I am working with MEMS structures and need to move a device (switch) in a operating point with quite high offset. (150V DC offset and a voltage swing of 60V) To generate the high HF Voltage (up to 5MHz) we use a System WMA-300 from Falco Systems.

Do get this DC offset on this system I was inspired by this answer to use a transformer (e.g. WB1010) to offset the input voltage.

Circuit Idea

In the end I want to record a response function of the device depending on the frequency (between 1Hz ideally (but 5kHz is fine) and 5 MHz) of the driving signal (between 120V and 180V). The signal that needs to arrive at the MEMS device should look as follows (Sorry for the badly drawn chirped sine). Chirped sine at the MEMS device.

When thinking about the solution, I was mainly afraid that somebody would operate the system wrongly applying a low frequency signal to the input coil thereby burning the primary coil. (50 Ohm source impedance, and 0.3Ohm coil resistance) Do I need to prevent this by adding a High-Pass filter in front of the primary coil?

Is there a better idea how to create this signal?

There is a better way to do it (see the accepted answer), seems the transformer is overkill.

  • \$\begingroup\$ It still doesn’t tally for me. You mention MHz then you mention 300 Hz. Please state plainly what final voltage you need immaterial of the method. Regard the individual AC and DC voltages as being in series so you can list them individually. \$\endgroup\$ – Andy aka Jan 28 '20 at 17:20
  • \$\begingroup\$ This really sounds like the kind of potentially hazardous situation where you should seek out someone experienced to assist. If this is in the context of a larger research organization (like a university) there are probably many such people around. \$\endgroup\$ – Chris Stratton Jan 28 '20 at 17:48
  • \$\begingroup\$ @Andyaka I tried to clarify the question. Chris Stratton: I agree, high voltages one should be careful what one is doing. \$\endgroup\$ – gyger Jan 28 '20 at 19:09
  • \$\begingroup\$ Draw a picture. \$\endgroup\$ – Andy aka Jan 28 '20 at 19:12
  • 1
    \$\begingroup\$ @pipe The research devices often are not so optimized as the commercial devices. It depends a lot on the rigidity of the slab. \$\endgroup\$ – gyger Jan 28 '20 at 20:48

The solution depends upon how quickly you need to vary the DC bias (if at all).

Just add a resistor in series with the DC source to the MEMS then a capacitor from the AC source to the MEMS.

The value of the resistor depends upon the leakage current in the MEMS, since they typically have very low leakage this could be 100k ohm.

The capacitor needs to have a reactance that is small relative to the resistor at the lowest frequency of operation. A value of 0.01uF would be appropriate for the frequencies you mention.

  • \$\begingroup\$ I have changed the question to reflect more closely what I need. Thank you for your answer. \$\endgroup\$ – gyger Jan 28 '20 at 20:06
  • \$\begingroup\$ @gyger - Your solution is fairly good but does need a transformer. The one I proposed only needs a resistor and capacitor and relies on the MEMS not requiring any significant DC current. I once built a system with over 3,000 MEMS actuators for manipulating light from optical fibers in a communication system. \$\endgroup\$ – Kevin White Jan 29 '20 at 0:01
  • \$\begingroup\$ Sorry for the late response, I very much like your solution (especially the simplicity). If I now have a 1uF capacitor and a 100kOhm resistance that would work out well down to 1 kHz (ZC=j150Ohm). I will accept your answer. \$\endgroup\$ – gyger Jan 29 '20 at 19:40
  • \$\begingroup\$ Just to give you an update, everything worked out that way. Excellent simple solution. \$\endgroup\$ – gyger Mar 20 '20 at 17:59

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.