2
\$\begingroup\$

Excuse the quality, sorry;

enter image description here

RED is GATE of Q2. YELLOW is GATE of Q4.

5V INPUT AT ALL STAGES

What are possible reasons that my H-bridge has a little 'bump' at the turn-off of the high-side?

It's reflected on the Vin(bridge) too;

enter image description here

Circuit:
enter image description here
(also here )

I've added a new shiny ceramic capacitor to the Vbridge and GND connection, but no dice. Could it be that having a ground common to my H-bridge ground is bad? Or perhaps the capacitor is refusing to discharge beneath 6V~ for a more obscure reason?

EDIT: It's driving a 15-20mH load, but the bump is present with no-load and resistive loads on the order of 1000R, 100R, 10R.

EDIT2: Also, I've noticed that if I put in above 17V to the bridge, the power supply I'm using just shuts itself down and I have to restart the circuit.

EDIT3: Here's a bunch of images of measurements I made.


Q2 DRAIN to GND

enter image description here


Q2 SOURCE to GND

enter image description here


Q1 SOURCE to GND

enter image description here


Q1 and Q2 GATE to GND (10V)

enter image description here


NEW WAVES

5V H-BRIDGE OUTPUT

enter image description here


10V OUTPUT

enter image description here


Q2 Gate/12V

enter image description here


\$\endgroup\$
  • \$\begingroup\$ What load is the bridge driving? What is the white waveform in the first image? \$\endgroup\$ – Jon May 12 '15 at 21:25
  • \$\begingroup\$ 15-20mH, High-side/Low-side. Red and Yellow respectively. \$\endgroup\$ – ARMATAV May 12 '15 at 21:26
  • \$\begingroup\$ The voltage at the gate would be intresting too \$\endgroup\$ – PlasmaHH May 12 '15 at 21:34
  • \$\begingroup\$ @PlasmaHH Well according to my first image that looks to be around 15V high-side, 10V low-side. Is that what you mean? \$\endgroup\$ – ARMATAV May 12 '15 at 21:36
  • \$\begingroup\$ @ARMATAV: ah, ok, I wasnt aware that you are referring to the gates with it, I was thinking you meant the drain/source \$\endgroup\$ – PlasmaHH May 12 '15 at 21:44
2
\$\begingroup\$

It's fairly straightforward. When Q2 is turned off, Q4 is already off. As a result, the output, LVCoilA, is left at an intermediate voltage. This is probably determined by a combination of the two MOSFETs' off current, combined with the capacitances of each MOSFET, producing a sort of voltage divider. If you look simultaneously at Q2 gate and LVCoilA, you'll see that the gate is set to the bootstrap driver's reference, which is LVCoilA, about 6 volts in this case. Since LVColA is also Q2's source, Q2 is turned off.

When Q4 is turned on, you'll see LVCoilA, as well as Q2's gate, get pulled down to ground.

As for the noise on the voltage bus, in order to make much sense of it, you must simultaneously display (for instance) Q2's gate and +12. As it stands, the two traces differ by x5 on the timebase, and there is no obvious way to be sure which +12 transients occur with which gate events.

\$\endgroup\$
  • \$\begingroup\$ Hey, thanks for the answer. So if I'm getting you properly, you're saying that my measurement of Q2 gate to GND is an incorrect representation of what is actually happening, and that Q2 gate to LVCoilA is turning off correctly? If so, take a look at this (ignore the level shifting); i.imgur.com/mhD8E05.jpg that is Q2 gate to LVCoilA, coupled with Q4 gate to GND. Am I good then? Also, on your note about the noise, I will definitely get you guys a proper reading when I'm home in an hour. The thing is that my power supply will shut itself off at 17V, and I think it's the noise doing it. \$\endgroup\$ – ARMATAV May 12 '15 at 22:31
  • 1
    \$\begingroup\$ I believe you are doing it right. The MOSFET only cares about its gate-source voltage, so that's what you need to measure. And for the love of God, will you please turn off the overhead lights when you take pictures of your scope? \$\endgroup\$ – WhatRoughBeast May 12 '15 at 22:55
  • \$\begingroup\$ Hahahahaha, it's so eye-gougingly bright though, I love it. Going to get those readings of the Q2/12V to you. \$\endgroup\$ – ARMATAV May 12 '15 at 23:20
  • \$\begingroup\$ Hey mate, here's the gate-Q2/12V reading; i.imgur.com/xKZdCh4.jpg Here's two readings, 5V and 10V, of the actual H-bridge output; i.imgur.com/Hhftwyh.jpg and i.imgur.com/LuXrsOs.jpg \$\endgroup\$ – ARMATAV May 13 '15 at 1:56
  • \$\begingroup\$ I suspect that what you're seeing on the 12v line is capacitive coupling of the gate drive through the drain of the MOSFET. Try adding more capacitance to your 12v line: a 10 uF, and a 1000 uF, each to ground, connected as close as you can to the MOSFETs. For the 10 uF and existing 100 uF, tantalums will be better than aluminums. Also, make sure that your grounds and 12v are close to the MOSFETs with thick traces or wires (I don't know your construction technique). Make sure your scope ground is as short as possible, and as close the MOSFET sources as possible. \$\endgroup\$ – WhatRoughBeast May 13 '15 at 2:16

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.