kelvin varley divider schematic or theory info ?

It's not only the super-precise resistors, but the high-quality low-resistance switches that make them expensive. You can find them on eBay for $50 to $500. For example, this top-of-the-line 7-digit Fluke 720A is at $393 with 4 bids and 80-minutes to go,

Other less-precise models go for less money, such as 99 cents,

No, those look the same, but are called ratio transformers. :-)

What about this one, four digits, but only $10.

e=2E

A Blumlein bridge uses a transformer for two of the arms. IIRR the Kelvin-Varley decade bridge had sets of eleven resistors, and one would parallel two of them with a divider whose resistance was nominally equal to the sum of the two resistors being parallelled, to allow you to sub-divide in decades

The British National Physical Laboratory at Teddingon sells a "good practice" guide for measuring reistance and impedance below 1MHz that ought cover the Kelvin-Varley bridge. For 25 UK pounds it would be a lot cheaper than a bridge.

"Alternating current bridge methods, for the measurement of inductance, capacitance, and effective resistance at low and acoustic frequencies" by B Hague seems to be available from Amazon if you want to spend a bit more money.

--=20 Bill Sloman, Nijmegen

Sounds good. The last time I priced the concentric-dial kind, they were $1500.

But a 16-bit DAC can be had for a few dollars.

John

A very beautiful post Tony!

Yahoo,

use. I

need

piece.

resistor

is no

Rodo,

You didn't try "kelvin varley voltage divider theory" using Google or you would have found this -

"Rodo" a écrit dans le message de news:4VGxf.38404$v84.14495@trnddc06...

need

resistor

no

These are not made of resistors, at least not I know of, but with accurate ratio transformers, which are more accurate and stable.

And given the accuracy you obtain for the price it seems OK to me. Oh, you're not telling us you expect few PPMs for $10, don't you?

One good way to get a high-performance Kelvin-Varley divider at a reasonable price is to buy a Fluke 343A on eBay, and use the Kelvin-Varley inside. These come up from time to time, and usually go for under $150, as this example shows:

Correction, looking at the manual, I suspect that instrument doesn't use a Kelvin-Varley divider. Sorry!

The switches need to be 2-pole, with low resistance contacts. 4-decades requires the MSB resistors matched to within 1 part in 10000 over all ambient conditions. $500 is not a bad price.

Below is a 4-decade Kelvin-Varley. It uses 3x 2 pole rotaries and 1x 1 pole. Vin | 10 +--o +---+--o +---+--o +---+--o | | | | | | | R1 | R2 | R3 | R4 | 9 | | 9 | | 9 | | 9 +--o | +--o | +--o | +--o | | | | | | | R1 | R2 | R3 | R4 | 8 | | 8 | | 8 | | 8 +--o | +--o | +--o | +--o | | | | | | | R1 | R2 | R3 | 7 | | 7 | | 7 | | 7 +--o | +--o | +--o | +--o | | | | | | | R1 | R2 | R3 | R4 | 6 | | 6 | | 6 | | 6 +--o | +--o | +--o | +--o | | | | | | | R1 | R2 | R3 | R4 | 5 | | 5 | | 5 | | 5 +--o | +--o | +--o | +--o | | | | | | | R1 | R2 | R3 | R4 | *4* /|\\ | | 4 | | 4 | | 4 +--o

I did a chip for "qrk" and company that had a 9-bit KV DAC used for tuning gm-C filters... tricky in that the successive sections had buffer amplifiers to prevent loading and avoid section-matching issues.

...Jim Thompson

The obligatory typo.....

The switch positions on the dials for the first three decades is the position *between* the 2 poles. Vout= 0.3211*Vref for the above sketch.

GR made a "decade voltage divider" that's a k-v. It looks a lot like one of their decade resistor boxes, and even old ones are stunningly accurate. Inside are manganin-wire resistors wound on mica cards.

And the old AC ratio boxes, full of transformers, show up on ebay. They're accurate to a few ppm.

John

Look around on Ebay for used Fluke differential voltmeters, such as the model 885, 887, etc. They all use a KV divider behind the dial switches. Damn good resistors too. As I remember, they are matched to a few PPM, and have excellent tempco. They have a vernier pot in the last position as well. There are manuals with schematics for some of the older Fluke DVMs on the web... One caveat, if you decide to try this route, is that the KV dividers used in these instruments may not be able to handle the voltage that you need to impose upon them. What voltage levels will you be working with??

Cheers!!!

I second that.

IMO, no self-respecting lab should be without one.

-- "Electricity is of two kinds, positive and negative. The difference is, I presume, that one comes a little more expensive, but is more durable; the other is a cheaper thing, but the moths get into it." (Stephen Leacock)

I still use a homebuilt 4-decade K-V, built about 25 years ago, (which is why I know something about it). :)

The K-V has the advantage that it only needs resistors to be accurately matched, easily done with a simple Wheatstone Bridge. A cheap way of getting a precision voltage attenuator if you are poor.

What do you use it for?

Not that often these days Fred. Used to do a lot of precision analogue instrumentation though. In the early days of being self-employed you usually have more time than money, so building your own test equipment can be a neccessity.