eBay item n. 254282702002.
Quite cheap, more than 10 available, and 100 microns of travel should make it quite useful for stacking with microscope subjects. Can directly take a 25 mm microscope objective.
I have no connections with the seller.
Piezoelectric focuser on eBay
Moderators: rjlittlefield, ChrisR, Chris S., Pau
-
enricosavazzi
- Posts: 1475
- Joined: Sat Nov 21, 2009 2:41 pm
- Location: Västerås, Sweden
- Contact:
-
viktor j nilsson
- Posts: 423
- Joined: Fri Mar 01, 2013 1:43 am
- Location: Lund, Sweden
-
enricosavazzi
- Posts: 1475
- Joined: Sat Nov 21, 2009 2:41 pm
- Location: Västerås, Sweden
- Contact:
I probably would if it were in the EU, but I am not traveling to the US until next year, so I will probably pass up on this item.viktor j nilsson wrote:That is very cool. I was just reading about these an hour ago.
Are you going to order one? At that price, I'd be tempted to pick one up for fun. Would be nice to split the shipping & import costs to Sweden.
This piezoelectric focuser requires a programmable voltage supply (I think) with low jitter and unknown output voltage, but once the needed voltage range and linearity are sorted out it should be relatively simple to build an Arduino-based stepper to control it. Also, one should not load the focuser with the weight of a microscope tube and camera, but separate the latter from the moving part of the focuser with a light-tight sleeve (a ring of velvet or rubber foam should do, since the focus travel is only 0.1 mm).
I retired recently (which means plenty of time but not much money) and no longer have a garage/workshop, so I am trying to concentrate on purchasing more-or-less ready-to-use items, as well as using what I already have.
--ES
-
viktor j nilsson
- Posts: 423
- Joined: Fri Mar 01, 2013 1:43 am
- Location: Lund, Sweden
I understand.
From the ad, it seems like~150v is needed for full actuation. So you would need the Arduino to control a high voltage relay. Probably doable, but I'm a little intimidated to work with high voltages.
I guess an alternative would be to find a surplus piezo controller, but that would still probably cost quite a bit.
Still a little tempting to hoard one for future projects.
From the ad, it seems like~150v is needed for full actuation. So you would need the Arduino to control a high voltage relay. Probably doable, but I'm a little intimidated to work with high voltages.
I guess an alternative would be to find a surplus piezo controller, but that would still probably cost quite a bit.
Still a little tempting to hoard one for future projects.
These look very interesting. I'm not sure how to mount them for the kind of normal setups I use, might be useful to play around with tho. I'm thinking of maybe using this as the mount for the subject rather than the lens assembly.
The ~150V should not be difficult and not a concern for safety if you design for such. Since the charge q for piezo devices is not required to be changed fast in our applications, so the current (dq/dt) is low and can be designed to be so low as not a safety issue.
A controller could be a simple DC to DC converter to up convert to the required high voltage at very low current, with a followup linear regulator controlled from a low speed DAC which is interfaced to the controlling computer.
Anyway, with enough folks involved might be a good group project to consider.
Best,
The ~150V should not be difficult and not a concern for safety if you design for such. Since the charge q for piezo devices is not required to be changed fast in our applications, so the current (dq/dt) is low and can be designed to be so low as not a safety issue.
A controller could be a simple DC to DC converter to up convert to the required high voltage at very low current, with a followup linear regulator controlled from a low speed DAC which is interfaced to the controlling computer.
Anyway, with enough folks involved might be a good group project to consider.
Best,
Research is like a treasure hunt, you don't know where to look or what you'll find!
~Mike
~Mike
-
enricosavazzi
- Posts: 1475
- Joined: Sat Nov 21, 2009 2:41 pm
- Location: Västerås, Sweden
- Contact:
A piezoeectric actuator is basically an analog device, so there are no intrinsic discrete steps. The relationship between applied voltage and displacement is not necessarily linear, but it is likely approximately linear within a certain interval.Adalbert wrote:Hi Enrico;
Does the step-size depend on the voltage or is constant = 0.001 inches ?
