Piezoelectric focuser on eBay

[mawyatt]

The Printed Circuit Boards arrived, as did the components from DigiKey. Still waiting on most of the other parts from various sources though.

Here' what the board looks like, turned out really nice IMO.

Best,

[Smokedaddy]

Sweet ...

[mawyatt]

The 1st test PCB has been assembled. Preliminary testing along the way shows the DC to DC Converter works as anticipated, producing 90~200VDC from 12VDC as set by control, the High Voltage Amplifier is completely stable and comfortable driving a high capacitive load (Piezo Element) to 0 to 60~187VDC as set by control (gain varied from ~15 to ~45). I've set the DC to DC converter to 180VDC and the HV amplifier to produce a maximum output of 163.8V with a maximum input of 4.095 volts (gain = 40). Anticipated Piezo Stage position resolution is ~25 nanometers, if the piezo elements are somewhat reasonably linear (will soon find out!). Without any gears or other mechanical contraptions (these Piezo Element Stages utilize flexure mechanisms without any backlash or mechanical noise), so this seems reasonable.

The voltage reference is set to 4.095V and the 2 HV indicators work, as does the HV output and Hi output current. Also checked out the reversing Output Circuit for bidirectional Piezo movement from a single supply voltage, it works as designed (this circuit doubles the effective range of the piezo stage by reversing the element connections, but doesn't effect resolution and only requires a single positive supply voltage, no negative supply required).

Haven't checked the 16 bit Analog to Digital converter yet for reading Strain Guage position feedback, that will come a little later.

No problems encountered other than an open resistor (cheap component) and reversed mount transistor, my bad

As you can see this is quite an involved circuit/system, hopefully rivaling OEM systems (that cost a fortune) in performance, but at a modest cost for us Macro enthusiasts

Now comes the Raspberry Pi interfacing and beginning writing the testing code in Python.

Here's a couple images of the 1st test boards, looks nice I think

Best,

[viktor j nilsson]

That looks beautiful! I'm picking up my piezoelectric stage at the post office tomorrow. Very cool that you were able to find a way to double the travel.

[mawyatt]

The 1st piezo test controller is now hooked up to the RPi and I've checked the interface addresses, connections and such. The High Voltage & Output loading resistors (low inductance 2W Metal Oxide) were changed (need to order the new values later and pay the darned shipping cost!), in the meantime just used two series 2W resistors to get the desired value. Didn't like the excess heat and decided to cut this in half, thus the change.

After resistor change went to created a simple test routine to drive the 12 Bit DAC and HV Amp output. Here's the 1st result showing a zero, 1/2 scale, full scale, and return to zero followed by a ramp up to full scale and ramp down back to zero. Note the Vertical Scale of 23 Volts per Division, the controller is set to a maximum output of 163.8V at Full Scale.

Now to get the ADC setup and operating, then a few more tests and such are in order and then the Piezo element will be connected.

Best,

[mawyatt]

Things are looking good at "Mike's Labs"

Decided to go ahead and hook up the Piezo Stage and created this file to drive the stage. Zero (0V), to 1/2 (82V), to full (164V), to 3/4 (123V), to 1/4 (41V), to zero, then ramp up (164V) and down (0V). Hooked up the Piezo Stage and it's working as planned, you can feel the movement, especially the large steps. Placed the stage on a sounding box and you can hear the movements.

Need to order some parts before we leave for a couple days, then when we get back start on the 16 Bit ADC to read the Piezo Stage Strain Guage.

Best,



Edit. Hooked up the Piezo Stage stain guage bridge interface and used a DVM (Fluke 87 with 0.1mv resolution) as a differential meter (have not tackled the ADC yet). The resolution isn't there but you can watch the DVM ramp up and down a few millivolts as the Stage is driven by the controller, I introduced a small delay between ramp steps so the DVM could respond. This confirms the operation of the stage and strain guage, very cool indeed

[mawyatt]

The 16 bit ADC with Fully Differential Programmable Gain multiplexed inputs hooked up and working, and boy does it work nice!! The differential input really helps with common mode noise, as does the input low pass filter. I’ve connected the strain gauge up with the bridge drive circuits and to the ADC as a differential signal. Long story short this works beautifully and you can read the strain gauge differential bridge output and see when you press on the stage.

