Dummy load questions

Because the EMC 32-1007 *is* an RF termination resistor, whereas the MC14741 is not. No one in his right mind would use single-layer wire-wound resistors as RF terminations. OK, you may get away with it if the frequency is low enough and you're not too picky. And you get to decide yourself what 'low enough' and 'not too picky' mean, because, as you observed, the relevant specs are missing.

You're definitely not very picky. The resistors are probably the dominant inductance anyway, so it doesn't really matter.

Sure, if you cool them hard enough. It says so right there in the datasheet.

Jeroen Belleman

By an interesting coincidence, I just blew up my power supply test dummy load that used four of those resistors: The room temperature resistance values are now well outside the alleged 1% specification. Oh well.

The reason these don't have any frequency specs is that they are not made for RF applications. They are "non-inductive" and they'll certainly work at the lower HF frequencies, but once you get into frequencies where the lead inductances and such become a problem, the VSWR will go up.

Yep. That's a real RF load.

Because they're physically two different types of resistors. One is made for lower frequency power dissipation. The other is made for attaching to a strip line and maintaining a constant 50 ohms well into the microwave region. What's your highest operating frequency?

Nice antenna you're making.

You do not need coax, but 6" seems awfully long. How big is this Bud Box?

My guess(tm) is that using those power resistors might be useful up to about 14 MHz. Beyond that, you'll see the VSWR climb. Find an MFJ antenna analyzer or return loss bridge, and sweep the dummy load up to your unspecified operating frequency. If there's a problem, you'll see it easily.

No, it's not rated at 50 watts. It's rated at 50 watts at 25C. As you go up in temperature, the power rating will decrease. At 100C (boiling water), it's rated at 37.5 watts max. Your ability to keep the temperature down depends on the size of the heat sink and any auxiliary cooling such as a fan or water bucket. Running devices at their maximum operating point is a bad idea.

Only if you can keep the resistor temperature down to 25C (room temperature). Suggestion: Learn by example: Note the use of mineral oil for cooling and fairly short lead lengths.

I've often wondered how people make multi-GHz high power loads and attenuators. Things must be distributed, to spread out the heat but keep the speed up.

I have a 20 GHz HP terminator, with APC7 connectors. Inside is what appears to be a sheet of glass, with a purple film, that extends from the connector all the way out to the walls of the can, with maybe some structure to the film pattern. But it's only good for a couple of watts.

At least it's at 25C ambient, and not 25C case temperature.

Inside a box the "ambient" will be much higher than outside "ambient".

LOL. This is just the sort of thing that Murphy whispered about in my head. (But I couldn't understand Murphy's whispers, which led to my SED post.)

2.25" square on the ends and 4" long. It's just large enough to accommodate two MC14741s, one on each side. Ergo, only the ends on a separate U had enough free space for the So-239. All of that doesn't matter now as the Bud Box will be returned to inventory.

I hang out at 20 and 40 meters, so it actually would've worked good enough for me "as is." But, "Any job that's worth doing is worth doing right" (rather than half-way).

Thank you for the links. This looks like the one for me, for now.

Keep an eye out on ebay for dummy loads --

Heath Cantennas are oldies but goodies, designed for HF amateur radio use. Filled with transformer or mineral oil, they'll typically take a few hundred watts up to 450 MHz for long periods, and up to a KW for brief periods.

Higher quality, look at Bird dummy loads -- type "N" connectors give you an idea of their frequency range, which is in excess of the old Heath Cantenna.

For not too much money you can get a dummy load that will take 100 watts all day long over a very wide frequency range.

If you want to make measurements, Bird also makes high-power attenuators.

Yes. Generally its in a pyramid form, with the heat being generated over a fairly small volume of resistor material, and then spread via layers of dielectric, ceramic, metal base, and finally the heat sink. The idea is for the capacitance in the resistor material, along with the inductance of the resistor material, to act as a distributed LC network forming a 50 ohm transmission line.

I had two microwave dummy loads that were rated to about 6GHz which I destroyed and later purchased. Unfortunately, I recycled them as scrap and didn't bother making any photos. So, a description instead. Inside was an amazingly small cylindrical resistor element. As I recall, about 4mm dia and 12mm long. The ends were silver plated. One end was silver plated to an N-connector. The other end went to a pyramid shaped transmission line that eventually ended at the far end of the dummy load. The finned extrusion formed the ground end of the transmission line as well as moving the hottest part of the heat sink away from the coax connector.

