Print

[K0ZN]

Hi.  I have an Alpha 91B I bought new in, I believe, 1996. I picked the amp up in Longmont at the factory and also bought a spare set of GU-74B's at the time. It is, and has been, a wonderful amplifier which I have used moderately over the 25 odd years with the same FOUR tubes without problem.

I can't tell you what you "should" do, but I can tell you what I have done, which, apparently has worked satisfactorily. I swap out the tube pairs about every two YEARS....run them for two years, then swap, etc. When I put the tubes that have been in storage back into the service, I simply turn it on IN "STANDBY" and let it run all night and part of the next day.  SO FAR.... re-start of the previously in-storage tubes has been a non-event. Again, this may not be the "best" program, but it has worked OK for me for 25 years.   "Your mileage may vary."

FYI: when you swap/install tubes be VERY careful that you get the grounding ring set correctly.
Pay close attention to how things look before you remove the tube. Not difficult, but quite important.

Side note about the tubes: two of them cosmetically look like they came out of a Russian tank (I have heard the GU-74B was used in tanks) that was destroyed in the battle of Lenningrad. (minor scratches, quite a few bent cooling fins, etc.) These were the "OEM" tubes....but, again, they have worked perfectly for 25 odd years. The "spare" tubes I bought new came in sealed Svetlana boxes and looked perfect; very pretty tubes (if you have a tube fettish!). Both pairs put out the same amount of power with the same control settings....meaning the internals must be pretty consistent.

Good luck with your amp.

   73,  K0ZN

P.S.  I would recommend operating glass envelope tubes at least every 90 days in whatever condition it takes to cause gettering.... again:
        "Your results may vary....."

[VE7RF]

I find the last comment about the GU74B tube run as a 600 wat tube very interesting it also reminds me of how many AL80 series amps (al80,al80a,al80b) heathkit SB1000, QRO1000 plus I may have left out a few more amps that use a 3-500Z tube which may be rated for 500 watts plate dissipation if properly cooled with a pressurized chamber with chimney. The Acom 1000 has been in production since 1998 or there about, I had first seen the amp at Dayton were Krassi was showing the amp opened up and he said to me "hey look its got a a KW on Six" and my thoughts were wow how to get to know all my neighbors.
These same tubes were designed in before Acom's birth the Alpha 91B  and many contesters bought these amps, many ran them at 2Kw or better all weekend long then Acom brought there A2000 out and many ran the pants off at 3Kw out but at that level they might have had a short life on those tubes but at around 2Kw they did fine and lasted a long time. You may not like Russian tubes at this time but those tubes are allot more rugged than some will make them out to be plus last allot better than 3CX800A7 at and around the 2Kw output level. I have seen a lot more Eimac 800 go south before the GU74B's.
If you run the Acom 1000 around 900W to maybe 1000W pep I see no real issue and on CW right around 800 cause the extra 200W will do nothing on the other end. I have seen some push the amp on You Tube vids for much higher output but at 1200 to 1400 wats you might be taxing the power supply a bit plus generating excessive heat for which the cooling system may not be properly designed. Driven between 60 to 70 watts should give you proper output plus your exciter is not really being pushed at that point but if you kick it in the pants at 100W that is not really needed unless the tube is going south.

The GU-74B has 600 watts of anode dissipation, NOT 800 watts.  Eimac never made a ...'4CX-800'...doesn't exist...never did.  Look at the oem russian spec sheet...  600 watts anode diss. 

[VE7RF]

Operating a single  3-500Z  at 550 ma, like in the AL-80B....  is pure looney tunes.

But apparently, that is the new fangled...'ICAS' rating for tubes.

I know why they did it though..to get the plate load Z down to a reasonable level, where a PI net can be constructed for all bands.  ( 2777 ohms, using 550 ma and 2700 vdc loaded)

If instead, you ran the single 3-500Z at say  4 kv loaded at 400 ma..to get the same 1 kw output, the plate load Z would be through the roof at  5555 ohms.... which would require huge values for the coil ..and real small values for the tune and load caps.  You would not be able to find a cap with a low enough min C on the upper HF bands.

[KD6VXI]

The Eimac 3cx800 is an ultra linear tube.  Designed for superb IMD, as it was designed for and used in MRI applications.

I know guys whom have a couple decades on their 800s. Run at 1 to 1.2kw out and properly loaded the 800 is capable of IMD below 45 dB.

You'll never get that out of a Gu74.

Sure, guys running 3 tube commanders at 5500 out aren't going to see that lifespan, but that's their choice.

Conparing the Gu74 to an 800A7 is like comparing a Ford to a pineapple.

I've built around both tubes.  The eimacs last a lot longer in the hands of a competent operator.

--Shane
WP2ASS / ex KD6VXI

[KX2T]

well if any of you out there wanna run FT8 on any of these tubes Eimac or the Russian tubes neither of them I would run at 800 watts output, maybe more like 500 would be far better and the 3-500Z I would run at maybe tops 400 watts but on CW and SSB there ICAS ratings would be fine. Since these new digital modes came out they are very much like the RTTY mode of years past, most seasoned hams did not push there  tubes at stupid hi power levels with that mode.
BT check some of the amps that went threw the ARRL labs using either the 3CX800A7 or the GU74B's tubes and you will find that both of these tubes will produce a very clean IMD signal which tells me its more the input signal going in. I have never seen my Acom 1000 flat top at all since 2012, never had any complains but here again that depend on the operator running these amps and how they load them then the tubes themselves, you push or don't load any amp improperly you can get terrible IMD out of any tube.
The 3CPX800A7's run in MRI's were a pair used at 1kw output and in most cases run for 600 hours then swapped, the larger tubes like the 3CPX1500A7's and there YU counterparts were run at higher power levels and then the 3CPX5000's were the big boys in the chain but mostly today these MRI's are solid state and the tubes are becoming extinct.

