0
\$\begingroup\$

I'm currently making a buck converter (rated output of 60V/10A) with IR2110 as its high side NMOS driver. The driver and buck converter circuit can be seen in fig 1 and 2. I've asked a bit on this forum for a different problem, here's a link to the post if you want to see it.

Gate driver circuit Fig 1. Gate driver circuit

Buck converter circuit Fig 2. Buck converter circuit

Currently, my IR2110 is experiencing heating issues if the buck converter's output voltage is high. Unfortunately I didn't do a systematic testing, but the buck converter is outputting 25V when the IR2110 suddenly heats up to 70 or 80°C.

I stumbled upon this application note about HV floating MOSFET driver, and it says that IR2110 can only be used if the buck converter's output voltage is in between 10 to 20V.

Warning about IR2110 in the application note Fig 3. Warning about IR2110 in application note (highlighted in blue) [page23]

But in IR2110's datasheet, it says that the floating output voltage can be up to VB (which is approximately 500V)

Recommended floating output voltage of IR2110 Fig 4. Recommended floating output voltage of IR2110 (highlighted in blue) [page 2]

So my questions would be:

  1. Can the IR2110 be used as a high side driver only (to drive a buck converter with 60V output voltage)?
  2. If not, should I change my configuration to synchronous? or should I just change the driver circuit? If your suggestion is to change the driver circuit, do you have any suggestion on which the driver I should use?
\$\endgroup\$
6
  • \$\begingroup\$ What switching frequency and duty cycle range are you running? Is your bootstrap diode fast enough? What Vboot voltage do you measure? Also 1 kohm pull-down is quite strong. Try 100 k? \$\endgroup\$
    – winny
    Jun 5, 2022 at 15:32
  • \$\begingroup\$ Huh, why is \SD labeled as active-low, but also grounded (presumably shutting the thing down)? Datasheet says this is active-high so the connection is correct but the label is wrong. \$\endgroup\$ Jun 5, 2022 at 15:49
  • \$\begingroup\$ Also, what happened to the high side supply, wasn't there an isolated supply for that? \$\endgroup\$ Jun 5, 2022 at 15:49
  • \$\begingroup\$ @winny I'm using 50kHz, and the duty cycle is approx 20%. The diode I'm using is UF4007 with a recovery time of 70ns max (from the datasheet vishay.com/docs/88755/uf4001.pdf). I'll try using 100k for the pulldown, but may I ask if changing this resistor will solve the heating issues? \$\endgroup\$
    – Kevin
    Jun 5, 2022 at 16:00
  • \$\begingroup\$ @TimWilliams Honestly, I'm not quite sure whether SD is active high or active low, but the driver is functioning correctly, with the VGS waveform being a square wave with 25% duty cycle. And about your second comment, are you referring to my 12V high side supply (VCC node in fig 1)? I'm using a Generic 12V adaptor that's not shown in the diagram \$\endgroup\$
    – Kevin
    Jun 5, 2022 at 16:06

2 Answers 2

1
\$\begingroup\$

This should clear up your main problem: -

enter image description here

It's figure 25's use of the resistor R1 and zener D3 that limits the buck output voltage to between 10 volts and 20 volts.

Can the IR2110 be used as a high side driver only (to drive a buck converter with 60V output voltage)?

Yes it can.

\$\endgroup\$
8
  • \$\begingroup\$ Ah I see. My bad, thanks for clearing that up! Anyways, there's no problem on that side, do you have any clue as to why the IR is heating up? Is it because I'm using 50kHz for the switching frequency? \$\endgroup\$
    – Kevin
    Jun 5, 2022 at 16:31
  • \$\begingroup\$ Do you really need a 22 uF capacitor for C2? That strikes me as potentially problematic. However, heating up to 70 degC in a local (close to the PCB) ambient of maybe 40 degC (a guess by me) doesn't strike me as too problematic on the face of it. \$\endgroup\$
    – Andy aka
    Jun 5, 2022 at 16:43
  • \$\begingroup\$ @Kevin did changing the 1N400x diodes to UF types help? Where is the post BTW? Was it your post? \$\endgroup\$
    – Andy aka
    Jun 5, 2022 at 16:45
  • \$\begingroup\$ For the 22uF cap, I'm using Tahmid's reference design (tahmidmc.blogspot.com/2013/01/…). Oh and also, I can also smell a faint burnt plastic smell coming from the IR2110, which concerns me (maybe the 70 degC reading is not accurate). \$\endgroup\$
    – Kevin
    Jun 5, 2022 at 17:03
  • 2
    \$\begingroup\$ @Kevin 22 μF is far too high for a bootstrap capacitor; that much capacitance on the switching node is likely to slow the switching significantly. Most DC-DC controller ICs that I've seen specify 0.1 to 1 μF. That slowed switching may be part of your problem, as it would increase energy lost in the transistor during switching. \$\endgroup\$
    – Hearth
    Jun 5, 2022 at 17:23
1
\$\begingroup\$

Aha, I found my answer I was looking for. It was indeed on another community:

https://www.eevblog.com/forum/projects/buck-converter-pre-regulator/msg4104580/#msg4104580

Plenty of discussion there as well.

Tl;dr: use current mode control; power the high side with a DC-DC so it doesn't turn off (UVLO) when the output is powered, at low output currents, or high duty cycles. Layout example as well, notice the tight layout with wide pour areas.

As for the present issue, measure Vb-Vs as others have mentioned. If Vs is peaking too low and this cannot be avoided (stray inductance of diode, forward recovery), series resistance in D1 may help.

\$\endgroup\$
2
  • \$\begingroup\$ I haven't tried measuring Vb-Vs yet, I'll give it a look tomorrow. By Vs peaking too low, do you mean Vs reaching negative voltage? \$\endgroup\$
    – Kevin
    Jun 5, 2022 at 19:59
  • \$\begingroup\$ @Kevin Yes, negative peaks. \$\endgroup\$ Jun 5, 2022 at 20:01

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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