I’m not very patient, so had to hook up the Piezo Stage and run thru a quick test and readout the position with the ADC. The initial result are in and it looks pretty good IMO for a just a quick test, here’s a plot of ADC differential readout voltage X axis in microvolts (yep microvolts, the ADC resolution is ~7.8uV, but this is with a 2.048V common mode voltage on the inputs, and the total range is only ~10mv!!)) and the DAC position command from 0 to 4096 in 256 steps (12 bits from 4.096 reference, the HV amplifier boost this by precisely 40X, so 0 to 163.84VDC at the Piezo Elements). Looks like what you might expect with a slight tilted “S” shape. The Piezo elements do drift some, so may need to “close the loop” with the ADC as reading the strain gauge bridge as position feedback, but the “hooks” are all there by design for doing this!!

Some basic code was written on the RPi and have the DAC and ADC both working to use for initial setup, debugging and testing.

Additional parts went on order this afternoon, this Piezo Driver uses a feedback compensation technique where the output capacitor ESR is part of the compensation network. I ordered additional special HV capacitors for this, the ones I have, the ESR is higher than I want, and since I was ordering stuff I went ahead and ordered larger HV decoupling capacitors than I have on hand to help reduce the DC to DC converter ripple. Used DHL to expedite, so should be here in a week, about when we will get back from a short 4~5 day trip. I'll spend more time on this in about a week.

Best,

[mawyatt]

Update, here's some images a of the controllers built up and ready for additional testing before shipping to users.

You can see the control pot and dial for manual positioning, and the image with added notes for connectors, jumpers and modules locations.

Soon these will be in the capable hands of others and we will see more testing, evaluation and characterization than I can do alone.

Best,

[Adalbert]

Hi Mike,
Excellent job, highly respectful,
ADi

[mawyatt]

Hi Mike,
Excellent job, highly respectful,
ADi

Hi ADi,

Thanks.

These Piezo element devices are very interesting and operate on a totally different principle than our stepper motors and rails. A ceramic element (actually many elements) changes dimensions slightly based upon the electric field across the element electrodes. These elements a placed within a stainless steel structure where they are in high compression. The structure has no bearings, nor gears or any other interface except with the elements. The structure actually "flexes" as the ceramic elements change size with applied voltage even though it's a solid chuck of SS. It has laser cuts that control the way the SS flexing behaves and this translates to the linear stage motion.

Anyway, to control the elements requires a very high voltage controller which we've developed here, much different than the lower voltage higher current motor controller for our focus rails. However the Piezo stages have a very limited range, in this case ~100um. They also have strain gauges in a bridge configuration, so you can measure the actual ceramic element induced strain and movement of the stage. We've included the ability to drive and accurately measure the strain gauge bridge to determine stage movement, allowing the possibility of semi-closed loop operation, the hardware "hooks" are designed into the controller.

Interesting devices indeed!!

I'm anxious to get these controllers to others so they can get involved and expand the testing and evaluation effort. I'm limited in my time, $, and only one person, so it's great to have others get involved.

Best,

[Smokedaddy]

I was reading about strain gauges this morning since I know nothing about them (not that this is a Physik).

https://www.omega.co.uk/prodinfo/StrainGauges.html

-JW:

[mawyatt]

The Physik Piezo Stages we have include a pair of complementary dual strain gauges located on the flexure section of the solid SS block. The four gauges are arranged in a bridge configuration to sense the stage movements due to flexing but reject any common mode movements due to vibration and whatnot.

The custom controller we've developed includes the ability drive the strain gauge bridge with a precise stable reference (4.095 Volts) and sense the differential output by way of a filtered 16 bit ADC linked to the Raspberry Pi interface. The differential signal is very small only reaching about 10mv full scale but has a common mode voltage of ~2.048V. This requires high precision instrument grade level sensing to evaluate this tiny differential signal accurately. This is all included in the controller and the software to sense the strain gauge output has also been developed and is fully operational. I use this to monitor the movement of the stage when excited by the controller. I've also checked the accuracy within the limits of my house lab capability.

Best,

[mawyatt]

Here's a few videos showing the controller and later with the Piezo stage using strain guage as a means to show piezo stage movement. Hopefully this will help folks with these controllers get setup and operational.

Best,

https://drive.google.com/file/d/16aWtFV ... sp=sharing

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

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

https://drive.google.com/file/d/1Ux9pkI ... sp=sharing

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

[Smokedaddy]

Excellent video(s) Mike including the narration.

[Lou Jost]

Just noticed the Hewlett Packard programmable calculator! Takes me wayyyy back.....