-2
  1. AI3V

    AI3V Ham Member QRZ Page

    Not to mention the terrible imd3 of the 4cx250.

    You will make a LOT of new friends with that tube on HF. :)

  2. KM1H

    KM1H Ham Member QRZ Page

    With GM3SEK's or a similar regulator circuit the 4CX250B can be at least a bit better than most of the SS transceivers.
  3. AI3V

    AI3V Ham Member QRZ Page

    Yeah, I agree.

    They are both pretty bad.

    Run a pair on 20m @1kw into one of your good beams and the splatter will be as loud as my barefoot rig and a dipole!!!

    I turned down the power in my HF rig to 75 watts or so, (2x mrf458 @13.8vdc), every little bit helps

    Last edited: Feb 4, 2016
  4. KM1H

    KM1H Ham Member QRZ Page

    The sweet spot for my TS-940 is about 75-80W which gives a -40 dB or so 3rd and that is not ARRL inflated dB. That runs into an Alpha 76PA 3 holer (3 x 8874's) at only 1200W out.

    The other rig is a TS-950 SD which has to be cranked down anyway as 150W is too much and at 80W it is -44dB. The amp there is a LK-500ZC (2x Eimac 3-500Z) which is then loafing at 1200W also.

    I dont beat my rigs or amps and never had a failure over many years of contesting and DX pileups.

    Ive been able to bring 250B's down to -33 or so at 2000-2200V in AB1, and almost -40 at 1100V at almost Class A in the transverter I built to test 6M conversion amps. It started as a late 50's P&H 150-6 TX converter only. Exciter is a TS-830 running about 10 mw Class A out of the transverter port.
  5. AI3V

    AI3V Ham Member QRZ Page

    How much negative feedback you running off those 250's ?

    Or are you running them allmost class a at 5 watts?

    Iirc the 4cx350 used in the frt84 40kw xmtrs I repaired on Diego Garcia amplified (grid driven) a 250 milliwatt signal to a few tens of watts.

    That was 35 years ago, if anybody has a technical manual laying around I would love to refresh my memory.


    P.s. the final in that 'mitter was a 4cx35000, and the rf output meter went to 60 kw.

    That and a 600' rhombic WILL open 20 meters (so I was told! :) )


    Pps, I think most hams would be surprised at how far you have to back off most "legal" emitters to really make them "clean"
    Last edited: Feb 5, 2016
  6. KM1H

    KM1H Ham Member QRZ Page

    No NFB.

    The transverter can run up to 140W and the 10 mw is amplified before feeding the 250B. The bias and RF drive is adjustable for whatever PO is needed so the IMD degrades a bit at full output. Some conversions only need 10-40W.

    I dont remember what the 50KW final (TMC maybe rings a bell) was at Wheelus AFB in Libya in 60-61 but I remember being at the receiving end many times when deliberate 20M QRM started wiping out the phone patch traffic I and others were running from K0NAB, NAS Olathe and they kicked it on. The rhombic was aimed at DC so we were right in line with its peak. We could do similar with a 32V2 into a modified BC-610 as the normal amp. Or the 610 bypassed and the 32V2 feeding into PP 1500T's with 1000T modulators when needed (surplused US Army amp). Power was about 5KW carrier, 20KW PEP. The antenna was a 6 el Telrex at 90' .
  7. W8NF

    W8NF Ham Member QRZ Page

    Ceramic versus glass is not the issue. Heater/cathode versus instant-on filament is the issue. Two dramatically different types of construction, two different methods of wear-out, two different sets of fragile conditions.

    Any tube with a notable warm-up time has a separate heater and cathode. The cathode is a nickel cylinder with typically barium oxide coated on it, as an electron emitter. The heater contained within it is merely to heat, not to emit electrons. The heater is simple tungsten wire.

    Any tube that's close to instant-on has a directly-heated cathode, usually known as a "filament". It's made of tungsten wire, impregnated with thorium, the "thoriated tungsten" filament, aka "coated filament".

    The indirectly heated structure (3CX800A7, 8877, 8874), despite warm-up time, is the "more elegant" system. For the same amount of cathode emission capability, it consumes approx 1/3 the heating power, and operates at a much cooler temperature, and is physically much smaller. Except for the smallest cathodes (e.g. 4CX250B), the barium is deposited on the nickel cylinder in stripes. Thus, electron emission comes only from the stripes. This allows the grid wires to be vertically oriented and aligned with the gaps between stripes, reducing grid current-the electrons flow through the gaps between the wires toward the positively-charged anode. Further, with such low grid current, they can use materials that can't take much the molybdenum grid wires drastically reduces the possibilities of secondary emission, and when you do get grid current, the electrons striking the grid wires don't create an added avalanche of electrons. The combination of small size, close spacing, well-behaved grid means lower distortion and much higher frequency performance (500MHz for most of these tubes) compared with directly heated variants.

    Another great benefit of the indirectly heated cathode is that the low temperature of the heater means less difference in heater resistance from cold to hot. You never need inrush protection on these heaters. And, they simply don't go through as much temperature cycle from cold to hot and back to risk an open heater. In 2,000+ tubes that went into production amps during my tenure at an MRI amp manufacturer, the only open filament we saw was a DOA tube.

    However! If anode voltage is allowed to drop too low, while anode current is high (a condition at an RF peak, therefore driving cathode voltage is negative, meaning grid is effectively positive), the electrons aren't being pulled so strongly by the anode - it's not so positive - but the grid is, therefore they're attracted more to it. This is the common cause of excess grid current, and those fine wires take tens of millisconds to heat up. If you exceed the power limit of the grid wires, the gold plating will evaporate from the grid and deposit on the cathode, killing emission. This takes less than a second. A grid current trip circuit really should be mandatory.