BR, ADi
The frame of the actuator uses flexure points, so there is no intrinsic play when reversing direction. The joints between the piezoelectric stacks and the frame, however, use rotating axes so there may be some play there. As long as the device is properly loaded (by gravity with the optical axis vertical), there should be no significant play. If you use it with the optical axis horizontal you will need to load the objective carrier, e.g. with a spring that pulls it toward the subject.
--ES
These devices have a strain gauge for use with feedback systems, so the possibility of "calibrating" the system movement to control voltage or digital word is available, or even implementing a classical feedback system, all if the strain gauge factors are known.
Strain Gauges (SG) are reasonable linear compared to the piezo elements when operating within their design range. So assuming the SG is within the design range they become the reference. A good approximation for calibrating the piezo would be to start with zero volts, read the strain gauge, run the piezo to a known 100um position and read the strain gauge again. Assume a linear relationship for the strain gauge places a simple straight line transfer curve between the two points. You can reverse the piezo voltage a get another SG reading at -100um.
A simple but accurate method to move the +-100um, assuming you have a good focus rail handy (Wemacro or THK for example), would be to use a high magnification lens (20X for example), use a good test target (silicon wafer for example). Get the 1st SG reading a zero position in focus (piezo voltage is zero), then move rail precisely 100um forward, use the piezo control to bring the image back in focus zero position and record. The piezo will be moving the lens (or subject) back to the zero position, thus requiring a -100um movement, record the SG & piezo readings. Now do the same in reverse moving the rail to -100um (make sure to compensate for rail backlash) and using the piezo to move back into focus and record the SG & piezo readings. You now have 3 SG readings for a +-100um movement. With the setup still in place you can make a few more measurements to fill in and will likely resemble a shallow, tilted "S" curve for the piezo.
You can now use this data for feedback control, or open loop piezo positioning.
Getting accurate high resolution SG readings may require a nulling bridge setup and amplifier to improve the resolution. Reading with a DVM on ohms scale will probably introduce large errors, unless you have a high quality DVM.
For the piezo HV single polarity supply, the cheap DC to DC converters on eBay may be slightly modified to produce and actually act as the HV converter and regulator together if they are designed the way I think they are (for lowest possible cost). This would produce an extremely simple yet accurate unipolarity controller for under $5!!
For bidirectional operation a couple of options are possible, the simplest is a relay which reverses the piezo polarity, the controller remains the same. The other option is to use dual converters, one positive and one negative. The negative converter might be easily modified also, but maybe not as simple as the positive converter. Think the relay is probably the simplest approach though.
Best,
Strain Gauges (SG) are reasonable linear compared to the piezo elements when operating within their design range. So assuming the SG is within the design range they become the reference. A good approximation for calibrating the piezo would be to start with zero volts, read the strain gauge, run the piezo to a known 100um position and read the strain gauge again. Assume a linear relationship for the strain gauge places a simple straight line transfer curve between the two points. You can reverse the piezo voltage a get another SG reading at -100um.
A simple but accurate method to move the +-100um, assuming you have a good focus rail handy (Wemacro or THK for example), would be to use a high magnification lens (20X for example), use a good test target (silicon wafer for example). Get the 1st SG reading a zero position in focus (piezo voltage is zero), then move rail precisely 100um forward, use the piezo control to bring the image back in focus zero position and record. The piezo will be moving the lens (or subject) back to the zero position, thus requiring a -100um movement, record the SG & piezo readings. Now do the same in reverse moving the rail to -100um (make sure to compensate for rail backlash) and using the piezo to move back into focus and record the SG & piezo readings. You now have 3 SG readings for a +-100um movement. With the setup still in place you can make a few more measurements to fill in and will likely resemble a shallow, tilted "S" curve for the piezo.
You can now use this data for feedback control, or open loop piezo positioning.
