grounded grid linear for AM

[kv5i]

I am interested in boatanchors on cw and am. I have run barefoot so far, but I have a 4-1000a amp I built years ago for dxing and contesting. I know it worked great on cw, but what can I expect from osing it on AM?
Thanks, Ed

[kb3ouk]

It should work fine for AM, but keep in mind that the carrier power is going to be about a quarter of what you got out of it in CW.

[K1JJ]

Depending on the power tube, as a ROUGH Guide, if a tube has excellent cooling and unlimited power supply power available, expect the following MAXIMUM power output. For some tubes, these maximums are hammering it while others not:

1)  SSB Linear:  Peak power (pep) output = Two times the plate dissipation.

2)  AM Linear:  Carrier power output = One-half the plate dissipation.

3)  AM plate modulated, Class C:  Carrier power output =  Two times the plate dissipation.

This is probably pushing the tube very hard in many cases - and can cause overheating or IMD problems.

Example:  With very high voltages, a PAIR of 3-500Z's can do 2KW output ssb pep, (1500w is more common)  -   about 500w AM carrrier in linear (pushing it hard)  -  and about 2KW carrier class C.  (more like 1500w carrier is better)

* In contrast, a  4-1000A, even though it is rated 1KW of plate dissipation, is really more like a 1.5 KW dissipation tube with lots of air -  if run with very high voltages.  It is easier to get the pair of 3-500Z's maximum power numbers listed above using a single 4X1.

Tom, K1JJ

[KM1H]

It should work fine for AM, but keep in mind that the carrier power is going to be about a quarter of what you got out of it in CW.

Carrier efficiency is around 30% while the AM sideband efficiency is the same as on SSB. The end result is some mathematical resultant but not as bad as it is often stated. This is where carrier control AM really helps, even a slight drop in carrier without modulation is a huge benefit.

The biggest problem with most GG AM linears is the already puny power transformer. I can run a LK-500ZC that uses a pair of 3-500Z's at 400W AM carrier all night even with the less than ideal cooling which is 1600W PEP at 100% symetric modulation. The iron happens to be the large 46# Dahl Hypersil external option which was standard on the LK-550 series.

Dont try that with a SB-220!

Carl

[KJ4OLL]

Very interesting discussion - I have a spare 30L-1 linear that I have considered modifying for AM use,
but not sure if it would actually be worth the trouble, in terms of improving the signal strength /copy quality
from the receiver's perspective.

- I have a DAHL RTTY main transformer available for installation in the 30L-1, resolves the original transformer
  winding overheating/melting issue.

- Updated Young Kim power supply already installed

- Svetlana 572B's installed

- suplementary cooling fan for the tubes ready

Would this be a good way to proceed to get my 32V-2 heard on more AM nets, or ?

73
KJ4OLL

[K1JJ]

Would this be a good way to proceed to get my 32V-2 heard on more AM nets, or ?
73
KJ4OLL

For AM, you are comparing a pair 6146's in class C = 125 watts carrier....  to four of 572B's in linear = about 175 watts carrier.    Because class C is more efficient on AM, the little 6146's will do ALMOST the same job, within one db..

If, however, you were to build a PAIR of 572B's into a class C plate modulated rig, they would put out an easy 500 watts of carrier.

Bottom line is if you want to drive a linear with the 32V2 to increase the carrier output power above 125 watts, you will need an amplifier that can dissipate about 1KW. This means a 4-1000A  linear, a pair of 3-500Z's  linear, etc.  Even so, this will give you only about 5 to 6db improvement, which is significant, but could be better.

A linear worthy of being driven by a 32V2 would probably be more like a 3CX-3000A7.

My suggestion would be to use the 32V2 as-is for the "little rig" and build yourself up a pair of class C 813's plate modulated by a pair. This will give you anywhere from 300-800 watts carrier, depending on voltages, etc. Parts relatively easy to get, you will learn a lot and a plate modulated  813 rig is a class act on AM...

Here's a schematic that has been built by some of the guys here:

http://www.amwindow.org/tech/htm/813/813.htm

Also check out the class E solid state rig designs by WA1QIX. They may be better suited to what you wish to do.

73,

Tom, K1JJ

[K5IIA]

Very interesting discussion - I have a spare 30L-1 linear that I have considered modifying for AM use,
but not sure if it would actually be worth the trouble, in terms of improving the signal strength /copy quality
from the receiver's perspective.

- I have a DAHL RTTY main transformer available for installation in the 30L-1, resolves the original transformer
  winding overheating/melting issue.

- Updated Young Kim power supply already installed

- Svetlana 572B's installed

- suplementary cooling fan for the tubes ready

Would this be a good way to proceed to get my 32V-2 heard on more AM nets, or ?

73
KJ4OLL

Is there any way to get your antenna any higher?

If you already have 100 or so watt carrier. A little improvement in antenna if possible and some audio mods to get a higher quality of modulation should have you rocking and rolling on any net you want to join in on. But if your antenna is maxed out then some brute for power is your only hope.

