The AM Forum

Hej !

Since quite some time I have an unused Eimac 3-1000Z waiting...
The idea is to build a PA for shortwave, I have seen some descriptions in old QST Magazines from the 60's.

So I'm looking for advice/experience from other Hams that may have some ideas!

Bill Orr's radio handbook has a two stage amplifier, a 4CX250 driving a 3-1000Z.

Any 4-1000 GG amplifier schematic can be adapted to the 3-1000Z.

I have a commercial 3-1000Z amplifier I just put a 4-1000Z tube in, works fb, but drive went up!

The 3-1000Z is like an 8877 / 3CX1500.. Not a WHOLE lot of drive for a WHOLE lot of Pout!


--Shane

There was a ARRL QST/HB article also back in the 70's I believe. Its a great tube but dont abuse it as they are very expensive to replace.

Carl
KM1H

Hej !

I have the Radio Handbook where the 4CX-250B driving the final 3-1000Z Amp is described, so I have some good ideas to use.

Then, to a possible "problem", I have a new socket for the 4-1000A, made of cast aluminium. (can't remember the  "SK" part number right now) Will a socket like this work? The pins will fit, but I'm afraid of the mechanical stess due to thermal expansion, I do not want to get any cracks in the glass base of the tube. The chimney is for a 4-1000A, hence it's a bit too tall, compared to the chimney designed for the 3-1000Z. Is this a problem?

Finding a new chimney and to have it shipped to Sweden may be a bit too exiting since glass have a nasty habit of breaking. However, the life of the tube is very important, if I need a shorter chimney I will have to find it.


 
Shane, could you please give me an idea of "operating conditions" when you had the triode in your amplifier? Plate voltage, Ip and so on!
Any "special design features"?   

Over here, they used to cut the bottoms off of large pickle jars and use them. What ever works.

The 3-1000Z is a zero bias tube at 3Kv on the plate.  That's how mine is run.

Their is not Iplate meter, so I can't tell you.  It runs a choke filter, and has EXCELLENT REGULATION.  3300 key up, 3150 key down at a kilowatt carrier (I run 8 gauge primary wire to the panel)

As someone else has said, the 3-1000 is PRICEY..  The 4-1000 can be subb'ed. 

Asking me about my amplifier is like asking a CBer about theirs (no pun intended).  Mine is switched cap on the in and output of the pI, and a roller inductor for tuning.

Any 4-1000 or 3CX1200 schematic should work.  The 1200 is the "solid state" version of the 1000Z and kinda the 1000A.

--Shane

The pins should float in an air system socket. It is a problem if they are tight against the insulator for thermal stress. It is not great for RF service due to long pins. Run wide copper flashing between grid pins and case ground. Also drill some holes in the sides of the air system socket to get more air flow. It takes a lot of air pressure to just rely on the hose connection. Maybe you can find a better RF socket for the tube.
The 4-1000A only had 4 1/2 inch holes to move air across the pins so it took a lot more pressure. The 3-1000Z didn't have the air restriction of the base making it have less restriction but also a weaker structure.
A fine amplifier tube. I think there was a 3-1000Z linear in the early 70s handbook or maybe the SSB handbook. It was only 1 stage.

This might be of interest

http://www.google.com/imgres?imgurl=http://members.cox.net/nx7u/images/LK2000_original.jpg&imgrefurl=http://members.cox.net/nx7u/BTI.html&h=300&w=400&sz=56&tbnid=vfOsU0LiZgI0QM:&tbnh=93&tbnw=124&prev=/images%3Fq%3DLK-2000&hl=en&usg=__ydhnJNU06QBU0yS6MjdmRseTa24=&ei=LdKwS8iqLIX7lwfHt9SRAQ&sa=X&oi=image_result&resnum=7&ct=image&ved=0CBcQ9QEwBg

BTI switched from a glass chimney to that brass duct seen in the photos. It directs air across the plate cap.

Kerri KC2UFU and Ken KC2UDZ are running one of these now. I listen to them all the time.

Dave wins the prize for the longest link ever posted on AMFone  ;D

That cast aluminum socket is a slug and seriously restricts air flow. The later plastic one is much better.

Many moons ago I made a chimney for a 4-400A out of a 2# peanut butter jar, it lasted about 2 months before fracturing. They are not RF and heat rated.

A big Coleman style lantern chimney may work.

Carl
KM1H

It's always nice to be recognized for something  :D

teh tnx.

Carl, what is ur opinion of eimac SK500 socket for 4x1?

tnx
rob

The aluminum is a slug, the plastic is fine.  I dont know model # without going downstairs and Im too lazy right now. ;D The aluminum can be modified as well as cutting out alternate sections in the tube base to reduce back pressure when using the aluminum one.

Carl
KM1H

Okay tnx i'll look for plastic. you just saved me a few shekels

R.

