Strange full H bridge "problem"

Thread Starter

surri

Joined Mar 11, 2019
11
Hi!

Hopefully someone can help me figure out what is going on with my full bridge.

I am using a h-bridge, A4953ELJTR-T (datasheet attached) to control a solenoid. When I activate the the solenoid I use 24 v, but to hold it about 5 v should be enough. But this driver are causing problems.

When I connect a resistive load to the h-bridge, run a pwm signal and put the probes across the resistor I get some strange curves.

Blue and yellow is the voltage across the resistor (200 ohm). Subtracting on from the other gives the red curve, the nice pwm signal.

B20190630_123458.jpg

So it seems the H-bridge increase the voltage on both terminals to reduce the potential across the load instead of have more "off time". Maybe this is a common method? Does anyone know what its called? It likes to do this at lower duty ratio and it makes sense.


Now the problem I was having was the solenoid would drop out at around 50% (12v) duty cycle if running with this h-bridge, 15% (3.6v) if running from a lab psu. This is regardless of the frequency. Same results with 30 khz and 3 khz.
I tested with a "normal" half bridge and the results were similar to that of the lab PSU (dropped out around 20%).

If I slowly decrease the duty cycle i notice something like a really quick (<1s) "phase shift" around 50% duty cycle, the oscilloscope trigger is unable to center the signals for a brief time. My guess is that the driver is perhaps modifying its behavior and this interruption is causing the solenoid to drop out so early.

Perhaps I have misunderstood the whole situation, not sure. But if anyone could shine some light on this I would very much appreciate it!
 

Attachments

dendad

Joined Feb 20, 2016
4,451
What is the input to the A4953? Have you looked at the input signals to see if they both go low at the same time?
Look at the truth table on page 4.
 

Thread Starter

surri

Joined Mar 11, 2019
11
What is the input to the A4953? Have you looked at the input signals to see if they both go low at the same time?
Look at the truth table on page 4.
Thank you for taking the time to help me!

Aha now I realize why I had the weird flick at 50% when running on 500 hz, its going to sleep at 1 ms of both low. Thanks! But running at 30 khz I still have the "weird" behaviour. One input is constantly driven low while the other is pulsed with 5v pwm.

Schematic of the driver part. driver.PNG
 

dendad

Joined Feb 20, 2016
4,451
One input is constantly driven low while the other is pulsed with 5v pwm.
I would think that is the way it should work. And if you change direction, the outputs swap.
On one of my boards, I use the A4950 chip, similar to this chip too.
But if you are just driving a solenoid, why are you using an H bridge in stead of a FET?
 

Thread Starter

surri

Joined Mar 11, 2019
11
Its also used to drive bi directional solenoids so it needs to be able to switch polarity. I have checked both input channels to the chip, the "low channel" is nice and low with little noise. The driving channel has a clean pwm signal at the driver pin.

I have recorded what happens with the current through the coil while lowering the duty cycle from 100% to 0%. Its quite odd. When the solenoid drops the core the signal becomes distorted, Im not sure if it causes of the drop or if the drop cause the distortion (back emf cause problem?).

https://imgur.com/a/KufEYfO

Then I did it again but with a half bridge driver. It doesn't have the weird distortion.

https://imgur.com/a/RxgV2v8

With the full bridge it drops the core at around 14v, with the half bridge at around 6V, with lab psu 5.8V.

Any ideas?
 

Thread Starter

surri

Joined Mar 11, 2019
11
Ok I solved it. If anyone else runs into this or a similar problem;
I think the problem was the driver running in fast decay mode instead of slow decay. This wasn't very clear to me in the datasheet that is was activated and the effect it would have on the coil.

I was sending a low signal on channel A and pwm on channel B. This resulted in fast decay mode. I am now sending a high signal on channel A and pwm on channel B. In this mode, when both inputs are high (pwm signal high) its "breaking" an electric motor, in my case, not de-energizing the coil.
 

crutschow

Joined Mar 14, 2008
34,282
When the solenoid drops the core the signal becomes distorted, Im not sure if it causes of the drop or if the drop cause the distortion (back emf cause problem?).
Yes, when the solenoid moves, it can generate a back-emf, but that should not cause any problems in your circuit.
 
Top