1. Home
  2. Electronics Design
  3. Potential transformer primary to secondary phase shift

Potential transformer primary to secondary phase shift

I've never been solid in the analog side of EE and I'm getting mauled by a problem at work. We have a test bench that is testing LVDTs. The configuration is we have a board generating the excitation voltage at the frequency the LVDTs are operating. Off of this excitation board are lines traveling a few feet into a DAC and then probably ~15ft out to the primary side (called Vp) of the LVDT which is housed in a linear actuator. The LVDT secondary is center-tapped delivering Va and Vb another ~15ft back to some other channels on our DAC.

When we measure this device on a different bench using short wires and a Tek scope we get very little phase shift from the primary to secondary side (on the order of 0-2 degrees). This is expected behavior. But when we test this in our other stand with the long wiring we are getting shifts in the 7-10 degree range. I'm fairly certain this is capacitance built up on the lines but I'm sort of stuck on what to do next to prove out exactly where the problem is and how to calibrate it out of the system.

If we had a large sample of parts we'd just do an A-B comparison from one bench to the other and if the results are repeatable we'd just use this as our offset but we unfortunately don't have a large part base to draw from at this point.

Reply to
DigitalPlease
Loading thread data ...

I've not played with LVDT's, but we are starting on some two coil magnteic susceptibility measurments, which is kinda similar.

What's the drive frequency? What is the coil inductance? What is the expected capacitance of the 15' of cable? If those numbers all 'work- out' and it is the cable capacitance, then is the 15' fixed? Can you reduce the drive frequency? Maybe you can add a bit of C to the other arm to compensate for the cable.

George H.

Reply to
George Herold

Sure, LVDTs and RVDTs have a fairly high source impedance, and cable capacitance can load them down a lot. High frequency excitation makes it worse.

It would be easy to measure the actual cable capacitance, and separately measure phase shift as a function of C applied to the LVDT.

LVDTs are often calibrated in-place, electrical output of the demodulator vs actual displacement. Some LVDT conditioners, like mine, have programmable phase shift in the demod, too.

John

Reply to
John Larkin

You need to measure your primary and both secondary inductances - mutual industances might be helpful too - and the series resistances of your primary and secondary windings.

You also need to measure the self-resonant frequency of one or other of the windings, but that is only going to give you a global figure for the inter-winding capacitances of all three coils since they will all be fairly tightly coupled. You should be able to divide up the capacitance into separate chunks that you can assign to the various coils, if it is worth the trouble.

Once you've got that kind of information you should be able set up a Spice model and get a fairly clear idea of where the phase shift is coming from. LTSpice would work fine, and has an inductor model that includes series resistance and parallel capacitance.

Plugging in the mutal inductances to turn the various inductors into the winding of a transformer is also pretty straightforward.

-- Nill Sloman, Nijmegen

Reply to
Bill Sloman

Well, twisted pair wiring has stray resistance, it's possible that the LVDT primary is just not at the same phase as the driving end. If you have a spare twisted pair, sense the voltage directly across the primary of the LVDT instead of tapping into the drive signal at the generator.

Reply to
whit3rd

You didn't specify the freq you are operating the LVDT at? It's very possible you have a SWR problem. Try loading down the end of the line with a low value R.. You most likely will get an attenuation but you can test for phase shift.

Other problem could be too much cap and too high of R in the lines, causing your basic RC shift component. This normally shows an attenuation of signal at the other end, if so.

Jamie

Reply to
Jamie

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.