The flange mounted terminators are a little different. They're intended to connect to a 50 ohm strip line on a PCB. The contact is usually 0.100" above the base, allowing a variety of PCB thickness to be used. If the OP were building a microwave load, I would have suggested a strip line PCB in between the coax connector and the terminator instead of a direct wire.

Unfortunately, I found a rather miserable example of how NOT to do it. This is an early 1980's vintage terminator and attenuator combination that is used with a Cushman CE-5 service monitor. Transmitting into the service monitor tends to be expensive, so this device was used in between to reduce the power levels involved. The calibration sheet is long gone, but I recall that it was claimed to be 1.5:1 up to 1GHz and could safely dissipate 100 watts for about 3 minutes. Note the fairly long leads between the BNC connector and the terminator. Also note that the terminator is directly mounted to the heat sink instead of the case. When I'm working with 60 watt radios, it gets quite warm.

Yes. I've seen high power microwave attenuators that have multiple larger value resistors distributed along the length of a 50 ohm transmission line. These are usually done coaxially rather than flat. They offer a higher upper frequency limit, but the power dissipation ability is approximately the same as the equivalent resistor area stuffed into one big package. I skimmed through Google Images but couldn't find any relevant photos.

I've never destroyed one of those, so I don't know exactly what's inside or how it works. The film is probably conductive. If the purple is closer to yellow-green, it might be:

Please be careful if you are thinking of taking apart a higher-powered one. The ones I've worked with most recently were Bird "Termaline" and a knock-off (with thinner aluminum fins) from another manufacturer. 50 ohms, 50 or 100 watts CW, probably 6 GHz. Many of them had a "BeO" sticker on them. Beryllium oxide seems to be kind of like asbestos: big solid chunks of it are OK, but fine dust of it - like from a busted terminator - is real bad news.

They were close enough for government work from 0.978 to 1.09 GHz. Bird's spec sheet says VSWR 1.20:1 max to 6 GHz for the 50 W version, and somewhere between 1.20:1 and 1.40:1 max for the 100 W version.

There was also a test fixture that I saw in the shop, but never used myself, with a huge oil-filled Bird 50 ohm load, probably 500 W. It was used for continuous tests of transmitters down around 108-118 MHz. I don't think there were any hazardous materials in that one; the trick there was that you had to close the vent plugs to ship (to prevent leaks) and open the vent plugs to run (to prevent explosions).

In the 50 W and 100 W sizes, Bird's terminators have square or hexagon fins. Some of the knockoffs were hexagonal (IIRC) but a few of them were perfectly round.

Protip: Perfectly round terminators (and attenuators) roll around and fall off the bench a lot.

Matt Roberds

. . .

That was the standard warning everywhere I've worked. The problem was that sintered beryllium oxide, porcelain, and ordinary aluminum oxide ceramics are all white and difficult to distinguish. So, any time something white was broken, we had to treat it like hazardous waste. No grinding of anything white allowed, even if we knew it wasn't beryllium oxide.

(...)

Even the small loads are oil filled, usually with mineral oil. In the past it contained PCB's which was rather unsafe. After blowing the end cap off a dummy load with about 100 watts, I needed a quick refill. No mineral oil handy, so I used vegetable oil. The lab smelled like a fast food restaurant for a few days until I found some mineral oil.

Not on my messy bench. There's far too much junk on it for anything to roll very far.

To be fair, the winding will be somewhat coaxial (well, co...helical?) with the casing, which suggests use as a lossy transmission line. Unfortunately, the impedance of a thin wire, in multiple nearby turns, is probably too high to match at 50 ohms (even with the epoxy dielectric).

Does seem to imply there might be a particularly well suited value, a bunch of which could be connected in parallel, to get a not-horrible load. You'd want to buy a whole range and TDR them, see which ones are the most constant over frequency.

Or just buy a real terminator. :)

Tim

One nice wideband terminator is a long piece of coax with a mediocre power resistor at the end. The coax losses take care of the fast stuff (twice!) and the end terminator absorbs the slow stuff. The thermal situation shifts as a function of frequency. This is especially good for time-domain signals (steps, impulses) where there is not a single frequency to absorb.

People make coax with a resistive center conductor, maybe for this sort of use.

Indeed. Running devices way beyond their specs is however good business, for some markets.

NT