2
\$\begingroup\$

I'm trying to drive a Peltier cell. The driver must allow both operation modes (cold/hot). To achieve this i'm using filtered 3.3V 100KHz PWM from a microcontroller to drive an H-Bridge (as far as i've read on the internet filtering PWM signals coming out from the bridge with an LC filter is useful for efficiency problems). The Peltier cell is a 70W 12V module.

This is the H-Bridge/Filter section: H-Bridge // LC Filter

This is the IR2110, bridge driver for A and B side: IR2110 A SIDE

I'd like to know if this circuit will likely work or if there are any mistakes in components maximum ratings (like free-wheeling diodes) or other stuff that can create magic white smoke.

Feel free to ask anything.

Thank you

\$\endgroup\$
  • \$\begingroup\$ In the first ~1ms or so of the startup the current through your peltier will not yet be stable and depending on the capabilties of the power source may briefly exceed the maximum rated current. The 44N is a pretty old device made for power applications, I would check if it really likes that frequency, and if the driver can run it, it has quite some gate capacitance. \$\endgroup\$ – PlasmaHH Jul 3 '17 at 11:24
  • \$\begingroup\$ For the startup i've got a soft start condition via software (some sort of ramp in PWM duty cycle) to charge capacitors. Is it enough? I've read the application note of the IR2110 and i've found an example circuit up to 500kHz with an IRF830. With the same test conditions, IRF830 have an input capacitance of 610 pF, while IRLZ44N's capacitance is 1700pF (quite high actually), but my PWM is much slower. Actually PWM frequency not a constraint and i'll decide when I'll test the circuit. I've set 100kHz lo lower the value of the inductance into the filter. \$\endgroup\$ – ozlow.own Jul 3 '17 at 11:40
  • \$\begingroup\$ Soft start should be fine, have a look at the things going on in the actual circuit with a scope though. Higher capacitance usually means slower to turn on with the same current capabilities of the drive which means more heat dissipated by one switch, then more switching also means more heat. If you are worried about magic smoke, do some calculations. Or prototype it and then measure how hot things get. \$\endgroup\$ – PlasmaHH Jul 3 '17 at 11:44
  • \$\begingroup\$ @PlasmaHH thank you ! I was concerning about diodes and MOS mostly because the current going back to the PSU while the PWM is into low state seems huge in SPICE simulation (tiny spikes up to 30A). I think this is due to simulation itself, which doesn't include any kind of "parasitic" resistance (traces etc.). \$\endgroup\$ – ozlow.own Jul 4 '17 at 11:58

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.