Getting accurate high resolution SG readings may require a nulling bridge setup and amplifier to improve the resolution. Reading with a DVM on ohms scale will probably introduce large errors, unless you have a high quality DVM.
For the piezo HV single polarity supply, the cheap DC to DC converters on eBay may be slightly modified to produce and actually act as the HV converter and regulator together if they are designed the way I think they are (for lowest possible cost). This would produce an extremely simple yet accurate unipolarity controller for under $5!!
For bidirectional operation a couple of options are possible, the simplest is a relay which reverses the piezo polarity, the controller remains the same. The other option is to use dual converters, one positive and one negative. The negative converter might be easily modified also, but maybe not as simple as the positive converter. Think the relay is probably the simplest approach though.
Best,
Research is like a treasure hunt, you don't know where to look or what you'll find!
~Mike
~Mike
ES, Viktor,
ES, thanks for pointing out these piezo drive focusers out. They should produce almost no vibration and continuous resolution positioning.
I've ordered the piezo drive and created a couple of circuit designs interfacing directly with the RPi without additional power supplies and such.
Totally self contained controller running from the nominal motor supply used for the stepper motors (9~24 Volts) with 12 bit resolution (4095 steps) on the piezo control voltage. This should be enough resolution to help cover the envisioned piezo non-linear S curve with a straight forward look up table to linearize the actual position vs. command position.
The concept involves modifying an inexpensive small DC to DC converter to up convert the motor supply voltage and use negative feedback techniques (either within the converter or by way of a post high voltage linear regulator) to precisely control the piezo voltage with 12 bit resolution derived from a 12 bit DAC directly interfaced to the Raspberry Pi.
I may roll a custom PCB if enough folks want to participate.
Best,
ES, thanks for pointing out these piezo drive focusers out. They should produce almost no vibration and continuous resolution positioning.
I've ordered the piezo drive and created a couple of circuit designs interfacing directly with the RPi without additional power supplies and such.
Totally self contained controller running from the nominal motor supply used for the stepper motors (9~24 Volts) with 12 bit resolution (4095 steps) on the piezo control voltage. This should be enough resolution to help cover the envisioned piezo non-linear S curve with a straight forward look up table to linearize the actual position vs. command position.
The concept involves modifying an inexpensive small DC to DC converter to up convert the motor supply voltage and use negative feedback techniques (either within the converter or by way of a post high voltage linear regulator) to precisely control the piezo voltage with 12 bit resolution derived from a 12 bit DAC directly interfaced to the Raspberry Pi.
I may roll a custom PCB if enough folks want to participate.
Best,
Research is like a treasure hunt, you don't know where to look or what you'll find!
~Mike
~Mike
-
Smokedaddy
- Posts: 1965
- Joined: Sat Oct 07, 2006 10:16 am
- Location: Bigfork, Montana
- Contact:
-
viktor j nilsson
- Posts: 423
- Joined: Fri Mar 01, 2013 1:43 am
- Location: Lund, Sweden
If enough folks sign up I'll design the PCB and get them made, this takes ~3 weeks to get back.
Well, looks like we've got enough for critical mass
I'll do all the electronics design and PCB design, but may need help with some of the mechanical adapters.
Best,
Well, looks like we've got enough for critical mass
I'll do all the electronics design and PCB design, but may need help with some of the mechanical adapters.
Best,
Research is like a treasure hunt, you don't know where to look or what you'll find!
~Mike
~Mike
-
ray_parkhurst
- Posts: 3438
- Joined: Sat Nov 20, 2010 10:40 am
- Location: Santa Clara, CA, USA
- Contact:
The piezo load should be high impedance, so little current needed, but if you added a follower/buffer with some current drive capability, the system you're describing would be very similar to the VCM controller built by Peter Lin. I talked with Peter about possibly creating a version of his SnS system which could drive a VCM instead of stepper for the Z-axis focus stepping, but he declined. I'd guess that the system you're describing could do this with the appropriate DC buffer amp. I wonder if there are any major issues with this? I'm thinking that this capability may have just won you another customer...