When I built my amplifier I wanted to use a 3-1000 but had second thoughts when I found out each tube is $1100 new.  That's awfully expensive when you're beating the p*ss out of them during the testing stage.  I decided to go with a pair of 3-500's and it works just as well.  I ended up with 1200 watts.

That's NOS too, not really new.   Don't know how long they have not been in production but yes, when planning an amp from scratch or halfway from that if you have a choice, 3x1 not the best way to go.  It is a great tube, but risky.  You gamble buying used; and vendor/auction prices are sky high.  If I had a pair 3x1 amp already, and a good set I'd stick with them because if you treat them right they'll last and really put out but I don't.  I considered quad 3-500 then decided that was too iffy and too much work, next came 3cx3K for contemplation but then a hb gg 4x1 project sort of fell into my lap, one with enough oomph to make it a pair 4x1, so that sort of decided the issue for me  ;D

You could always pick up a 4-1000a and build a rig. once you get it working right install the 3-1000. A good tube will last a long time.

4x1 is a good, rugged, hard to kill tube. The one in my big homebrew rig was used when I built it, and I've beat the pants off of it for 10 years now. Still using the same tube!!

And it sure looks purty with the plate glowing red!!

Have you done this before, Frank?

My Harris wants about 200 watts of drive with the 4-1000 in it, to produce legal limit.

Before, approximately 100 watts of drive pushed it to legal limit, I'm told. 

The difference is a 3-1000 vs 4-1000 at app 3Kv.

--Shane

I ran a 4-1000A GG for years and today it sits in my buddy's shack about 5 miles east. 160 through 10.
I could have dropped a 3-1000Z right in because I had a string of Zeners for variable bias. Resting current will go up.

A good 3-1000 cathode driven with 3200 volts on it under load will make legal limit with 60-80 watts under carrier conditions.

A good 4-1000 in the same exact situation with all the grids strapped will need 110 - 130 watts drive for 1500 watts carrier.

That's what I measured in the 4-1000A
So if you have a clean layout the 3-1000Z will give you a bit more gain.

The 4-1000 in GG likes plate voltage - 5 to 6 kV makes it much easier to drive than at 3 kV. Max rating per Eimac is 6 kV. A small globar resistor should be used in the plate parasitic suppressor at this voltage. The plastic socket is much better than the older cast aluminum one.

That's for sure.

No reason the 3-1000 can't be run at that voltage too with zener bias.

I put 5500 on my amp once. Made real power but blower wasn't rated for the air flow required. I ran a pair of BC610 transformers across a 240 volt line.
It ran just over 3000 volts and makes 1500 out with just over 100 watts drive.

Back in my 4x1 building days I used one or two in either grid drive or GG, depending on the customer requirements.

The PS started at 4200V but soon went to 5200VDC key down as the standard. A local company wound the iron.

With sufficient drive the output was 2500W per tube at least thru 20M with good pulls from local stations.

Carl
KM1H

Hej !

Thank you all for all comments!

So, to some conclusions(?) regarding the socket.
I have a new SK-500 (this is the aluminium socket).

1) I understand by looking at it, and this is also in line with comments in the thread, it will cause high backpressure for the cooling fan. Possibly this can be improved by making some extra holes in the socket.

2) Is there a possible problem for the 3-1000Z to use a SK-500? Will it cause mechanical stress for the pin to glass seal. I'm not sure that I fully understood the previous comments about that.

3) Or should I start looking for the plastic SK-510 socket?

4) Any suggestions/comments regarding Power Supply? 

As you all can understand, I want to get things at correct as possible from the beginning, things can go wrong in a big way with a high power amplifier...

After you've read all the notes and if your still wanting to use a 3x1 then the aluminum SK500 socket should be fine at the 3x1's 'lower' plate voltages and "somewhat lesser" cooling restrictions.  I thnk it'd be elegant to use the 3x1; who else will ever use them for something more than a shelf queen in a (gasp) collection  ;D?

I have a SK500 in hand and just want you to know that the pins can be aligned accurately simply by very slightly enlarging 4 out of 5 holes in the base, loosely installing the pin connecting hardware, inserting the tube and then tightening the nuts to the new spacing requirements.  This will set up the cold alignment.

Since your very slightly flexible GG copper strap and braid will "nail down" permanent spacing between three of these pins permanently to the rim of the socket and float as a unit that satisfies the bulk of thermnal stress for those three pins. Double nut the GG connections, leaving the three holes slightly loose to the base phenolic if you want 'maximum' flexability.

The other two pins for filament can be left floating as far as the sockets base phenolic goes as long as the tube pins are tight to the clamps and the double nutted hardware from pins to filament supply leads are tight.   

Sounds complicated but simple in practice.
Also even if everything is stock tight, keep in mind that the phenolic in the socket insulator also expands radially with increased heat.  Don't have the exact figures but I'll bet it's not too different than pyrex glass.

