I stumbled across this picture of a special Stanford SR620 counter somewhere on the net.
Gerhard
Attachments
I shelved that project for now, I became somewhat unhappy with several of the members and asked myself why I was bothering with ankle biters. That thread has since been closed. I still have the stuff, just stored away.
So now I am spending my energy setting my workshop up for post-employment.
Some of my friends retired without hobbies, not a good thing.
Jn
The worm and gear I will be able to adjust the depthing, and I expect to be able to get to zero backlash. My rotary table has the ability to adjust it down to zero as well by lowering the depthing.
Preloading is also viable if I can't zero it out, I'd probably use a dc brush servo setup in constant current (torque) mode.
At work, the upper magnet array is preloaded by gravity, and the magnetic attraction to the lower array increases the force so much that they need to put springs in to fight the forces so the motor can pull it up. The lower arrays can actually go zero preload as the magnetic force will be more than the array weighs. Springs are used there to keep the preload from crossing zero. I also detailed to the controls people how to use the encoders on the motor and load to measure the flexure of the 66 mm diameter ball screw and consequently, how to adjust the springs to reduce motor load variations at closed gaps. Required accuracy is 1 micron, but the ballscrews change length in the 45 to 70 micron range. One vendor got into real trouble as they couldn't get stable motion using the load encoders, so would normalize position of the motor encoders before a move, use the motor encoders in the PID loop,and the two ball screws flexure variations would lockout the device because of tilt.
Needless to say, the PID algorithms for top and bottom arrays can go nuts if the programmer is not that well versed (so to speak). Trying to gain it up to reduce errors can easily get into trouble as the huge variations in forces make the frequency response of the control loop change with position. Too high a gain to reduce errors when the arrays are close can cause oscillation when the gap is opened, the forces reduce, and the unity gain point shifts enough to zero the phase margin.
Alloys..the recirc ball leadscrew is stainless (16 ppm/C), I'll probably make all the EDM mechanism in alum (25.5ppm) or brass (16ppm IIRC). I will probably use G-10 (or G-11) as the insulating structure for the wire guides, it is 16 ppm.
I know the tie point for the ballscrew on the base, the bearing attach point on the slide, and the extension of the metals from the slide bearing point to the wire guides.
But honestly, temp variations can be quashed with a temperature controlled chamber, I would probably just go 5C above ambient and keep it there.
The very first thing I will do is a homing routine on a known entity where the wire contact is sensed, and zero the slide position.
The real unknowns will be, how hard will it be to clear the residue out of the cut, as a lower operational voltage reduces the kerf which reduces clearing waterflow. Also, at that level, will the debris change the flashover voltage because of conductivity.
jn
Last edited:
No, quatloos.
Commercial Off The Shelf. Try and keep up with the acronyms.
Transformers acronyms - YouTube
jn
And if that form of EDM fails, you can make some cool beats so the teens and 20somethings thing you're an old school DJ.
Look forward to seeing where you go with it.
Might want to read up on some of tech Kern does on their mills -- aluminum chassis with integral plumbing to keep the chassis constant temperature (and epoxy-granite elsewhere for vibration absorption). Chasing zeros is both a fun and maddening pursuit!
Thought about going to linear motors, by the way?
um, I was an old school DJ... back in the 70's/80's.
Aerotech was pushing their linear motors for my machine upgrade. We already had a full machine drive system, a 4 meter long leadscrew about 1.5 inches diameter, so just need the motors with gear reduction.
But the quotes on even NEMA 42 3 phase brushless with encoders are unbelievable. One motor costs more than my lathe, mill, table saw and 3d printers put together. I suspect even including all the other tools I've bought.
So a linear is not in my ballpark.
I wonder if there are chinese linears out there, the rails and ballscrews I've found on the web are way too cheap.
Right now I'm doing up the design for the worm gear drive using openscad. It's great to see all the interferences before making chips.
jn
back in the 70's/80's.Aerotech was pushing their linear motors for my machine upgrade. We already had a full machine drive system, a 4 meter long leadscrew about 1.5 inches diameter, so just need the motors with gear reduction.
But the quotes on even NEMA 42 3 phase brushless with encoders are unbelievable. One motor costs more than my lathe, mill, table saw and 3d printers put together. I suspect even including all the other tools I've bought.
So a linear is not in my ballpark.
I wonder if there are chinese linears out there, the rails and ballscrews I've found on the web are way too cheap.
Right now I'm doing up the design for the worm gear drive using openscad. It's great to see all the interferences before making chips.
jn
Last edited:
I never said 'later in life' I said 'later in his career'. Which is accurate. He discovered the 'Josephson effect' in 1962 (not sure when it was called that first). So 1970 is 'later in his career'.
</pedant>
um, I was an old school DJ...
I'm trying to imagine you with the chest wig and medallion here...
Story goes that Josephson went barking mad and started investigating metaphysics later in his career. You can't sack a nobel laureate so he was just left to try and levitate his desk with the power of his mind. The reality may be slighly less embellished
I did get a tour of a voltage calibration lab in the 80s when chosing a university. I have never seen so many digits on a volt meter before or since! First and last time I saw a standard cell as well. I am glad some will spend a good chunk of their careers coming up with more accurate ways of measuring fundamentals but I lack the stoicism to be one of them.
Yeah, he ended up being a bit eccentric
I first got to hear about a Josephson Junction Reference (JJR) back in 1988/9. I was working on a 5 1/2 digit panel indicator (programmable and could take a wide range of process sensor signals directly in). We had an ancient HP 5 or 6 digit lab DVM that wasn’t up to the job, so I got a new 8 1/2 digit HP meter. We used to get the HP Journal in and they discussed the design of the 8 1/2 digit meter and the use of the JJR during its development in one of the journal features.
Very exciting times for me personally as a 30 something.
Last edited:
Shockley was off the scale weird. Lots of stories about the difficulty of working with him - most folks bailed after a few months. Massive ego issues and then there was the eugenics things that really turned people against him. Even Bardeen found him too much to deal with.
Funny, some highly gifted folks are humble, nice to people around them and deliver the goods (Einstein is the gold standard) and others are complete rectal orifaces.
Last edited:
Don't even think of it. None of my DJ friends in the other clubs in Newport were like that. There was one patron like that, reminded me of Mr Kotter, spent all day lying on the beach doin the tan thing. Most of the women didn't go for that type. It was certainly different in Manhattan, but I turned down a job there because I figured there was more opportunity to meet gorgeous and willing women if I were an electrical engineer, as opposed to a club DJ in Manhattan...
Remember, I didn't say I was smart...
Totally agree. The Nobel winner guys I've met were really down to earth, amazingly so. Only one guy was unpleasant, but he was a professor from Harvard.
Even the ALPHA antimatter lead guy Jeffry Hangst is a really great guy, as well as his cohorts. Totally fun to talk to.
jn
Last edited:
I had a professor that worked with him in the late 50's he did have a serious car accident with permanent brain damage.
Wow! That thing's been counting 10 GHz for 300 million years!
Oh, the real SR620 can deliver abt. 12 digits in a second, it is not such a simple
counter. Single shot ~ 20 ps time resolution. I love that thing!
Gerhard
I met a few as an EE. One told me "I cant believe I'm in love with a geek"!