Piezoelectric focuser on eBay

ray_parkhurst · Post by **ray_parkhurst** » Wed Aug 14, 2019 8:10 pm

Lou Jost wrote:Just noticed the Hewlett Packard programmable calculator! Takes me wayyyy back.....

My wife and I still use our 32S-II's. But the weirdest thing is that my 15C from High School still has the original 3 batteries in it, and still working. The battery low indicator comes on in the cold months but it still works. We're talking 40 years here.

mawyatt · Post by **mawyatt** » Wed Aug 14, 2019 8:29 pm

I remember in grad school when a friend showed up with the original HP35, I couldn't believe they had crammed my Post bamboo slide rule and CRC Math Tables book into it!! It cost $300 and way out of my range, so was a long time before I could afford an HP Calculator.

My 32S has been with me longer than I can remember and I've only had to change the batteries once!!

Best,

Mike

Lou Jost · Post by **Lou Jost** » Wed Aug 14, 2019 8:34 pm

I had the HP25 when I was in high school, a dream come true. My favorite pastime back then was trying to make the fastest possible code to generate prime numbers...You can tell I must have been a great hit with the ladies

mawyatt · Post by **mawyatt** » Thu Aug 15, 2019 7:24 am

I have fond memories of looking into how the calculators respresent functions such as the exponentials and trig functions.

They don't just truncate a power series at so many terms, for example the e^x = Sum x^n/n! or 1 +x/1! +x^2/2! + x^3/3! ... They use coefficients from Chebyshev Polynominals instead of the reciprocal integer values shown in the series. These coefficients are derived based upon the method and bit length representing the numbers and series truncation length for the function. The idea is not to just blindly truncate a series at some length using the series which may cause a warped error distribution, but use the Chebyshev coefficients to create an equalized error distribution. In the e^x case if you truncate the series for positive real x, then the result will always be less than ideal. So in this case the coefficients would not be 1/1, 1/2, 1/6, 1/24 and so on but the Chebshev coefficients.

During a cafeteria lunch at work with a colleague, his wife (she was a Fulbright math scholar, studied in France where LaPlace studied) and another mathematician, I was discussing this series representation. Turns out they were working on ways to improve the long term accuracy of the Ring Laser Gyro using Kalman filters, and were limited by the finite word length of the computer. Using the Chebyshev coefficients in the Kalman filter calculations dramatically improve the results and became the standard Kalman filter code for RLGs!

Anyway, fond memories indeed!!

Best,

ray_parkhurst · Post by **ray_parkhurst** » Thu Aug 15, 2019 7:48 am

I bought an HP41C my junior year of high school, and learned quite a bit about how it worked (and machine language programming in general) by doing "synthetic programming". When a memory card is inserted, the calculator only allows access to the address range of the card itself. But if you remove the card while the calc is on, there is a bug which then gives you access to all the registers. I remember using that capability to make some interesting programs. I was very proud of my hex to decimal and decimal to hex converters, which by using synthetic programming could be done with the absolute minimum code.

ChrisR · Post by **ChrisR** » Thu Aug 15, 2019 9:13 am

I used mine to make words like ShELL OIL when you turned it upside down....... $\:D/$

bklein · Post by **bklein** » Fri Aug 16, 2019 5:00 pm

I'm not finding anything when I'm searching for this item# on ebay. Still available? If so, what's the trick to find em?

mawyatt · Post by **mawyatt** » Sun Aug 18, 2019 6:59 am

bklein wrote:I'm not finding anything when I'm searching for this item# on ebay. Still available? If so, what's the trick to find em?

These Physik Piezoelectric Stages sold out quickly I believe, they were quite a bargain!

The seller Starlight-Photonics has indicated that more should come available soon.

The controller we've developed here works well with these Physik devices and should work with any similar piezo devices. These devices require a very high voltage source to drive, and conventional low voltage motor controllers won't work, neither will conventional amplifiers since they are designed to drive low value resistive (and higher inductive) loads. If you hook up a conventional amplifier to these devices it will likely oscillate because they are highly capacitive (many microfarads!!!), and likely to destroy the delicate ceramic elements, maybe the amplifier too!!!

I'll be happy to explain the details if folks are interested.

Best,

mawyatt · Post by **mawyatt** » Wed Aug 28, 2019 12:42 pm

An Instrument Grade High Precision Bi-Directional Complementary Output (+- Voltage) has been developed at "Mike's Labs" for driving various Piezoelectric loads, the Physik Piezoelectric Stage included.

