The AM Forum

Someday I'll finish my amp (seen in the avatar) and start putting out lots of carrier. I'd like to build (or buy) a 50 ohm nonreactive dummy load capable of handling about 300 watts continuously. Naturally there are a number of ways to do this at 1.8-30 MHz easily:

a) buy an ME-165 from Fair Radio, good for 400W and has an SWR/power meter, but $125 plus shipping.

b) solder a large number of resistors between two plates (PCB or brass/copper screen). But 2W carbon comp resistors aren't that easy to find any more. Can I use carbon film or metal film? Or even wirewound if I have a large enough number in parallel? Will the inductance of (say) thirty 1.5K resistors in parallel be nominal at HF?

c) use thick-film resistors. Mouser has a Caddock TO-247 package good for 100W (at the mythical 25C case temp) but still handle 50-75W at realistic temps, and they can be mounted directly to a large heat sink, for $9 each.

Any thoughts?
thanks
Charles

Buy a Bird load
or use a dipole.

I use a linearly diposed, air-cooled dummy load. It's good for 10-20 kW for extended periods. Cheap to build.

Charles,

I have two Bird units:

- a Bird power attenuator Model 8322,  200 Watts, 30 dB. of attenuation.

- a Bird Termaline load Model 82A rated at 500 Watts.  

The power ratings are conservative, for continuous duty I think.  Both of these units are basically the same size.  You can go with the thru-line attenuator only and have your cake and eat it.  You want the attenuator for working on a transmitter with a spectrum analyzer.  

You can use the attenuator as a dummy load, and you don't even need to put an additional 50 Ohm load on the output connector.  At 30 dB of attenuation, the load mismatch with the output open is insignificant.

If you have access to a Bird catalog, there are probably many other model numbers that are usuable too.  Mine all came with N connectors and I leave BNC adaptors on them.  The termaline dummy load I got at Dayton many years ago for about $60.  My 30 dB. attenuator was about $93 on e-Bay 2 years ago, so they aren't too expensive.

It hardly seems worth building one when you can buy one that's good to 500 MHz.  What you build isn't going to be good to UHF.

Some of the surplus places come up with Carborundum (non inductive) resistors for not very much cost.

Also there are Heath Cantennas that show up for not too much $$, as well as just the resistors that show up alone, eBay of course is a source.

The Cantenna uses a single 50 ohm Carborundum resistor...

If you look into the design, you'll find that there is a "shield" around the resistor - its purpose is to add a bit of capacitance which makes the thing look flat in SWR out to 30Mhz.

You can build an oil cooled dummy load using the Heath Cantenna scheme, or do something using Copper tubing/pipe of an appropriate diameter (if it's still affordable)... etc... they're good for quite a bit of power short term...

      _-_-WBear2GCR

I thought that tube was solely for improving convection. But adding capacitance makes sense too.

Do you just need a load for testing?

I would try connecting several wirewound power resistors in parallel. Twenty 1K 20 watt wirewounds in parallel should produce a 1 KW+ load (with a small muffin fan on them). The parallel connection reduces inductance.

 If you want to get religious, resonate out any residual inductance with a parallel capacitor. Tune to peak at the frequency of interest. The inductance will turn into an open circuit at resonance.

This can be experimented with if you have a generator and an RMS voltmeter like an HP-3400 and a 51 Ohm Carbon Resistor. Hook the 51 Ohm resistor in series with the output of the generator. Connect it to the top of your experimental load (the bottom along with the meter ground connected to the Gen ground). Measure the output of the generator. Measure the RMS voltage at the connection between the series resistor and the load. The voltage should be exactly 1/2 at that point if you have 51 Ohms.

If the generator has a 50 Ohm output, you do not want the series resistor. Simply measure the generator output with the load unconnected. When you connect the load, the voltage will go to 1/2.

Measure at several points 160M, 75M etc.. noting the correct setting of the calibration capacitor.

Just thinking out loud..

Mike WU2D

So I have been told!  :D

May help in the convection department as well...

        _-_-WBear2GCR

MFJ sells the dummy load in a bucket like the old Heath.  Kind of expensive but you can buy just the resistor from them for a more reasonable price.  The bucket antenna is nothing more than just a brass strap mechanically attached to the bottom and then it is folded up both sides of the resistor and the strap is soldered to the outside of an SO 230.  The center of the 239 is attached the same way.  SS hose clamps would work.

Bathed in transformer oil it is good for 100 watts over 10 minutes or 1 KW for 1 min.  Derate it for casor oil or mineral oil.  Get an empty paint can and a hole punch and have fun.

Plain old light bulbs make a fine HF load for on the cheap.
3 ea 100W, 120V bulbs in parallel is very close to 50 ohms and will take your 300 watts CCS.

We have also used electric range and oven heating elements on broadcast band transmitters, they might radiate a little on HF.

And a couple of added thoughts,

If buying a used load,  try to measure the R of the element prior to or at the time of purchase.  Abuse will often change the value .... over entoastment  usually causes the R to increase.

The MFJ oil inna 1 Gal can load appears to have a resistor at about one half of the power rating of the old Heath Cantenna.

In very little time,  I was able to make a 73 Ohm load from the MFJ unit.   Could be useful,  I guess,  but for me am using a couple of IBirds.  Much larger resistor and superior engineering.


YMMV  GL   Vic  K6IC

Thanks for the input, guys  ;) I'm sure that pesky part 97 has something in there about not using antennas as dummy loads  :P

Of course I could buy one (and probably will look for a Bird as suggested). I hadn't thought of using an attenuator with the other end unterminated. But it's more fun to build it myself...

