EEVblog Electronics Community Forum

Electronics => Projects, Designs, and Technical Stuff => Topic started by: DenzilPenberthy on July 12, 2013, 03:03:50 pm

Title: 0 to 100v Piezo Driver ideas...
Post by: DenzilPenberthy on July 12, 2013, 03:03:50 pm
Hi everyone,

 I'm designing a simple system for driving some piezo actuators for some colleagues who are scanning surfaces under an atomic force microscope.

 A similar system I built recently was nothing more complicated than an Arduino and a DAC which can produce an output between 0-10v

 Now I've been asked to make a 2 channel (x&y) one with 16 bit resolution which can supply up to 100v, with a switchable gain so 0000 to FFFF could correspond to 0-10v, 0-20v, 0-50v or 0-100v for example.

The piezo load is 10nF and it only needs to work up to around 10 kHz which I reckon is a shade over 60mA

 If the voltage didn't have to go so high then this would be fairly straightforward to implement with standard op amps.

 High voltage op-amps are available but they're pretty expensive, the Apex PA85 here at Farnell is £172.51 each! Also they've only got 14 left and won't be restocking them.

http://www.farnell.com/datasheets/28904.pdf (http://www.farnell.com/datasheets/28904.pdf)

'Proper' analogue stuff with actual transistors is a bit of a weak point of mine so does anyone have some bright ideas how to implement something that would fit my application?

What about one of these op-amps which is good for 140v for £25 then a transistor on the output to enable it to supply more current?

http://uk.farnell.com/linear-technology/ltc6090cfe-pbf/op-amp-precision-1mhz-19v-us-330uv/dp/2334037 (http://uk.farnell.com/linear-technology/ltc6090cfe-pbf/op-amp-precision-1mhz-19v-us-330uv/dp/2334037)

All thoughts gratefully received!
Cheers.

Title: Re: 0 to 100v Piezo Driver ideas...
Post by: Marco on July 12, 2013, 03:56:23 pm
You can simply connect the load to 100 volt with a mosfet to ground, drive the mosfet with a low voltage opamp and use a divider to generate the feedback. You need to level shift and invert the input with a second opamp.

PS. you can connect the load to ground as well, but then you need one P and one N transistors.

PPS. like this http://www.digikey.com/schemeit/#f79 (http://www.digikey.com/schemeit/#f79)
Title: Re: 0 to 100v Piezo Driver ideas...
Post by: nctnico on July 12, 2013, 04:11:27 pm
Use a 0 tot 50V power supply (pretty simple to make with a step-up switcher). Feed the 50V through an inductor to the piezo and a mosfet to ground. Just pulse the mosfet for half the cycle. The inductor will take the piezo through the rest of the cycle.
Title: Re: 0 to 100v Piezo Driver ideas...
Post by: Marco on July 12, 2013, 04:16:00 pm
I don't think the piezo in question is supposed to resonate, I think they are linear actuators.
Title: Re: 0 to 100v Piezo Driver ideas...
Post by: nctnico on July 12, 2013, 04:22:51 pm
Another way is to use a step-up transformer if it needs to be linear. That will keep the opamps within a reasonable price range. TCA0372 is a nice one.
Title: Re: 0 to 100v Piezo Driver ideas...
Post by: mikeselectricstuff on July 12, 2013, 06:21:23 pm
Take a look at audio amplifiers - higher power ones will be getting close to the voltage range you want
Title: Re: 0 to 100v Piezo Driver ideas...
Post by: Marco on July 12, 2013, 09:32:36 pm
Could someone look at my schematic and comment? Pretty sure the basic concept is sound and suited to his low bandwidth needs ... of course simplified a bit, a pull up transistor on the base of Q2 will improve it's turn off speed and it will probably need a RC circuit in the feedback path to stop it from oscillating.

(http://i278.photobucket.com/albums/kk105/_MfA_/piezo_zps41107481.gif) (http://s278.photobucket.com/user/_MfA_/media/piezo_zps41107481.gif.html)
Title: Re: 0 to 100v Piezo Driver ideas...
Post by: ejeffrey on July 12, 2013, 09:53:43 pm
You really want a push-pull stage to drive a piezo, not just a passive pull-down resistor.  Mike's suggestion to look at audio power amplifiers is a good one.  You will need to check for stability driving capacitive loads since that is not normal for audio amps.
Title: Re: 0 to 100v Piezo Driver ideas...
Post by: Marco on July 12, 2013, 10:10:48 pm
Then add a push pull stage, although if you don't mind burning 10 Watt in the divider 1000 Ohm will get you a RC time of 10 microseconds which seems sufficient  for 10 kHz. The problem with most audio circuits is that the audio crowd hates heavy feedback and opamps ... very high linearity and DC amplification are not really on their agenda.

Something like this is closer to what he needs then IMO. Although with some needless complexity due to unneeded bipolar output.

(http://circuits.linear.com/img/439_circuit_1.jpg) (http://circuits.linear.com/Signal_Gain_*_Filter)
Title: Re: 0 to 100v Piezo Driver ideas...
Post by: TerminalJack505 on July 12, 2013, 10:42:17 pm
If you have a copy of The Art of Electronics, check out Figure 3.75.  It's a 0 to 1kV low-power, push-pull piezo driver.
Title: Re: 0 to 100v Piezo Driver ideas...
Post by: DenzilPenberthy on July 15, 2013, 09:55:36 am
Loads of great ideas there, thanks very much! Haven't been near a computer all weekend to check back on this thread. I'll go and find a copy of AoE and check out their circuit.

Cheers!
Title: Re: 0 to 100v Piezo Driver ideas...
Post by: krivx on July 15, 2013, 10:03:09 am
I have done this before and used the Apex HV op-amps. I would look at the costs of your actuators, depending on the piezo model they can cost considerably more than 172£/ea. I wouldn't spend time trying to come up with a cheaper solution if the budget is already there.
Title: Re: 0 to 100v Piezo Driver ideas...
Post by: Frant on July 15, 2013, 11:36:35 pm
I have done this before and used the Apex HV op-amps. I would look at the costs of your actuators, depending on the piezo model they can cost considerably more than 172£/ea. I wouldn't spend time trying to come up with a cheaper solution if the budget is already there.

Apex hybrids are often used in such systems. Piezo actuators are already so expensive (easily $10k) that a couple of $200 components is not a big deal. A state-of-the-art scanning/positioning stage (1nm resolution, 50 nm repeatability) with a closed-loop controller included can cost up to $100k, and a typical AFM configuration more than that.