10 watt potentiometer

[Top Shelf]

Hey folks, first time I've posted in this section. I'm going to build the impedance measuring (rig) as suggested by Vance Dickason (pg. 158-figure 8.5)-Loudspeaker Design Cookbook, which states that I need a 10 watt pot. Anyone have a suggestion for a good manufacturer for such a device? Having never dealt with pots before I don't know what resistance variance I need and different percent accuracies. I have looked at several manufacturers and am confused. Can I use an adjustable power resistor that is 25k ohms and 50 watts say from the manufacturer "Ohmite".
Also I have a frequency generator ,the Neutrik Minirator MR-1 and am wondering if this will suffice for this test and negate the need for a frequency counter due to the fact that it has a built in freq. display?
Any help would be greatly appreciated,and yes I have done a search thru this site for this type of question but if you know of a thread dealing with this that would be cool.Thanks!, Jeremy

[wakibaki]

My paperback copy of Dickason (fourth edition) has 142 pages, and Figure 8.5 is an impedance Nyquist plot???

The impedance measuring techniques in my copy use fixed resistors.

'Potentiometers' carrying power are often 2-terminal devices called rheostats and they're a disappearing breed.

I think you need to come at this from a different direction, perhaps a new copy of Dickason?

w

[Top Shelf]

I'm really sorry, I was going on the assumption that everyone had the the expanded sixth edition (or newer). And he also

recommends trying various fixed value resistors in "a calibrated resistance substitution box (one that can handle sufficient current so

you don't burn out the resistors) or a 10 watt potentiometer. Find the impedance value at any frequency by noting the voltage at

that frequency with the woofer connected. Remove the woofer and substitute RX and vary it's value until the same voltage is

obtained. RX is then removed and measured with an ohmmeter to get an impedance value"

I would like to think it would be much easier and convenient to have a single resistor that "does it all" instead of try and see with a

bunch of different ones.

Is there a recommendation for a good quality unit with 10+ watt handling and a 1% tolerance that some of you have used before?

[Conrad Hoffman]

The first thing that comes to mind is the typical medium sized Ohmite rheostat. No tolerance needed because you're going to set it and measure it. I suppose one could build a switchable resistor bank just like the usual substitution box, but with power resistors- IMO, a giant PITA. Another way to do it is with a few rotary switches and lengths of resistance wire rather than resistors. Cut the correct lengths and solder loops from terminal to terminal. If rotary switches are unavailable or too expensive it can be done on the cheap with a 0.156" header and some plugs. Solder the resistors or resistance wire from pin to pin (on the bottom) and connect to the appropriate terminal with a wire on a single hole header. I built a whole KVD using that technique and it worked well. There's also the time honored radio tuning method where you just use an alligator clip to clip to a coil at the right spot. Clip to a big loop of bare resistance wire, then measure it.

It seems I need to update my ancient Loudspeaker Cookbook!

CH

[richie00boy]

Disckason gives another method which is less fiddly and time consuming, and doesn't require hard to get parts...

[Top Shelf]

Thanks Conrad very enlightening. Richie00boy are you speaking of the voltage divider method instead of the current source method where it only uses the 1000 ohm 5 watt resistor? I wanted to use the current source method due to Vance stating that the voltage divider test is not true at higher impedances and the current source can be done at higher power levels(power amp).

[wakibaki]

Originally Posted by Conrad Hoffman

It seems I need to update my ancient Loudspeaker Cookbook!

Hmm. Me too.

w

[richie00boy]

You are talking about the voltage divider method, I am talking about the current source method with the 1k ohm resistor. I think you are getting mixed up with which method is what. The current source method does lose a little accuracy with very high impedance peaks, but in my experience it's small, and the bigger the current source resistor you can get away with the smaller the error becomes.

[Top Shelf]

I could be mixed up,but i'm 35 and dementia hasn't settled in yet...I think? I have the american version thats probably a misprint. Either way, my book tells me... 6th edition LDCB ( Figure 8.5- pg.158 Impedance measurement-current source method.) Here's the chain: sine wave generator-amplifier-frequency counter-speaker and/or substitution resistance-volt meter( on negative lead)series parralelled with a 0.1 ohm 5 watt resistor. then on (pg.156-figure 8.3 Impedance measurement-voltage divider method)and this chain of components: sine wave ocillator-frequency counter-1000 ohm/5 watt resistor-A/C voltmeter-speaker ,Is my book wrong? Just so I know so I can order the 7th edition and maybe get the right facts. Thanks

[richie00boy]

OK just looked at my 6th edition and I think it's a terminology issue p156 is what Vance calls the constant current method and p158 the current source method. I would say that that constant current by definition is a current cource. Anyway, the method I would recommend is that on p156 which avoids you messing around with switch boxes and tends to give easier to get results.