Piezoelectric focuser on eBay

Have questions about the equipment used for macro- or micro- photography? Post those questions in this forum.

Moderators: rjlittlefield, ChrisR, Chris S., Pau

mawyatt
Posts: 2497
Joined: Thu Aug 22, 2013 6:54 pm
Location: Clearwater, Florida

Post by mawyatt »

Viktor,

I've just done a quick simple test using a 50X lens with a Nikon D850 at 16X in Live View Video mode with the setup shown with Trinamic controlling a THK KR20 rail with 400 step motor. This arrangement produces an effective magnification of 800X and viewed on a small 200mm HDMI monitor.

Used a resistor lead end, then tip of needle as subject and commanded the Piezo Controller over various ranges and by refocusing the subject using the focus rail I could determine the subject movement introduced by the Piezo Stage.

I've found the piezo stage doesn't work well when the control is reversed. After some tests and watching how the stage behaves I've thought about this and I believe this is what's happening. The long thin piezo element stack is wedged into the solid SS fixture walls under a high compression load during assembly, effectively squeezed between the stage walls. This is the only means of attachment to transfer force to the stage. With the applied electric field in the normal mode (positive) the ceramic piezo elements expand in length and "push" against the stage walls, thus inducing movement. When the electric field is reversed the ceramic elements contract in length, but the stage movement is limited to where the elements are no longer in contact with the walls. The elements will continue to contract with larger reversed fields but the stage will not respond since the elements are no longer in contact with the stage walls. You can hear the slight "click" sound when the element disengages with the wall and a louder "click" when it reengages when the field is removed. There is some slight movement other than along the preferred axis when the elements disengage and reengage.

The bottom line is these Physik Piezo Stages don't respond well below about 30um reversed. However they do respond nicely in the forward direction to about 400um and the linearity seems good over this range. The piezo stage and controller have a scale factor of about 0.1um per bit.

The orientation I've shown (arrow up) has the stage moving upwards with positive applied electric field.

Anyway, hope this helps some folks with these interesting Physik Piezo Stages.

Best,
Research is like a treasure hunt, you don't know where to look or what you'll find!
~Mike

viktor j nilsson
Posts: 423
Joined: Fri Mar 01, 2013 1:43 am
Location: Lund, Sweden

Post by viktor j nilsson »

Mike, thanks for the update! I understand if you are a little disappointed that the travel was limited in reverse mode, but that's hardly a deal breaker for me. But hey, did you just say 400um travel in normal mode? Where did that come from, I only remember hearing about 100um travel in normal mode? That is really good news, if that's not a typo.

I checked the spec sheet for the PI P601 actuator (which I assume is the same piezo element as in the eBay unit) on PI's homepage. It specifies the "Push/pull force capacity in motion direction" as between 30/10N and 15/10N depending on the version:

https://www.physikinstrumente.com/en/pr ... tor-202600

If I interpret this correctly, one probably shouldn't put more than 1 kg or so on it.



By the way, I am now fairly certain that these units come from an Illumina sequencer, just like the Nikon CFI Plan Apo 20x 0.75 VC objectives that has recently flooded the market.

https://www.ebay.com/itm/PI-Physik-Inst ... 2659080368

The eBay ad above shows an identical unit said to come from an Illumina sequencer.

And one commenter in this blog post:

https://blogs.swarthmore.edu/Illumina+G ... own/?p=125

has taken apart an Illumina Hiseq 2000 and found a "piezo actuator for the Z-Stage (Physik Instrumente P-601 with driver E-601 and E-801 Sensor module) with mounted Nikon CFI Plan APO VC 20x Objective."

Although I haven't found a picture of the Hiseq 2000 piezo unit in situ, I believe that it's this model as the pictures I've seen of the GAIIX model looks quite different.

Seems like there will be quite a few of these on the market over the coming years. Will be fun to put them to good use!

mawyatt
Posts: 2497
Joined: Thu Aug 22, 2013 6:54 pm
Location: Clearwater, Florida

Post by mawyatt »

Viktor,

I was disappointed and surprised at the same time :shock:

The range is indeed ~400um in the + direction with 163.8 volts as the maximum applied voltage (the controller can go higher but I've limited myself to this level), but only ~30um in reverse - direction until you begin to see the effects I described above. These are also quite good from a linearity standpoint from the limited tests I've done, even at 400um movement range.

The controllers work beautifully with these Physik Stages, and the 16 bit strain gauge readout is an added plus :D

My feeling is the quality stepper motor rotor/stator cog positions are reasonably accurate and the THK KR20 1mm pitch rails are also reasonably accurate. So using a high magnification lens to estimate the position by means of in or out of focus seems reasonable. I set the piezo controller, then move the Z axis THK KR20 rail with D850 & Mitutoyo 50X to refocus the tip of a pin or other object and record the amount of movement required by the rail as the same as the Piezo element movement under the piezo controller command. Wish I had a better environment to evaluate the position accuracy, but this is the best I can do since retiring.

