Half bridge driver

[MrHeckles]

I have built this circuit but it does not appear to work as i expected.

I assumed that it would drive a motor quite happily at what ever speed was selected via the variable duty cycle PWM input. If a low duty cycle was selected and the motor needed to slow down the inverse pulse going to the lowside mosfet would provide a regenerative braking action, thus slowing down very quickly.

What this circuit does in real life however is a little different. At a low duty cycle setting the motor runs a little jerky slowly. As the duty cycle is increased, the jerkiness increases and then at a high duty cycle, the motor stops.

At the highest setting (about 90% duty cycle) the high side mosfet gate only seems to be getting 5v (which is a bit odd I think). At the low duty cycle settings the high side gets less than that, but the lowside seems to get a nice 11.5v (give or take).

Could someone let me know if this thing SHOULD work properly (and maybe the driver chip is damaged) or whether the design could cause this issue?

Thanks

 

[audioguru2]

Hi Mr. Heckles,
Your boost voltage pin VB is missing a diode.

View attachment 37964

 

[MrHeckles]

I have now put in that diode. It made no difference.

I removed the cap across Vb and Vs. It made it run a tiny bit better.

It still seems like the motor is being driven, then stopped about 1/second. I cannot get the highside out to go above about 3v max. If I try to up the duty cycle going into the chip, the motor will stop completely.

Should the setup of the half bridge drive a motor if the high side and low side are turning on alternately?

Is there any chance that the circuit design is good, but the chip is busted?

 

[audioguru2]

Hi Mr. Heckles,
It seems to be a tremendous waste of power to accellerate the motor with the high side then brake it with the low side on each and every cycle!
But hey, if you want to keep the motor, Mosfets and power supply nice and warm ...

Didn't you do it like this?

View attachment 37965

 

[OutToLunch]

isn't the lower mosfet just acting as the freewheel diode does when the upper switch is off? it's not applying a braking force to the motor, just providing a low impedance path for the current.

I would agree that the boot circuit was not implemented properly.

Also - where is the decoupling capacitor for the driver IC? There is a very good possibility that even if he did hook everything up right, the turn on of the upper FET would cause the input rail to sag and cause the driver to trip a Power On Reset. I would also suggest decoupling the upper FET with some ceramics.

 

[audioguru2]

isn't the lower mosfet just acting as the freewheel diode does when the upper switch is off?

When the upper Mosfet turns off, I think the motor's very short duration inductive spike is negative, followed by considerable time when it is a positive generator.

it's not applying a braking force to the motor, just providing a low impedance path for the current.

The lower Mosfet conducts the generated current therefore is a brake.

where is the decoupling capacitor for the driver IC? There is a very good possibility that even if he did hook everything up right, the turn on of the upper FET would cause the input rail to sag and cause the driver to trip a Power On Reset.

True! The big electrolytic cap is an inductor at high frequencies, it needs a ceramic bypass cap in parallel.

I would also suggest decoupling the upper FET with some ceramics.

It might be a good idea but isn't done on the datasheet.

 

[MrHeckles]

I put a 0.1 uF ceramic cap in parallel with the big electrolytic, put one between Vcc and Ground, put one back between Vb (after the diode) and Vs, and connected the drain of the highside mosfet to the +ve rail.

It all works very well now. The highside output is up to 18v from the driver. None of the fets get hot. The motor speed can be changed rapidly up or down.

Thanks guys.

Would I be right in assuming that the lack of the decoupling cap for the driver would be the main culprit for it not working?

 

[audioguru2]

It all works very well now.

Horray!

Would I be right in assuming that the lack of the decoupling cap for the driver would be the main culprit for it not working?

That's right. Whenever a demand for high frequency current was requested from the power rail, its inductance of the electrolytic caused its voltage to flop to ground.
Which radio frequency does it use?

 

[MrHeckles]

The PWM pulse is from an earlier post of mine, the current limiting SG3524 setup. It's running a frequency of about 65 kHz.