    Also, remember that warm-up time? What happens if you try to pull cathode current before full warm-up? You can rip barium oxide off the cathode...reducing cathode emission at least, and possibly the remaining particles cause interelectrode shorts. This damage also occurs very fast.

    The lifetime of an oxide coated cathode tube is not at all debatable. Eimac has always maintained 20,000 hours of heat on-time as a proper life expectancy. Note that it's specified as heater on-time, not "cathode current being drawn" time. As long as the heater is on, the barium oxide on the cathode is slowly evaporating, coating other internal parts of the tube. You don't have to be drawing current. Happily, because the heater operates at such low voltage, you are not risking any kind of failure by cycling the heaters on these tubes. If you know you won't be needing it for a while, turn it off. Just know you'll have to suffer the next power-on cycle.

    So - for an indirectly heated tube, 20,000 hours, unless you pull cathode current before it's warmed up or overheat the grid.

    When I was in the business, our main MRI customer returned all old tubes back to us, with a record of heater on-time. We kept a half dozen amps powered up in the burn room, putting more hours onto those heaters. We never had one fail before 21,000 hours, more typical was 23,000 and we had some go 25,000. I still don't understand why the failure isn't gradual. We'd run the amp with heater, HV and cutoff bias, and send it through a tune cycle once a week and check gain. Saw no dropoff in gain until the end, then it was fairly swift.

    By the way, all this info is in Eimac's "Care and Feeding of Power Grid Tubes", a relatively fun two-afternoon read. I think it's 9 PDF parts, available from Eimac's website.

    The directly-heated cathode is a different bird. It operates at a much higher temperature, so high that each time you turn it on, you send it through a "creep" temperature range defined by Larson-Miller effect. I'm not a materials scientist, so I apologize to those in the audience for my simplification here.

    As the temperature of the thoriated-tungsted filament goes through this Larson-Miller point, a slight crystalline re-alignment takes place in the metal. When spun in filamentary form, tungsten's response is to grow slightly longer and decrease in width. It does not help to "soft start" the filament - it passes through this temperature, whether fast or slow - and, whether turning on or off. In the 572B tube, the M-shaped thoriated tungsten filament is supported by fine spring wires, which hold it in tension. As the tungsten stretches, the spring wires maintain filament shape, and this can go on a long time before a filament ends up open-circuited. In the 3-500Z, however, the filament consists of counter-wound spirals, mechanically fixed at both ends. So, as the tungsten in the 3-500Z grows, the twin spirals have no way to length, and they eventually flop to the side, creating a filament-to-grid short. Despite decades of claims that this is caused by parasitic oscillations, it doesn't explain why the same symptom occurs in SE triode audio amplifiers operating with wide-band resistive loads and only 800V on the anodes. The "valve audio" enthusiasts all know not to cycle power on a 3-500Z!

    I do not recall Eimac's book indicating a lifetime for a thoriated tungsten filament structure. Many industry references claim 30,000 hours, as long as cathode current versus pulse duty cycle are observed.

    I've worked for three companies who were Eimac OEM customers, so I processed a warranty claim or two. Std warranty is 2500 hours for most transmitting tubes (not just Eimac). Remember, warranty is not life expectancy - they're only long enough so that you can shake out any potential manufacturing defects.

    A directly-heated tube is amenable to rejuvenation, in which you'll get a few more dozens of hours. If the thorium has been depleted, then raising the filament voltage 5-10% will bring more of it to the surface, restoring emissions...but at this filament voltage, the remaining emissions will deplete quickly. Eimac wrote an app note about filament management that said this is a good practice to do...don't keep standby spares at all. Wait for emissions to decline, then raise the filament voltage and then, order your new tubes.

    My thoughts on the 3CX800A7: go for it. It's still used in currently-manufactured commercial equipment, such is not true for the 3-500Z, or any other glass tube. And, having such a high frequency capability, means that the phase shift inside the HF bands is so low as to make oscillations almost impossible to create, as long as you really have the grid grounded properly.

    I'm not a fan of the 3-500Z...seen too many amps damaged by filament-grid shorts. It's an inevitable failure on that tube, except in the old Henry amps that pulled so much cathode current that they depleted tubes before there'd been many on/off cycles.

    20,000 hours of heater on-time is only 2 1/2 years if you leave the amp on 24 X 7. But you probably didn't plan to do that, did you?


    Dave W8NF
    Last edited: Feb 5, 2016
    WA7PRC and KA9JLM like this.
  8. KK4YDR

    KK4YDR Ham Member QRZ Page

    Now that is the kind of write up that I was looking for. Fantastic amount of information. Your energy in writing that will not go wasted as I am not the only one whom will benefit from this thread.

    In reference to grid protection circuits, I was really looking at getting an AL-800H by Ameritron which have all the protective features built in. It does have a grid protection circuit apparently built for the 800 tubes as they are delicate, which you further reiterated through your post. I know there are other amplifiers out there, but ones using the 800 tubes are harder to come by and I unfortunately do not have the necessary knowledge to fully home brew an amplifier at this time in the hobby. However, I can buy one that someone else spent big bucks on the Masters and PhD in EE to build and sell.
  9. KM1H

    KM1H Ham Member QRZ Page

    What tubes were these?

    Yep, sure:rolleyes:

    Dont expect that level with the AL-800.

    Id offer that the "designer" didnt have a degree but when you have copied others ideas for decades a few are bound to fly well. Consider that any EE really worth his paper would want to go with a low paying hobby company and you start to understand why commercial and military gear companies dont have a fraction of the problems.
  10. KA9JLM

    KA9JLM Ham Member QRZ Page

    Don't fall for the new one in the box, with no picture of the amp for sale on the internet.

    Even if the receipt for the box is included.

Share This Page