That does look like the same Physik stage, but for $200 + shipping!!

I believe these Phyisk and similar piezo stages may prove very useful for high magnification work, especially for those that don't have an industrial precision grade rail like the THK types. Using a good standard rail like the Wemacro, Stackshot or MJKZZ rail as "coarse" and the Physik Piezo as the fine stage for stacking, this now seems quite reasonable and inexpensive. Of course there are many other uses I'm sure where these will be employed.

Looking forward to you and others continuing with the testing and providing valuable inputs.

Best,
Research is like a treasure hunt, you don't know where to look or what you'll find!
~Mike

mawyatt
Posts: 2497
Joined: Thu Aug 22, 2013 6:54 pm
Location: Clearwater, Florida

Post by mawyatt »

Just completed a couple initial videos showing the custom controller running a routine to drive the PI piezo stage. The stage is driven in 1/4 full range steps from start to full and back, then ramped up and down in and sequentially slower ramps.

The PI piezo stage is mounted to a Wemaco vertical stand base with an ARCA Clamp, on top of the PI stage is a small 40mm XY&R precision positioner with a small ball head. The subject is a bend pin tip, viewed at 10X.

If you watch the 2nd video carefully you can see the pin tip move up & down and follow the patterns as mentioned and shown on the oscilloscope trace in the 1st video. When viewing the pin tip thru the 10X on the small HDMI screen, note the smoothness in the movements.

I may try and get the Voice Coil Motor controller setup and operating since I have a couple speakers now.

Best,

https://drive.google.com/file/d/1_ivuW7 ... sp=sharing

https://drive.google.com/file/d/1pnRIFX ... sp=sharing
Research is like a treasure hunt, you don't know where to look or what you'll find!
~Mike

mawyatt
Posts: 2497
Joined: Thu Aug 22, 2013 6:54 pm
Location: Clearwater, Florida

Post by mawyatt »

I've been able to do a little more with the Piezo stages, controller, and get more information for the folks experimenting with these stages.

The PI Stages discussed so far need to limited in applied voltage range to +130 and -30VDC, they are bi-directional somewhat but highly asymmetrical. So stay within those limits or you are likely to damage the PI Stage.

With the custom controller mentioned I've been able to operate the PI stage in a "Somewhat" negative feedback "Close loop" fashion utilizing the PI Stage Strain Gauge and the 16 ADC & PGA with RPi. "Closed loop" meaning driving the PI stage to a commanded position, sensing such position and provide error correction to force to that position, and repeating (looping). "Somewhat" meaning following the "Close Loop" algorithm until an acceptable position error is sensed by means of the Strain Gauge then releasing the control loop and moving onto something else like waiting for another position command.

Another member researching these piezo stages has implemented such a loop as well with the Arduino, but with a different stage and utilizing an OEM driver.

We now have the Physik Instrumente E-610.S0 main driver PCB and E-801B and E-802 plug-in boards (from eBay supplier of the PI Stages), these are likely the OEM diver for the PI Stages discussed here. After a brief study of the data sheets and hardware these are extremely complex completely analog control boards, no digital control within the driver or feedback systems. It appears that over 15 precision op-amps are utilized on 3 separate PCBs including a precision differential hi gain sense amplifier, dual "notch filters", PID control, slew rate limiting, driver core and so on. This is a continuous time, continous amplitude, pure analog control loop requiring no computer or digital interaction. The system responds to an analog command voltage, which could be from a DAC, a power supply, or a simple potentiometer. The control transfer function would probably take an entire page to write with no closed form solution (requires numerical methods) in closed loop configuration, so it's going to take some time to digest this!!.

Evidently these PI Stages are capable of 5nm resolution and +-10nm repeatability when operated with the proper control algorithms and controller. Working on a very advanced controller that should be able to approach these levels of precision but without the expense and complexity of the OEM solution, and work directly from the RPi. To make this clear, we're not coping the OEM controller, and don't copy the work of others unless it's common knowledge, available for general public use and always note the source. What we are doing is looking at the OEM design approach and how it's implemented and using this to improve the knowledge base on the subject, which allows better decisions in the upcoming custom design phase. The original custom controller presented here was completely without knowledge of the OEM controller and it's interesting that a completely different approach to driving the PI Stage was chosen, but similar approach to allow external manual control of such. The OEM controller utilizes a more conventional inherently low-impedance voltage mode output stage (complementary NMOS & PMOS source followers) that is specifically stabilized with the highly capacitive loads and must be selectively current and slew rate limited. The approach for the custom controller mentioned here is a complementary bipolar (NPN, PNP) transconductance stage with high output impedance that converts an input voltage to an output current, this output current is "forced" thru the higher output impedance to achieve voltage gain at high voltages. This technique is often utilized in advanced analog chip design to drive capacitive loads and is inherently stable with large capacitive loads, modern LDO regulators are a prime example of such. The OEM controller designers may not have been aware of this design technique, or had other reasons for their design approach (speed may have been a factor, but we're not concerned about speed for macro use), but their approach certainly is working quite well!

