On Mon, 21 Oct 2002, Roman Black wrote:

*>A call around some of the Jaycar stores may reveal
*>old stocks of the model I got about 3 years back,
*>model QM-1480. I really think the 2mH range is
*>needed if you want to wind smaller inductors.

Ok, I will check. Although I have no idea what or who Jaycar is.

*>I also have a good RLC bridge my dad built as
*>part of his training and gave to me when I was a
*>pup, but unless you want to tune resonant circuits
*>I think they suck as a device to measure capacitance
*>or inductance, especially when you have hundreds
*>of parts to sort through.

I do not aggree. A bridge is THE instrument to measure L and C with.
Especially a bridge with variable generator frequency. I'll eventually
build one that will use an external signal generator as source. Bridges
are especially useful if you are trying to tune or match parts to a spec
because the indication (at least on a manual bridge) is relative to the
reference, so you read error wrt wanted spec directly.

Has any1 got a bridge schematic or link that uses synchronous quadrature
detectors to read Re and Im simultaneously ? Because this is what I want
to build.

*>There still has been no response from any inductor
*>gurus here about building a good analogue inductance
*>meter?
*>
*>Surely all it needs is an oscillator using the
*>inductor under test, and an analogue meter showing
*>freq which is then calibrated in uH??

The scale would be a square root one. Not nice for linearity. You could
calibrate it by slowly adding turns to a pot core ... lots of time. But
less than a weekend's work.

I've tried to make a L meter using W = L*i^2/2. At constant frequency one
switches a current into the coil (through a precise current limiter), then
cuts the current and rectifies and integrates the resulting flyback pulse,
to obtain some charge in a capacitor.  When discharged continuously
through a constant current source, this voltage should be proportional to
L, less nonlinear losses. But it is not quite, even if using ideal diodes
etc in SPICE. With real world components linearity is terrible.

Peter

--
http://www.piclist.com hint: The PICList is archived three different
ways.  See http://www.piclist.com/#archives for details.