Dear All,
we are using the DRV8955 motor driver. We set its MODE to 330kOhm to GND (four independent High-Z, voltage on MODE_1.23V) and IN1-4 and EN1-4 pull to GND.
After that if we pull-up any one output by 1kOhm to VM=12V and on the output can be measured 6,3V. I think it should be 12V because the high-z state. On the oscilloscope all input and out pin are static state.
Do you have any idea?
Hi David,
thank you for your amswer, there is the picture here:
You can see on the picture the used settings and the external 1kOhm pull-up resistor. The driver is in the 330k to GND mode (in theory, I hope).
On the scope image:
Blue channel: VM voltage
yellow channel: DRV8955 OUT1 voltage
pink channel: DRV8955 MODE pin voltage
You can see: if the MODE pin voltage is 1.23V, the OUT1 voltage is about 6.3V, I think i should be 12V.
If I changed the MODE pin voltage to 0V (Mode 0), the current control is active and the OUT1 output level changed between 0V and VM, it is normal.
I do not understand, what is this current leakage in the 330k to GND mode. I repeated this test with the DRV8955PEVM board and the result was the same.
Hi Pal,
I just got word back from design. It looks like this voltage drop is caused by the leakage current.
The response from design is below:
"When the device is ON but OUT is Hi-Z the leakage is around 5.8mA nominally which means it will drop around 5.8V with a 1K resistor.
For nsleep=0, the leakage is 1.8mA nominally.
This leakage is present due to the strong pull downs in the high side gate taking current through diodes present between powerfet source and gate."
Best,
David
Hi David,
thank you for your analysis.
It means, when we drive an unipolar actuator (or DC), the recirculation current (fast decay) will have a little distorsion. After stopping the actuator this 5.8mA always flows through the coils causing residual magnetism, so we always have to switch off VM.
In the Low-current sleep mode the situation is the same. Do other pins have this kind of leakage?
Hey Pal,
David is on PTO the next couple of days, I will help support.
In the Low-current sleep mode the situation is the same.
I think it should be much less (about a third), down to 1.8mA in sleep mode, can you test/confirm this?
Do other pins have this kind of leakage?
No, it should just be the output pins.
This leakage current will exist in any of our integrated devices, you would have to switch to a gate driver to fully eliminate it. Look for the R_Hi-Z spec or similar in the datasheet if you look at other devices, though not all our datasheets list this spec. For example below from DRV8143 the leakage current will be between 12V/1.4kΩ=8.5mA to 0.185mA, though likely somewhere in the middle.
Best,
Jacob
