Unspecified Specifications

[moffy]

Recently I have been trying to come up with a 0 - 10MHz volume control. After fussing around for a while with VCAs, JFETs, Gilbert cells etc I thought about an ordinary 1k cermet pot, but of course the BW is never mentioned. By setting the POT to midscale (max resistance) and using a frequeny generator I got a BW of 30MHz! I was pleasantly surprised. I thought I would ask for other 'Unspecified Specs' that would be useful to know or just fun.

Over to you.

[T3sl4co1l]

It's probably even more than that, your circuit just isn't well matched (i.e., probing it with a scope, you will find the probe has relatively low characteristic impedance at high frequencies; a constant-Z probe would show different results).

In general, wires and traces cluster in the 100 ohm range, characteristic impedance.  Resistors near the same range (say 22 to 220 ohms) tend to have high cutoff frequencies, with values farther out having a ~proportionally lower cutoff.  (Example: if a 1k resistor has a cutoff at 500MHz, a 1Meg resistor has a cutoff at 500kHz.)

So a somewhat lower impedance pot may be better suited to your application.

Other features to expect: single turn, vertical adjust types will have a single horseshoe of resistive material, and sit relatively close to the PCB ground plane.  The length of the track affects frequency response (everything here is proportional to length, of course), so small SMD pots are better than large THT pots, and vertical is better than horizontal adjust.

Needless to say, multiturn pots are much poorer -- the track is a helix going around the shaft, so will have lots of inductance and weird "hook" effects (distributed capacitance); as will any wirewound type.  These are best left for low frequency application, like adjusting DC control voltage to a VCA or such.

Tim

[BrianHG]

In an old design, I would take the output of a 1k pot & pass it through a J-Fet voltage follower amp.  It was flat on any adjustment setting all the way to 50Mhz at least.  At the time, this was the limit of my scope and function generator.

[Gregg]

I'll just leave this here in case this has happened to others 

[moffy]

Thanks for the responses very informative! Some things just aren't in the data sheets.
Another thing I had to measure recently was the transmission characteristics of coloured perspex, to see if I could make a transparent/coloured perspex box to block any emissions from my laser etcher(7w, 450nm) escaping. I searched the Net but the data was largely lacking, though it appears that orange is the preferred colour for blocking blue/UV light. I then rang around perspex suppliers but they also didn't know the transmission/absorption characteristics of the perspex. Finally I just went to Jaycar and bought a super bright 470nm blue LED, a battery case and resistor and made my own tester. Then I found a supplier that would let me 'test' their orange perspex, and voilà the 3mm orange perspex completely blocked the blue light.
They then manufactured a very nice box, at a reasonable price for me to put over the laser etcher.