attenuator pads

[W4AMV]

Precision pads are expensive and necessarily so if you desire accurate and constant attenuation into the GHz range. However, for HF work it's a shame to compromise a wideband pad for an HF application.

If you have a good junk box assortment of 1 or 2 watt resistors you can build a decent T or PI pad that will handle reasonable power. As it turns out the T pad is ideal because it's construction form factor works out nicely with connectors. 16-to-24 1 or 2 W R's sandwich between PC cards will provide 50 W. Need more power, add more R's. The game you have to play is how much attenuation do you need, what power handling and what's in the junk box. The sandwich arrangement of R's keeps the inductance low, a bit of space between each R allows free air to flow and the PC board mounting helps pull some of heat out. I ran a small fan past this 16 W pad (16-R's, 4x4 rows and columns) easily handles 50 W. This pad is 5 dB, had a 3, had a 6 but I needed a weird one, 5 dB. The attenuation shown is within a few tenths of a dB of desired and the match out to 200 MHz is excellent, nearly 30 dB return loss.

[Opcom]

Are those temperature sensitive types? Beware old carbon ones that are let to get too hot when soldering. Some modern resistors look like carbon but are not. That is a good job done there!

[W4AMV]

Hi Patrick.

Thanks. No, these are standard 1 W carbon film. Not carbon composition or metal film. You are absolutely correct that any temp shift of the R's will lead to attenuation shift, errors in measurement and so on. However, it is interesting to note that just 16 of these, which would lead to 16 W pad, is actually much better than that. Two reasons, one, the PC card sandwich tend to provide a bit of a heat sink and two, the first bank of R's closest to the P INPUT side are asked to take the majority of the heat. The remaining 2 banks of R's are just loafing I have a small fan blowing air over this little assembly and key down 50 watts, it's ice cold.

Alan