My 4-1000 amp had 6200 volts on the plate under load.  Big PDahl Iron.  With a 100watt SSB radio on 80 meters if I wispered it would SLAP the Bird 2500 slug HARD in the corner.  I have no idea how much that amp actualy made.  I could run it at 800 to 900 watts AM Carrier it it did not seem to care. The blower was VERY loud and the room would shoot up 30 degrees in 10 minutes of use. I sold it off. 

It was a 4PR-1000 tube though.

C

Hej !

OK, I now have the situation under control, checked the procedure "align socket with pins", everything is now clear.  :)

My plan is to run the tube at "quite low" power, or in other words following the Eimac Data Sheet for linear SSB service.

Here in Sweden we have an advatage when it comes to feeding the power supply, our mains voltage (that you will find in every room) is 230 V / 10 A (50 Hz) or possibly in newer installations 16 A.
Every (almost) house is feed with three phase 400/230 V 16 A at least. To find 400/230 V 20 or 25 A is not uncommon, and even 32 A! Actually my apartment is feed with 400/230 V 16 A. So, for QRO operators, power usually is not a problem.   

The metal socket was made for use with a high pressure blower. An airflow switch should be used with the system. The metal socket is perfectly OK if you have a good enough blower. "No" fans or low speed blowers, but something like a 3-4" diameter squirrel cage, 2-3" wide, running at 3450 RPM or more. The higher the rotational speed of the blower the better it will serve if you have the metal socket. The big xmtr here has a 4-1000A and the metal socket, and a good size 3750? RPM blower. It can run for hours with rated dissipation and no problems. The construction is such that the "hose inlet" of the socket is open to a pressurized compartment. I cant' stand it when people, and even manufacturers, use fans for cooling a tube like that. I don't know how much 3-1000's cost, but they are still made by "Amperex" IIRC - I have some spares for RF generators used in semiconductor manufacturing.

If 3-1000Z cost is a factor, you might be able to trade what you have for a set of 3-500Z hardware and the tubes will be more like $150 each new Chinese. I use them in a modulator and in 40+ hours of tests, never have had a failure.

Hej !

Thank's for the input! A pressure sensor connected to the socket has been included in the plan, actually it is a MAP (Manifold Absolute Sensor) from an automotive application. 

Since I already have the Eimac 3-1000Z I will keep it, I did not pay that much for it, and it is new. However, my concern is that IF it breaks it will not be pleasant to find a replacement.

ME!!!  I'd LOVE to find one that I could trade for or something to put in my Harris.  The 3x1 is about vaporized now.

I also have a 3CX1000 I'd stuff in if I could find a bayonet socket for it that cost less than my Harris :)


--Shane

The 3-1000Z can be replaced by a 4-1000Z.  Just make sure you GROUND pins 2,3 and 4.

It's a seckzi tube....  SOMEDAY I'll find one to return my amp to stock.  I LOVE the drive requirements.  I'd also consider doing grid driven on my 4-1000 that's in the 3-1000 RF deck now.

--Shane

How about modifying the erector set Chinese 4CX-1000A socket? Start by removing one layer and see what needs to be changed. The original 3CX1000A7 socket was for grid drive and I modified mine for GG.

Carl
KM1H

I found the 4CX1000 socked is a source of spare parts for the 4CX3000A socket so I bet it was a standard sice contact.
6200 on a 4-1000A Oh MY...

From my experience running the 4-1000 at 6 kV pay attention to your plate parasitic suppressor. Mine was made of about two or three windings over 3 paralleled two watt carbon resistors, I forget the value but I guess 50 or 100 ohms each. (As in the published Eimac design intended for lower B+). When they failed - obviously by overheating - the results were dramatic to say the least! I was told (by Rich Measures I think) that the tube has a tendency for a VHF parasitic. I guess the elevated B+ increased the load the on the supressor. But the amp sure did play VERY well - for a while untill the great explosion, which was about like discharging a shotgun in the room.  :o The 3  two watt carbon resistors were replaced by a small globar resistor, nichrome wire, a new plate choke, another 4-1000, and all was fine.

The power supply survived just fine, 50 ufd oil cap and K2AW rectumfryers.

I have a 4CX1000 socket here.  In the current config, the cathode end is floating (on the 3CX1000) and the end of the fils that DOESN'T have the cathode connected is grounded.  That was the only problem I had with modifying the socket.  I'd have to elevate the tube above the base of the socket by 1 'pin' to get it to ground the cathode.  To drive the cathode, I'd have a HELL of a time, as the tube would have to drop so far into the socket the anode sits on top of the socket :(

I'm trying to use the Eimac socket, it's what I had on hand.  Once I disassembled it, I found that a couple of the (I assume) BeO insulators where cracked and split.  So, that project was shelved until I can source another socket.  In the meantime, I've come into some 4-1000s, so it's on hold.

I'd LOVE to get a low drive tube in it instead.  65-80 for full output would be nice!

--Shane