Change Out 3-500Z's for GS-35b's?

[W9GT]

Anyone have experience replacing 3-500Z triodes with Russian GS-35b’s?  I realize that the GS-35b’s are much more robust tube…probably somewhat similar to 8877s s in their capabilities.  3-500Z tubes, especially Eimacs are becoming more and more expensive and GS-35b’s are more available and less expensive.  Therefore, it seems like it might be a good idea to change tubes.  

I have an existing amplifier that I built over twenty five years ago, that uses a pair of 3-500Zs.  It works great…only replaced the tubes once….but the tubes are again becoming a little soft and the output has dropped off, especially on the higher freq bands.  I have been toying with putting in a single GS-35b, or even a pair to replace the 3-500Zs.  One GS-35b would be enough, but since amp is a two-holer…might use a pair.

I would have to replace the filament transformer and I'm just toying with the idea and considering how much tinkering I might have to do with the tank circuit and input network to make it work with those tubes.    I just wondered if anyone else might have done something similar.  I have plenty of power supply…3500 v @ an amp or more…adjustable with a variac.  

73,  Jack, W9GT

[K1JJ]

Hi Jack,

A noble quest, indeed.


Normally:

Well, depending on one or two tubes added and more power input, the tank impedance will need to be lower by making a smaller L and larger C for each band.  You could use the same E/I parameters and the same tank impedance if desired too.

Can the exisitng antenna relay handle more power?

The exisiting caps (C1 and C2) may arc if they are air and not vac variables.

The HV power supply will need to be more robust and the fil xfmr will need modification, as you said.

Even the metering may need mods since you will likely pin the plate meter now...

Tube socket changes with chimneys added, of course. Blower system added to mate to the chimneys.


But since you already have a big HV supply and the amp is homebrew, it may be do-able with one tube more easily.


On the surface it sounds like a cool project, but in reality, I'll bet it would be easier and a better product if you started from scratch and built a new amp designed around TWO GS-35b's.




T

[W9GT]

Hi Tom,

I appreciate the input.  Well, most of the components in the tank are already overkill, but some changes might be in order.  Might go to a vacuum variable for plate tuning.  It probably makes more sense to just put in one tube.  No socket is really necessary, as the tube can be mounted to the chassis in GG configuration.  I could cut out a bigger hole in the chassis where the two 3-500Z sockets are located and make a cover plate with one hole for a GS-35b. Chimney can be hi-temp silicon rubber stuff that I already have...blower would need to be added, but I have one that would work.  It just bothers me that new 3-500Zs are over $200 each even for Chinese clones and I can almost get two GS-35bs for that!  3-500Zs made a lot of sense when they were $95 a piece, not so much anymore.

73,  Jack, W9GT

[K1JJ]

Jack,

Yes, just the cost and availablility of the 3-500Z's is a good reason to upgrade the amp.  Since you have a rugged infrastructure already in place, might as well go with two tubes and make it worthwhile. Sounds like you can work out the details with no problem.  If it was a commercial 3-500Z amp, it may be a problem, but since it's already homebrew, you've got lots of options.

Post some pics when you may progress!

T

[Steve - WB3HUZ]

If you already have a good pair of 3-500s, they should last you for years. I wouldn't bother doing the change over unless your 3-500s have died.

That said, if you run the new GS-35 plate voltage and plate current at the same levels as the 3-500s, why would you need to change anything (excepting the obvious like socket, fil tranny, etc)?

BTW, I have a GS-35 socket if you are interested.

[W9GT]

Yeah, I agree Steve.  That is what I was thinking.  If I just ran the GS-35(s) at similar voltage and current, the effective plate Z would be the same.  Of course, differences in tube grid and plate capacitance would have some impact and I haven't researched that thoroughly.

Unfortunately, my existing pair of 3-500Zs are showing signs of deterioration, and output on higher bands has dropped off a bit.  That is why I'm thinking about replacing them.

73,  Jack, W9GT

[W2PFY]

How would 4-400's work in there in GG? Same socket, I think the filament requirements are the same.

[WA1GFZ]

Jack,
How about a pair of 4-400s? The 4-400s have a tad less gain and will be more stable. I helped a friend convert a pair of 4-400 to 3-500s years ago and we had to chase some problems to make the switch to triodes.

[W9GT]

Yes, 4-400s are OK, but not as good as 3-500Zs in GG.  I have used them and got about 300-400 watts less output.  They work great for class C though!  Good for AM rigs, just not so hot for GG linears.

73,  Jack, W9GT

[WA1GFZ]

sounds like you are drive limited Jack.

[WA1GFZ]

How about 3cx800s? I see them on ebay pretty often as MRI pulls.

[W9GT]

Yes those are nice, but in demand due to their use in many commercial amplifiers.  Same is true for GU-74b/4CX800As.  Those used to be commonly available on eBay for $35.00, now more like $200.00 plus.  I was going to build an amplifier using three of them, but decided to go another route.

I also have a nice amplifier that I built even longer ago (late 70's) that uses a 4CX1000A or 4CX1500B.  That amplifier is heavy duty all the way, with vacuum variables, a big rotary inductor and super heavy duty power supply.  I have been working to modernize it a bit by adding the G3SEK "tetrode boards" for control and screen regulation, etc.  Tetrodes will really crank with very little drive, but power supply and control is a headache...certainly not as simple as triodes in GG.