[KM1H]

With the 4 572B's and the Dahl iron you will be limited to what the iron can deliver since the voltage is well below a 572B capapility. The Svetlana tubes are considered 572B Lite due to the smaller anode and soft glass used and around 125W PD each giving you the rough equivalent of a single 3-500Z.

A 30L1 stock is good for about 600W PEP so if you can get a 175W carrier you are doing good since your still stuck with the 811A HV.
A Clipperton L or Hunter 2000B is a better choice for 4  572B's but neither has a robust PS either.

I dont see much benefit since you will need to reduce drive to around 20W and that is not easy to do with a 32V2 and still maintaing good linearity by itself unless its thru a pad. This would make a nice amp for a DX-35, 60, Knight T-60 and some Globes.
The 32V2 needs to see some big glass tubes that can handle it such as a pair of 4-1000A's or even a ceramic 3CX3000A7 or YC-156

Carl

[KJ4OLL]

Hi,
Thanks for the great tips and ideas!
I would gladly avoid "hot rodding" the 30L-1 if I could build a homebrew.
 
I like the 813 design! The "E" is interesting, but I fear solid state amps...

Are there any known good, stable homebrew designs for the 3CX3000A7?

My SSB amp uses a 4CX1000A, and I'm accustomed to working w/ them.

I have a few spare 4CX1500B's, not sure why. Possibly use one for am AM homebrew?

Really need a better antenna too.
The multi-band "trap" vertical resonates at one narrow frequency:



I have the antenna tuned so that the best SWR is at 3.85, but it looks like this away from the resonant point:
BAND   Mhz         SWR
80M         4.0         9.8   
           3.85         1.6
           3.8         3.9

I want to put up a wire antenna, possibly center-fed w/ ladder-line, but I have no way to get it much above 40'.

So plan "B" is to use brute force (amp), and possibly retune the vertical antenna for 3885, as I seem to end up there most.

73
KJ4OLL

[W2VW]

The vertical on 80 meters will not deliver performance desired by most active AMers.

The low angle of radiation will make your signal very weak within several hundred miles.

Center fed doublet at 40' isn't terrible. 60' would be great though.

[W1RKW]

I built the 813 rig that Tom/K1JJ mentions in his previous post.  I can attest to the performance of this lash-up. It's a solid design.  My 813 lash-up easily produces 550w carrier.

[Opcom]

It's only efficient at full output. Use the biggest tubes you can.

[KM1H]

The 3CX3000A7 is one of the easiest tubes to build around as it very stable and forgiving. You can use the grid as a dummy load

The YC-156 is even easier, just bolt its flange, which is the grid, to the chassis. With 5KW of PD you dont need to worry about efficiency and at a QRP 1000W carrier it doesnt require much air or HV   The 15v 15A filament is easy on the electric bill altho its takes 6-8 minutes to warm up while the 3000 is instant on with 7.5V and 50A.

Both tubes are readily available as pulls but stay away from a Svetlana 3000.

[K1JJ]

The 3CX3000A7 is one of the easiest tubes to build around as it very stable and forgiving. You can use the grid as a dummy load

The YC-156 is even easier, just bolt its flange, which is the grid, to the chassis. With 5KW of PD you dont need to worry about efficiency and at a QRP 1000W carrier it doesnt require much air or HV  The 15v 15A filament is easy on the electric bill altho its takes 6-8 minutes to warm up while the 3000 is instant on with 7.5V and 50A.

Both tubes are readily available as pulls but stay away from a Svetlana 3000.

Agreed!

BTW, about 15 years ago, I got stuck with a pair of Sweatlana 3CX-3000A7's.   Their IMD was horrible no matter what I did.  What kind of problems are the Sweats famous for?

The YC-156, being designed for clean, finicky MRI  service, is probably the BEST deal out there for hams in all respects. Heck, run 3-4KV with 50 watts drive and you have a legal limit amplifier that is probably the cleanest thing on the planet - due to the excellent IMD specs (-40 3rd) and the fact that it is being run at 1/4 below its "loafing" level.     The filament power required is low because its an indirectly heated filament.

What's not to like?   The only problem is they use solid state devices for MRI now and the YC-156 supply has dried up.  CBer's have grabbed them and there are many used tubes that look like they had a blow touch torture session.

T

[KM1H]

BTW, about 15 years ago, I got stuck with a pair of Sweatlana 3CX-3000A7's.   Their IMD was horrible no matter what I did.  What kind of problems are the Sweats famous for?

Short service life in BC service and cant be rebuilt as can Eimac and Chinese. Some AM and FM stations have boxes of them stacked waiting for a rebuilder to come along.

Alan Bond often has good output YC-156's for sale. The YC-179 and 3CPX5000A7 are similar with the latter being the parent. They are a 3CX15000B7 internals with a 5000 anode cooler.