The Printed Circuit Boards arrived the other day and we have one assembled and ready for testing. The design is based upon a fully complementary output (using 4 output transistors) capable of supplying both + & - high voltage outputs which allows the piezo elements to be single ended (ground referenced) rather than "floating" like the Physik Piezoelectric Stage for Bi-Directional operation. The controller uses a commercial external to the PCB High Voltage DC to DC Converter that supplies +- 175VDC for the precision output driver stage and sits on top of the main PCB shown below. As with the original design, the controller is cable of driving high capacitive loads, now with +- output voltage capability and uses the standard RPi interface.

3 initial setup programs were created to allow complete testing and confirm operation with the RPi also suppling full Stack & Stitch capability and show operation while simultaneously reading the Piezo Stage Position with the integrated 16 bit ADC. A hardware offset capability is included to allow "zeroing" out the offset bias of the Strain Gauge Bridge.

Everything is working as expected

Best,

Mike

Various controllers from "Mike's Labs" for Piezo and Voice Coil Stages

Initial Piezo Stage Controller with Bi-Directional capability for "Floating Loads" with integrated + HV DC to DC Converter

Complementary Output (+- Voltage) Piezo Stage Controller with Bi-Directional capability for "Floating or Grounded Loads" with external +- HV DC to DC Converter

mawyatt · Post by **mawyatt** » Sat Aug 31, 2019 8:24 am

The Physik Piezo Stage has been configured and mounted to an ARCA clamp for mounting the the base.

Things are getting close for some optical tests and images with the Pizeo stage and controller

Best,

viktor j nilsson · Post by **viktor j nilsson** » Sat Aug 31, 2019 11:35 am

So that's how you've decided to use it! I've been meaning to ask. I still haven't decided yet what I'll do, but I think I will opt for a vertical, microscope-like setup.

mawyatt · Post by **mawyatt** » Sat Aug 31, 2019 12:11 pm

Viktor,

This allows the Piezo Stage to mount vertically or horizontally since it's attached to a standard ARCA clamp.

Awhile back I decided to "standardize" on ARCA where possible, I've found the quality traditional (not quick release) ARCA clamps and rails are very sturdy and stable. I usually directly mount a long ARCA plate to the setup base and long ARCA clamps on the bars/fixtures that holds the focus rail & camera/lens. This is on the WeMacro stand base, these are really very good and a bargain (the stand parts cost more that what William at WeMacro is charging!).

On my other setup which is based around the Thor Labs optical base and 95mm precision vertical bar, I use the same concept but the clamps are Thor clamps not ARCA, these are better clamps but much more expensive.

Best,

viktor j nilsson · Post by **viktor j nilsson** » Sat Aug 31, 2019 1:06 pm

How is it handling the weight of the stage? I haven't read any spec sheets, but I was under the impression that it wouldn't enjoy a heavy load.

viktor j nilsson · Post by **viktor j nilsson** » Sat Aug 31, 2019 1:40 pm

bklein wrote:I'm not finding anything when I'm searching for this item# on ebay. Still available? If so, what's the trick to find em?

https://www.ebay.com/itm/254282702002

The seller has listed more of these stages now, at a slightly higher price ($57 vs $37).

mawyatt · Post by **mawyatt** » Sat Aug 31, 2019 1:47 pm

Weight shouldn't be an issue with these piezo stages within reason, they are designed to position heavy objectives, and the XYR stage I'm using isn't heaver than some Mitutoyo objectives. These are the small 40mm types and not the bigger 60mm type XYR stages. Of course there will be a offset bias in position, but should remain somewhat constant.

With heavy loads you might experience some asymmetry in movement range. Some similar piezo devices on Physiks Instrumente website might have similar characteristics I would think.

I have noticed something similar to backlash tho, this is reading the position by the strain gauge using the 16 bit ADC on the custom controller. This is a pretty sensitive sensor and converter and I've created a RPi routine to utilize this (you will get this soon

), if you place your finger on the stage a press with about as much force as to massage a beetle you can see this effect. The effect is to not return to the exact same position reading when coming from far away, the controller is issuing the exact same voltage, so it' something with the stage. I'll dig more into this when I get everything hooked up and working, which hopefully will be soon!!

Best,