[VIDEO] I had an interesting fail -- steel cabinet collapsed and fell

I blame Hillary's email server.

Well, that is a little unusual, as it had a steel back, not the typical 4 verticals with just a couple cross braces. I've had a few of those either fail completely, or start to show signs of twisting, and I readjusted the cross braces to resist the twist. Your cabinet clearly twisted also as the first part of the collapse.

Well, sorry for the mess all over the floor.

Jon

Yes, i saw how it twisted a second before failing.

I would never have thought that it could be possible.

On May 19, 2017, Ignoramus17154 wrote (in article):

I?d bet on a busted spot weld or a missing fastener or two. With a rectangular back panel, all four corners must have bolts, or a twist is likely.

Joe

These things are actually QUITE flimsy, without the cross braces. Also, there is some method to doing the cross bracing properly. I do not actually know the correct scheme. but, you want to avoid the cross braces being put in compression, as they will buckle. I do occasionally tighten the screws on these things, as they seem to work loose over time. If the bracing screws get loose, it will collapse if heavily loaded. The screws for the shelves also need to be kept tight, as they also help resist the verticals from slanting.

Jon

I still have the cabinet laying in my yard. I will take a look.

Much to learn on a seemingly simple subject!

By The Way your cnc boards are still working for me.

Yeah, nothing that seems simple is actually simple! And, stuff that you just expect to sit there and do their job, sometimes doesn't. Still don;t know how the screws on these cabinets actually get loose over time, but when I go around tightening them, a bunch are in fact a bit loose.

Great. Glad it is working for you. A guy just returned a big order. I guess when he found out how many wires were involved, he chickened out. He has a 3-axis Fanuc-controlled mill with brushless motors and a tool changer. No WAY he is going to find a bolt-on solution that just plugs in and has the machine up and running over a weekend, which I guess is what he was hoping for.

Jon

My mill had DC motors and everything worked really nicely.

All three axes work.

Except for one thing. For 4th axis, I have a rotary table made by Troyke with a resolver (not encoder). I bought from you a "resolver to encoder signal converter".

This worked initially, HOWEVER after a while some errors started creeping in and sometimes I would get completely erroneous results, for A axis jumping randomly by thousands of degrees or some such macroscopic number of degrees.

I disabled 4th axis for now but I want to fix it. Maybe I should fit an encoder in place of that resolver or figure out why this converter is not working.

Overall I am happy but 4th axis would be a great plus for me.

It is shocking a top loaded shelf pressed out of sheet metal could wobble and fall over.

Never seen one of those fall apart before.

Hmm ... I've only seen resolvers in aircraft instruments. Those require 400 Hz power (26V IIRC). Is what you have also 400 Hz, or do they make 60 Hz resolvers, too?

If they are 400 Hz, could some subsystem which is supposed to generate the 400 Hz failing?

Good Luck, DoN.

I want to say if you have a slave pair - one is a encoder and the other a resolver. in other words - transmit and receiver.

One on a knob that is rotary or on a rack - and the other on a surface that is moving. Sometimes in reverse the surface 'tells' the pointer where to point.

Martin

Hmm ... that sounds more like synchros (AKA "selsyn"). An ac signal applied to the rotor, and three phases (Wye connection) of output. Thus the synchros had five leads, two for the rotor, and three for the Wye connection of the stator.

Apply power to both rotors, and one will move to track the other. Instead, take the second rotor, connect it to the input of a servo amplifier, and the servo will rotate the one to which it is connected until at 90 degrees, at which there is zero output, and a slight motion will provide signal either at 0 degrees phase or at 180 degrees phase, causing the motor to rotate the servo until the output is zero again. (This allows driving things which are heavier than the synchro is capable of driving directly.)

A resolver, however, has one rotor signal with two stators at 90 degrees, and isolated from each other electrically. This produces output signals on the two stators as a sine and a cosine of the rotor's angle. (Some resolvers have two rotors at 90 degrees, so they can process the output of a normal resolver with only one rotor winding, so you can add two angles. (The resolvers which I am describing are quite small, to fid in aircraft instruments.) I don't know whether there are larger 60 Hz resolvers as there are larger 60 Hz synchros (the 60 Hz ones were commonly used on shipboard, where the extra weight of the 60 Hz versions was not a significant penalty. They were the ones called "Selsyn"s. And while all 400 Hz synchros which I have seen are all pretty much the same, the receivers on the 60 Hz ones had an inertial damper on the shaft, while the transmitters did not)

The resolvers have either six leads (two for the rotor, four for the two independent windings) or eight leads (two for each winding in the rotor, and two for each winding in the stator.

So -- I would really expect the Troyke rotating table to have a synchro and a servo amp, but it could be a resolver, and some way to connect that to a servo motor and amplifier.

Enjoy, DoN.

I have always fastened (at the top) tall bookcases to the wall. I guess cabinets should be fastened as well. They tend to be overloaded by default.

That was good Don. The 60 cycles/per/second - they were of that age - off B-36 or B-52 frames. The Bronze bodies are like 5 pound coffee cans. Heavy. I have to drag them out and check the wires...

Air force used many of them in various Analog systems...

Martin