So the task ahead is to create a custom stable Closed Loop controller that can drive the PI Stage to these incredible levels of precision, or at least approach this level, but without the high complexity and cost of the OEM controllers.

Moving ahead :D

Best,
Research is like a treasure hunt, you don't know where to look or what you'll find!
~Mike

Smokedaddy
Posts: 1965
Joined: Sat Oct 07, 2006 10:16 am
Location: Bigfork, Montana
Contact:

Post by Smokedaddy »

Does anyone have a safe guess what sort of weight one of these Physik(s) is able to handle with no issues?

-JW:

enricosavazzi
Posts: 1475
Joined: Sat Nov 21, 2009 2:41 pm
Location: Västerås, Sweden
Contact:

Post by enricosavazzi »

Smokedaddy wrote:Does anyone have a safe guess what sort of weight one of these Physik(s) is able to handle with no issues?

-JW:
The question is relevant, but I think it should be put into context. I don't think that the best use of these piezo stages is to carry the weight of tube lens, extension tubes and camera. I believe that they should just carry their own weight and the objective.

I could envision a system where there is a gap of about 1-2 mm above the stage, between the stage and the rest of the equipment. If necessary, this gap can be made light-tight with a gasket of black foam or a similar elastic material. Even with the full travel of about 0.4 mm, the change in distance between objective and tube lens (or the change in distance between objective and sensor in a finite system) should introduce no detectable drop in optical performance.

The flexure system does introduce a sideways shift that accompanies the vertical travel, but this should be a small fraction of the vertical travel. I don't think it will be a problem except perhaps with deep stacks at very high magnification, but will leave it to others to comment to this.
--ES

Smokedaddy
Posts: 1965
Joined: Sat Oct 07, 2006 10:16 am
Location: Bigfork, Montana
Contact:

Post by Smokedaddy »

enricosavazzi wrote:
Smokedaddy wrote:Does anyone have a safe guess what sort of weight one of these Physik(s) is able to handle with no issues?-JW:
The question is relevant, but I think it should be put into context. I don't think that the best use of these piezo stages is to carry the weight of tube lens, extension tubes and camera. I believe that they should just carry their own weight and the objective.
I was thinking of using it as a holding platform for a specimen with my MM-11 EPI setup. That would be pretty easy to make. However I would love to be able to use it with transmitted light for diatoms and the reason for my question. Adapters could be easily made to mount an objective, some sort of super light weight one off bellows, or maybe a 3D printed PLA tube, and my 50D but I haven't a clue if it could handle that. I'm guessing no.

-JW:

PS, probably could only handle a Raspberry Pi camera. :roll:

enricosavazzi
Posts: 1475
Joined: Sat Nov 21, 2009 2:41 pm
Location: Västerås, Sweden
Contact:

Post by enricosavazzi »

Smokedaddy wrote: I was thinking of using it as a holding platform for a specimen with my MM-11 EPI setup. That would be pretty easy to make. However I would love to be able to use it with transmitted light for diatoms and the reason for my question. Adapters could be easily made to mount an objective, some sort of super light weight one off bellows, or maybe a 3D printed PLA tube, and my 50D but I haven't a clue if it could handle that. I'm guessing no.

-JW:

PS, probably could only handle a Raspberry Pi camera. :roll:
Assuming one mounts the stage on a fixed column, a support for the camera, extension tubes and tube lens can be attached to the same column. A 1 mm gap between the stage and the camera support is all it takes to allow free movement of the stage, and in this way you have no problems with reasonably heavy cameras.

Using the stage as a specimen carrier should also work fine, even if you use a normal microscope stage resting atop the stage. A normal microscope specimen table is probably no heavier than half a Kg.
--ES

mawyatt
Posts: 2497
Joined: Thu Aug 22, 2013 6:54 pm
Location: Clearwater, Florida

Post by mawyatt »

Agree, the camera, tubes and lenses might be too much.

I'm using these PI P601K devices to move the subject and have mounted a small ball head and 40mm XYR stage without issue. This doesn't weight much more than a Mitutoyo.