Coincidentally I just repaired a Tek 130 L-C meter that I got at Dayton a couple of years ago for $10 (missing the fuse holder cap and power cord, usually not a good sign). It just needed a fuse, a thorough cleaning and recalibration. So I will try it out on a couple of wirewounds and see just how inductive they are. I also have a Tek 190 constant-voltage signal generator to use for "divider" experiments as Mike suggested.

I agree that three paralleled 100W bulbs would be pretty close to 50 ohms at full brightness. Unfortunately the resistance of the filaments is nonlinear and is considerably lower when the bulb is "cold" or dimly lit.

-Charles

Back in my novice days, a single 100w bulb worked fine as a dummy load for a 6146 rig... they're good enough for testing, if you need to test, and easy to read for max power out too!!  ;) ;) ;D

Wonder how a nice big 500w halogen bulb off one of those "torchiers" or outdoor/worklights would do...

         _-_-WBear2GCR

I use a dozen 600-ohm Glo-bar carborundum resistors in parallel.  Not sure the exact rating, but they are each about 18" long and 1" in diameter.  I  mounted them vertically so that the hollow space in the middle would create a chimney effect and improve convection cooling.  I can run it at a kw for hours at a time, with no smoke.

I notice the resistance changes slightly when hot, but the SWR meter still reads the same - about 1.05:1.

I used to have some large wirewound "noninductive" resistors that made fair dummy  loads for 160 and 75, but I had to use a variable cap to tune out the reactance.  They were otherwise conventional vitreous-enamelled wirewound power resistors, wound with two layers of wire, each one wound in the opposite sense as the other, so that the magnetic field of each coil cancelled out (most of) the field of the other.

Lets get serious. Has anyone found a cheap & dirty way to build a dummy load for testing BC rigs that can take maybe 10KW? How about oven elements?

Sounds like somebody still has some 4cx5000's hanging around. :-X

 "Lets get serious. Has anyone found a cheap & dirty way to build a dummy load for testing BC rigs that can take maybe 10KW? How about oven elements? "

 just when were having fun..................... Try these boys, there near rockchester

http://www.globar.com/ec/resistor.php.html 

this one auta work for 9KW

Type Style Inside Diameter (inches) Outside Diameter (inches) Thickness (inches) Peak Energy (joules) Available Resistances
                                                                                                                                                            Minimum   Maximum
911DS Solid Disc - 1.60±0.06                           1.000±0.040                                        9000                          1.6          100

  or    stove elements,  plate glass and tinfoil  for  capacitors, and HVAC tubing for coils. build a big tank ckt... 
  klc

Use an old hot water heater.

Take a large plasdick garbage can and suspend two large metal plates facing each other. Hook the plates to the transmitter and fill the can with water. Add salt and stir - and move the plates closer or farther apart- continue to add more salt until you see 1:1 on the Bird under power. With enuff power, the water will eventually boil from the heat. Probably good for 50KW for 5 minutes?  Load stability may be an issue, but keep moving the plates.

Just like a big water-cooled, lossy RF capacitor.

I've never tried it, but was told it was used long ago by the OT's for really big BC rigs... ;)

It wud be fun to experiment with a scaled down 1KW version

T

HMMMMMMMMMM............a JJ QRO slim-a tron

Yes, a very good analogy. Salt water is salt water.

T

The 10 KW shortwave I worked on used a couple 200 watt resistor in a can(pretty good size).  They pump water through it to cool the resistors. 

Ah yes, the big can with plates and saltwater... 

That is exactly what we used on the 50 KW 194 kc AN-FRT4 up at BW-8 Greenland.

A;so had a small Press Wireless PW-15 that took a bit longer before the water boiled but with the outside temp at -20f, we had time to do what we needed to do before things became too unstalbe.

Oh, the FRT-4 fed a 850 foot stick and the PW-15 was on a 600 foot stick...

Ralph, W3GL.

If your really REALLY slick, you can make one of the plates stationary, and slide the other on a track with a piece of threaded rod and a small gear motor and make this JS'ed frankinload remotely adjustable. (unless you planned on boiling a couple hundred ears of corn, then you can put it outside preferably far from the house.)

I would go for a 55gal drum, or maybe a 275 gallon oil tank.

I didn't have a spare 275 gal tank, or room in the shack for one, or a floor that can hold a ton of steel+salt water  ;D Anyway I only need a couple of hundred watt capability.  So I made this contraption out of the junkbox. Total for parts would be less than $10...
(http://i13.tinypic.com/332ng91.jpg)

Ten 470 ohm, 10 watt "sand" cased wirewounds, a piece of copper foil, and an old air variable. Without the cap it's fine on 75 and 40, but SWR rises to 2:1 on 20 and 2.5:1 on 15. I don't have an analyzer handy so I don't know what the reactance actually is... but tune the cap and the SWR is nearly unmeasurable on 20 and 1.2:1 on 15. I plan to mount the "hat" on a piece of copper-clad board with a hole in it, and a muffin fan on the other side, so it can handle well above the 100W dissipation in still air. Shorter leads and a strap (instead of wire) connecting the cap, and it'll probably be fine on 10m too  8)

-Charles

Charles,

Congrats on a very nice use of junk box parts, OM!

Most would think the wire wounds wouldn't work, but you solved it with the variable cap for the higher freqs.  The fan is a good idea too.

I once made a dummy load using (24) 100W lightbulbs.  It required an L/C network and changed impedance the first few seconds as the bulbs warmed up... ;D

Yes, a network to match/tune out the reactance, it is a great idea.

73,
T