I'm working on a project that requires charging and discharging a piezoactuator (150V, 1.5μF) extremely fast, IE total slew time allowed ~50μs (so somewhere around a 11A pulse). My background is in materials science, so after a lot of stumbling around in the dark I've wound up with the following circuit based on what I have available for now: enter image description here

I forgot to relabel X4-A and X5-A, they are STGP19NC60HD IGBTs, and Vcc/Vee for TL082 is +/- 18V.

Essentially a function generator provides a +/- 10V square wave, where at +10V optocoupler 1 (OC1) turns on and subsequently turns on IGBT X4-A, which allows the piezo to charge, as IGBT X5-A is off. When the square wave inverts to -10V OC1 and X4-A turn off, TL082 inverts the -10V, turns on OC2 and therefore turns on X5-A to discharge the piezo. While I know there are a lot of things with the circuit that probably don't make sense (there are some artifacts left over from a few other inspiration sources), this circuit slews the piezo at the rate set by the RC constant set by R7, as by (mediocre) design. The only problem is the piezo never reaches the voltage that is expected; 150V-voltage drop across IGBT X4-A (~few volts). Instead the voltage applied to the piezo is limited to, and varies directly with the gate voltage (~X4-A Vge - 2V).

I've read through a few other SE posts that seem to have similar problems, but none have passive elements downhill from the emitter, ie the emitter seems to always run straight to ground. I know that there is something very fundamental I'm not understanding about the IGBT's operation but I can't seem to figure it out or find a working circuit to modify. It would be great if there was some tweak to fix the voltage issue, but I wouldn't be surprised if this thing has to be completely scrapped and started from scratch.

  • \$\begingroup\$ I would scrap it and go with a MOSFET drivers which have dead-time control so the IGBT FET gates are not crossing over shorting the supply aka shoot thru. Emitter followers here don’t have enough input current thus inadequate output current to charge gate capacitance, Ciss and fall time may be excessive. \$\endgroup\$ – Tony Stewart Sunnyskyguy EE75 Aug 9 '18 at 22:20
  • \$\begingroup\$ Your output will never be higher than X4-A's Vg-2V, but you could see an improvement by taking the 'bottom' end of R2 off GND & connecting it to X4-A's emitter instead (yes the output node). As you have it now produces a voltage-divider effect between R9 & R2 resulting in Vg on X4-A never being higher than about 130V. As Tony has already noted - proper (high-side) drivers would fix this too. \$\endgroup\$ – brhans Aug 9 '18 at 23:06
  • \$\begingroup\$ Also, your positioning of R7 in the GND leg of X5-A means that you're only controlling the slew rate on the falling edge. If controlled slew rate is important to you then you should either put R7 in series with the piezo, or add another resistor for X4-A so you control the rising edge too. \$\endgroup\$ – brhans Aug 9 '18 at 23:12
  • \$\begingroup\$ R7 was one of the first things I realized didn't make sense after looking at the schematic, I added another for X4-A to spread out the heat dissipation. Would a proper high-side driver look something like this: analog.com/en/products/monitor-control-protection/… \$\endgroup\$ – Marin Aug 9 '18 at 23:33

Your Answer

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

Browse other questions tagged or ask your own question.