PI P601K adapter by JW :D

Image
Image

Best,
Last edited by mawyatt on Sun Nov 17, 2019 6:34 am, edited 1 time in total.
Research is like a treasure hunt, you don't know where to look or what you'll find!
~Mike

mawyatt
Posts: 2497
Joined: Thu Aug 22, 2013 6:54 pm
Location: Clearwater, Florida

Post by mawyatt »

Continuous closed loop piezo operation will likely be necessary with heavier loads to keep the precise position. At nanometer levels these stages have load deflection, some creep and hysteresis effects which require mitigation for long term stability, one of the reasons we are developing the continuous Closed Loop Controllers for these specific piezo stages.

Very difficult task with the tiny movements, and sensing such for precise sub-micron (actually nanometer levels) position and control. As shown these devices are highly complex, but has been investigated and now reasonably understood to allow proceeding with the developments.

The good news (really good news actually) is that the initial control algorithms and details have been implemented and preliminary test results confirm superb nanometer levels of control and performance

We have 2 different systems in operation now at Mike's Labs, one is based upon modification of an OEM driver with a custom interface and control output, the other is a unique custom design driver and controller. Both operate from a Raspberry Pi and initial test programs written to allow position control and nanometer level stepping.

We are waiting on a final set of custom PCBs and components to implement the full control concepts on both systems which include the ability to directly control and readout actual position at nanometer levels from a single custom PCB, either mated with the OEM driver or stand alone.

Encouraging results indeed :D

Contact us if interested.

Best,
Research is like a treasure hunt, you don't know where to look or what you'll find!
~Mike

Smokedaddy
Posts: 1965
Joined: Sat Oct 07, 2006 10:16 am
Location: Bigfork, Montana
Contact:

Post by Smokedaddy »

I've been experimenting with the Piezo off and on for a few weeks. Big learning curve for me but I'm trying. I'm still modifying my MM-11 for this particular setup and need to make a few other changes. I'm mounting a tilt/tip stage on it now but need to machine a few things again for the Piezo ... hopefully tomorrow. My micron gauge (shown below) isn't calibrated so I tried another plunger style gauge with .001" increments and it read a total movement of .014" (or 355.6 um) and did so consistently for the full range of movement on this particular Piezo, so I'm going with that. I also triple check these measurements with the .001" plunger gauge and kept getting the same results.

1/4 scale - 1024 - .004" or 101.6 um
1/2 scale - 2049 - .007 or 177.8 um
3/4 scale - 3072 - .009 or 228.6 um
Full scale - 4096 - .014" or 355 um
Step size - 355.6/4096 = 86 ±

Here's a link to a video I did with a different setup (different numbers too) a few days ago testing the Piezo in action. You can read the description regarding what's going on.

https://www.youtube.com/watch?v=ID_FFl_-kV8

So using my old setup (shown below) I shot an image stack for test purposes with my MM-11 EPI setup and a 10x BD Plan. I haven't done much EPI shooting for a while, so I might be forgetting something. I tend to do that nowadays. <duh> This was stacked in Zerene and is a 100% crop then cleaned up a bit in Photoshop. Oops, the image below was a 100% crop and 1600 pixels wide but I forgot to change it to 1024, so it's not shown below at 1600.

I used 430nm steps (0.43um) and 31 images were taken (should of done more). The platform I made that is attached to to the top of tje Piezo wasn't level. I need to work on that. I still need refine my skills on the 'start' focus better. Hopefully this will be useful for one of my applications. If not, no stress on my end.

Image

Image

Image

Image

-JW:

mawyatt
Posts: 2497
Joined: Thu Aug 22, 2013 6:54 pm
Location: Clearwater, Florida

Post by mawyatt »

Nice work and excellent setup!!

This is Open Loop operation and the P601K Piezo Stage seems to be working well with an expected slight curved position vs. command voltage behavior at the top end according to your position measurements.

Keep up the great work!

Best,
Research is like a treasure hunt, you don't know where to look or what you'll find!
~Mike

Smokedaddy
Posts: 1965
Joined: Sat Oct 07, 2006 10:16 am
Location: Bigfork, Montana
Contact:

Post by Smokedaddy »

Per Mike's request ... here's my graph from this evening IF I did it correctly. I plotted the entire range of 355um of my Physik with points from zero to 4096 in 256 increments to give me an idea of how straight (linear) the stage is. My total range is 355um with a 4096 count, then each count represents 355um/4096, or ~86.7 nanometers.

https://www.photomacrography.net/forum/ ... c&start=36

Image

-JW:
Last edited by Smokedaddy on Tue Nov 19, 2019 8:31 pm, edited 5 times in total.

rjlittlefield
Site Admin
Posts: 23603
Joined: Tue Aug 01, 2006 8:34 am
Location: Richland, Washington State, USA
Contact:

Post by rjlittlefield »

What's being graphed, and what are the units?

--Rik

Post Reply Previous topicNext topic