My thinking with replacement of the 3-500Zs is just based on using cheaper, more available tubes.

73,  Jack, W9GT

[WA1GFZ]

Man I must have seen a dozen 4CX1000s and 4CX1500 past couple flea markets. But then you could get HUZ's socket and drop in 1 gs35 into the single tube amp. I learned the hard way about large tubes and bread slicers.

[W2VW]

There might be a few people here who have actually been on the air with a GS-35B or two.

I know Stu has an amp working. I replaced an 8877 with one about 6 years ago. Still have the Harris AM-7224 here.

You seem to have the main problem covered. The GS-35B will need in the mid 3kv range to work with available drive from the typical slopbucket sized rig.

One tube will play into the same tank as your pair of 3-500Zs without any changes. A little more plate voltage and plate input power can increase. All the better if your variac is cranked down a little to get the present 3300 volts. Regulation will be better near variac's 100%

One GS-35B sounds like a winner as long as the air plumbing and noise isn't a project killer. Gain will be slightly less than the glass tubes and you will need more bias. A pair of 3-500Zs at 3.3 kv is at least a 16 dB amplifier when biased in class B. Expect more like 13 dB gain with the communist valve.

A pair would require pretty much everything to change to take advantage of the second tube. Frankly, I don't understand why anyone would design a QRO amp with a pair of tubes these days. There are too many good bargain tubes with handles out there.

[W9GT]

16 dB??? that seems a bit optimistic!
Well, with a new set of 3-500Zs and 100 Watts of drive, the most I ever saw was 1500-1600 watts out (to a dummy load).  This was on 80 and 160M.  A little less, maybe 1400 watts or so, on 10M.  

I have heard guys talking about considerably more than that with the GS-35b, on the order of 2K to 2.5 K out with 100 W of drive.  Don't know if those claims are true.  Of course, you can get that kind of power fairly easily with a tetrode (grid driven), 'cause they will really crank with only a few watts of drive....lots-o-gain available.

Yeah, tubes with handles are cool!  Only problem is, you might invest a lot in an amplifier, or even in just modifying an existing amp for one of those "pulls" and find out its no good.  

Seems like the single GS-35b might be a nifty way to go...just wish I had a two hundred watt exciter, rather than only 100W  .

73,  Jack, W9GT

[W2VW]

16 dB??? that seems a bit optimistic!
Well, with a new set of 3-500Zs and 100 Watts of drive, the most I ever saw was 1500-1600 watts out (to a dummy load).  This was on 80 and 160M.  A little less, maybe 1400 watts or so, on 10M.  

New 3-500Zs made 1400 watts on 10 meters with 110 watts drive carrier conditions in an SB-220 last time I checked. If the meter was correct the B+ was 2600 volts under load. This is of course not what the amplifier is designed to do.

Maybe the 3.3 kv is not under load when measured or the reading is off some.

Just my experience. Others who have had this stuff on the air might chime in.

[KM1H]

About 12dB is the theoretical max for a 3-500Z in RF service and I certainly wouldnt want to see a SB-220 running 1400W on 10 either. Besides the math doesnt work either and then add in the input circuit and tank inefficiencies. The most Ive seen is ~1200W using an accurate instrument, not a Bird. Now in a monoband CB amp its possible for a few minutes

A 4-400 has slightly less gain than a 3-500 in GG, usually about 100W per tube for the same drive; a little bit less gain at below 2500V and also pulls may not be up to snuff. The PL-175A is right up with the 3-500Z. Ive cranked 1600W out of a NOS pair in a LK-500 with the external PAC-5 on 6M; you would be surprised how much power is lost in a typical bandswitching HF amp.

Carl
I dont know of any commercial amp besides an old Henry where a GS-35B would be able to function well. All the others are limited in at least one area.

[Steve - WB3HUZ]

A pair of 3-500s with 4kV will make 2 kW on the low bands. I've done it. Those tubes, like most every other Eimac triode become more efficient/have more gain with higher plate voltage.

A pair with 3 kV should be good for the legal limit.

Unless you have more than a 100 watt driver, a pair of GS-35s doesn't seem to make much sense.

[WA1GFZ]

200 watts should drive a pair of 4-400s to 1500 watts.

[KM1H]

The LK-500 can squeak out 1500W with the external PAC-5 but thats it. The PT-2500A and AL-82 can both do it with the higher B+ but they also are famous for IED events. It takes a set of brass nuts to run them at 4KV

200 watts should drive a pair of 4-400s to 1500 watts.

Maybe with some almost duds, they also wake up with more HV.

[Steve - WB3HUZ]

Why? The Eimac sheet rates them to 4 kV.   

But there's probably not much difference in the available power when running at 3500 volts and a little more margin for 'error.'

[KM1H]

Cuz they outgas more at that voltage and start arcing and have to be reconditioned or discarded. Plus the Chinese ones often will not high-pot much beyond B+ much less at AM potentials. A tube arc at 4KV can do an awful lot of damage to itself and attached components if there is no or poor protection circuitry.

Nobody is going to hear the difference between 1500 and 1200W and I get tubes that have run 24 years of contesting and DXing. I also reduced the filament voltage to 4.8V; they were at 5.2 when I first tested the amp.