synchros

You've probably already found it, but in case not the Synchro/Resolver Handbook from DDC is a handy reference.

They're also a good source for S/D and D/S components. We integrate their RD-19230 onto a couple of cards.

NAI does make good stuff. Worked with them in the past on some small shipboard data processing equipment.

Take a look at the test gear from CSI . We've got a couple of the DSM-4 meters and a DST-400 transmitter that are okay (although I wish the transmitter was a single dial and not discrete up/down buttons for each digit). Not sure what the cost was. They also have some S/D and D/S components.

Cool. I'll get some quotes. Thanks.

I also need a benchtop 400 Hz source, a few watts maybe. Most 400 Hz synchros are 26 volts RMS, but some or 90 or 120. Any ideas?

The commercial ones are expensive, and we're cheap, so I've been there. Get a big-ass audio amp, feed 400 Hz from a function generator to the gozin and put the gozout through an appropriate transformer. Replace the FG with any old microcontroller (I think I used an ATmega48) set up as a DDS source at 400 Hz driving a DAC for a fancier version. The deluxe model (which we haven't needed) would close the loop on the output voltage but just using the volume control and a DMM has done well enough so far.

A VFD.

Cheap! Tripath is a delta-sigma "class T" amp.

Consider a VFD drive.

A source

Interesting, but are those unfiltered PWM drives? We have one motor controller here that trashed all the test benches in the building... it radiated ghastly ringy switching transients everywhere. We had to surround it with home-made LC filters.

I've got a device that converts a transformer-type resolver to quadrature, so it can be read by anything that can read an encoder. You get 4096 counts/rev on a standard 2-pole resolver.

See for more info. ($150 each, so a better deal than the industrial test stand gear.)

Jon

You don't actually need to run them at the design voltage or frequency, they are basically transformers, as long as you stay below the V-T product, they just work.

Jon

If ouy are not going to run an electronic shaft with two synchros, much less will work. Remember that the synchro is just a weird transformer.

There is certain sense to stay near the design frequency of 400 Hz, but the thing will work with much less than

I had a resolver (which is just a two-phase synchro) in a flight simulator instrument, and it was just fed with a LM 741 with no problems whatsoever.

I'm sure it is a PWM scheme to get the sine-like output. But then again, so is a class T amp.

You just do what you have to do, right? If noise is an issue, go linear.

Good luck, hope you figure it out. Servos are fun to play with.

We're designing some synchro acquisition and simulation stuff and I need to really explore the voltage/frequency/angle space to make sure we're doing everything right.

For now, I guess I'll use a cheap audio amp and a function generator and a step-up transformer.

US

y,

K.

there's not much room for filters inside those small VFDs and they are meant to drive a motor so ...

a high powered pwm amplifier like some of these

-Lasse

Don't forget the "Scott-T" connection to go from three phase syncros to two phase resolvers. A couple of properly tapped transformers or op-amps with the right resistors will do it.

For just "playing around" 400 Hz synchros will run on 60 Hz reduced voltage.

I have a pair designed for 60 Hz that are strong enough to use for rotating an antenna with no amplification between.

I remember picking up a pair of "something" at Eli Heffron's junkyard in Somerville, MA... connected together, no power supply, turn one, the other one turned. Synchro's or resolver's ?? ...Jim Thompson

Yeah, one of the transformers needs a ratio of sqrt(3)/2 = 0.866. But a couple of 1:1 transformers will work if you do the math to account for the angle shift. I have a Gertsch synchro simulator, a box full of transformers and switches, and I need to simulate resolvers, too, so a scott-t would be great.

Synchros, resolvers, and their test gear are all really expensive, even surplus.

Can't play around, I've got to be serious! Like, 12 bits serious.

I just thought it was fun to do that; it spins longer that way. It's especially impressive when you consider the load from the four brushes onto the slip rings.

It doesn't sound like you have much fun.

I don't recall asking for that. If anybody does have any synchros or resolvers or test gear in their junk pile, I'd be happy to talk to them.

It's a discussion group, and you're just a crabby old git. Like Sloman, all you do is whine. Why would anybody want to live that way?