[KD6VXI]

A pair of 4-400s makes an easy kw carrier and over 4kw pep, plate modulated.  500Zs have no problem at that level, either.  Granted, they will lose emissions faster than when run at lower levels, but they will do it.

It's the shabby colling in ham amplifiers that limit what you can run digital.

The YC156 was an amazing tube.  Last I checked Reid Brandon still had a NOS one, sans PRM nuts for mounting.

I often thought of snatching it uo, but if you ever need another tube, they are almost unobtainium now.

ACOM gets away with the GU74 and decent IMD because the tube is biased deep into conduction.  Which is another reason the tubes don't last as long.

The 3cx800 was used in broadcast as well, FM band.  The transmitter escapes me right now, maybe someone else can chime in.

Ham amps have subpar cooling.  That's their limiting factor.  Along with other reasons, but they just don't have the cooling for high duty cycle modes.  Just ask the guys that burn up commercial 8877 amps in contesting.

My Harris has a 3-1000z tube.  I can run a kilowatt (have pushed it to 1700 before, 3kva supply in it) for days.  Any mode, any length of time, 24x7 rated, all day.   The State Department demanded it, as they where used for RTTY in embassies.

--Shane
WP2ASS / ex KD6VXI

[VE7RF]

A pair of 4-400s makes an easy kw carrier and over 4kw pep, plate modulated.  500Zs have no problem at that level, either.  Granted, they will lose emissions faster than when run at lower levels, but they will do it.

It's the shabby cooling in ham amplifiers that limit what you can run digital.

The YC156 was an amazing tube.  Last I checked Reid Brandon still had a NOS one, sans PRM nuts for mounting.

I often thought of snatching it uo, but if you ever need another tube, they are almost unobtainium now.

ACOM gets away with the GU74 and decent IMD because the tube is biased deep into conduction.  Which is another reason the tubes don't last as long.

The 3cx800 was used in broadcast as well, FM band.  The transmitter escapes me right now, maybe someone else can chime in.

Ham amps have subpar cooling.  That's their limiting factor.  Along with other reasons, but they just don't have the cooling for high duty cycle modes.  Just ask the guys that burn up commercial 8877 amps in contesting.

My Harris has a 3-1000z tube.  I can run a kilowatt (have pushed it to 1700 before, 3kva supply in it) for days.  Any mode, any length of time, 24x7 rated, all day.   The State Department demanded it, as they where used for RTTY in embassies.

--Shane
WP2ASS / ex KD6VXI

For proper biasing, the GU-74B should be idling at 225 ma...PER tube.   That's 450 ma for a 2 x holer..and a whopping 675 ma for a 3 x holer.  This is, if you want 1/2 decent IMD.   Typ they are run at less than that.

Eimac made  FM broadcast cavities for both the 3CX-800SA7 ( 750 watt  CCS 24/7)...and also the 2 x hole version  (1.5 kw  CCS).

As far as cooling goes, toss in the myriad of under rated components into the mix.   Everything from tank coils, insulators, interconnecting wires, and grossly under rated bandswitches....and also roller coil contacts..and also relay contacts.

Bandswitches are rated at 60 hz / DC....NOT RF.    They have to be  de-rated for RF.  You also have to factor in ......

Mode
Duty cycle
Ambient temp
Temp rise over ambient
Frequency.

Going from  160m...up to 10m  is a 16:1 ratio.   Sq root of 16 = 4.   A tank coil on 10m will dissipate quadruple the amount of heat  vs the same coil stock wire ga / tubing OD  on 160M.  You can do similar calcs for other freq combo's.

Best way to deal with heat is..... don't generate it in the 1st place..or as much of it.

When using FT-8 /  RTTY / FM...duty cycle is high...... vs  SSB / CW.   On  SSB, average plate current is only 1/2 of   FT-8 / RTTY / FM .  Heat dumped into the coils, bandswitch etc, is quadruple when using FT-8 / RTTY / FM.    Duty cycle on the B+ supply is  double.    Anode diss is higher.

Cracked ceramic bandswitches on Ameritron amps...when using  FT-8  is all too common.

[W4MSL]

Just learned that Array Solutions provides testing for GU-74B tubes for $35.

https://www.arraysolutions.com/index.php?route=product/product&product_id=14890

Seems like pretty comprehensive testing, and I'll probably send my NOS GU-74B to them, if it ever arrives from Ukraine. I'd rather place a known-good tube into what I hope to be long-term storage than one about which there are questions. Someone (not here) suggested not testing/conditioning a GU-74B until actually ready to use it. That might be sound advice, but is not so intuitively. I've also been told that these tubes contain no getter, so the term gettering might not be accurately applied to the GU-74B, if I understand that correctly.

[KD6VXI]

I don't know about 'no getter'.

They are gettered by filament heat alone.  They don't have to be run like a 500Z, 400A, 1000, etc where the tube is gettered by showing color.

24 hrs with nothing but fils voltage and you're good to go in any commie ceramic tube.

They have gettering.  If you don't think so, throw a commie tube that had been sitting for 30 years in an amp with all voltages on it with zero filament time.

--Shane
WP2ASS / ex KD6VXI