DRV8801 over current trip threshold

Hi Igor,

There is not much tolerance when it comes to the SENSE voltage. It is very precisely set at 500 mV. As soon as the 500 mV are met, the H bridge is disabled. This is after the TBLANK period of course. I believe the TBLANK period is somewhere around 1.5 us.

If you short the SENSE pin to GND, then there can not be any ITRIP events. You can still get OCP events, which behave 100% identical to an ITRIP event except that it is not based on the SENSE voltage but on the FET current. As soon as the current goes above something like 3A, the H Bridge gets disabled.

Whether the H Bridge gets disabled on an ITRIP or an OCP, it will remain disabled for the next 1.5 ms. As soon as the 1.5 ms are up, the H bridge will enable itself again (given the input signals of course).

Hope this info helps,

Jose Quinones

Jose, I think you are wrong here.  Go back and re-read the DRV8800/DRV8801 datasheet.  The 500 mV specification on the SENSE voltage is not an overcurrent trip point, it is the Absolute Maximum Voltage rating for the pin.  Exceed +/- 500 mV on the SENSE pin and the part is damaged.  There only overcurrent trip on the DRV8800/8801 is the 3A internal (I-OCP) protection circuitry--and it is independent of the SENSE voltage.  Are you sure you work for TI?  Because you are giving out inaccurate information.

Actually, he is correct.  Sense is internal, through resistor to ground ' sensing ' voltage' at the pin. http://www.ti.com/lit/ds/symlink/drv8800.pdf (Page 6)

So voltage actually at the pin will trip it.  You are also correct in that if you apply voltage for any reason over the 500mV on the pin ( but why would you? ) you would damage it.

Part of the internal FET H-Bridge protection senses the current to also cut power to the H-Bridge (internal voltage/amp monitoring for both the Sense pin and the current draw over the H-Bridge) so in the end, check your specs on the resistor from Sense pin to ground and make sure it's tolerance is +/- 1% alternatively use a variable resistor with center to one end to adjust and a series resistor to meet halfway to the value of the Sense resistor.

When it comes to wiring up the DRV88xx series, i start with the recommended setups then drop in the values/components I need for my use.

hope this can calm things down and straighten it out,

Eric

Hi John and Eric,

During normal operation, while driving an inductive load, as soon as the SENSE pin becomes 500 mV, the H Bridge will be disabled for about 1.2 ms. In essence it is not possible to go above this value (for a long period of time) unless you directly drive the SENSE pin with a voltage source.

Like Eric said, there are not too many reasons why you would want to do that. Curiously, our H Bridges are every now and then being used as everything but a motor driver, so it is always a good idea to know where the limits lie.

One last note. DRV8800/01 do not have a current regulation like the other devices (e.g. DRV8802/12/13/14, etc) because the TIME OFF is too long (1.2 ms). However, you can set the ITRIP according to the equation ITRIP = 0.5/SENSE and it will allow the H Bridge to be disabled at a current value smaller than the OCP threshold. Do note the behavior after an ITRIP or an OCP event are virtually identical.

Best regards,

Jose Quinones

Eric

Can you double-check this with the product engineers?  I have been told that reaching 500 mV will damage the part.  It is shown as the absolute maximum voltage on the SENSE input.  It would be ridiculous to have the maximum possible [damage] voltage also the trip point.  Sure the part might go into recirculation mode, but it might also be destroyed.  Who knows which will happen first?  Note page 8 of spec--Vsense maximum is 500 mV.  Anything beyond this value "may cause permanent damage to the part."  I'd sure like to know just what the part does inside--as I am using it in a new design now.  Of course, I am staying well away from 500 mV on the Sense line.

For the following, refer to Figure 3. DRV8800 FUNCTIONAL BLOCK DIAGRAM from http://www.ti.com/lit/ds/symlink/drv8800.pdf

Look at the bottom of the diagram.

In the ( not connected for purposes of making the diagram simple ) section marked Motor Lead Protection.  Power goes to the FET H-Bridge, possibly through it then out the Sense pin, through a resistor to ground, thus completing the circuit.  The FET H-Bridge is designed to have as much loss as possible through the FET H-Bridge in to what is driven leaving the ( up to value of ) 500mV.  

The 500mV would be easy to sense from the bottom of the FET H-Bridge AT the pin marked SENSE since that pin is reqired to complete the bridge to ground.  Any higher ( either by fault, loss of ground, etc ) will damage  the FETs in the H-Bridge.  The very low voltage of only 500mV is due to the resistor TO ground AT the pin itself.  With that in mind, I for one am glad I do not have to add the transistors/logic/FETs to do that outside of the package by making the equivalent of it with something like a LN2003AN and support components.

With such a low value and an external connection available, it is simple ( not easy ) to add an op amp to that pin and compare it to the rated 500mA so the MPU can have information that a fault occurred. 

Anything more on this subject would be best left googling " making FET H-Bridge "

Eric

Hi, 

I've also got a problem about DRV8801 that couldn't figure out. as you said unlike DRV8802, DRV8801 doesn't have current regulation engine, but what I understand is when DRV8801 meets the ITRIP point, the H-bridge will be disabled, and then current decay according to the MODE pin. on the other hand the DRV8802 WITH current regulation engine, once meets the ITRIP point, it also disable the bridge and trigger current decay mode, looks their behavior are identical, doesn't it? please give me some explanation.

another problem is we all know that DC motor's speed can be controlled by using PWM, many of DRV88XX chips are current control by PWM signal, once I send a PWM signal to these chip, how this signal will be used to control speed and current at the same time. for example if  I wanna increase motor speed by generate higher frequency PWM signal, the frequency is so high that the current through the motor winding doesn't have enough time to reach the ITRIP point, how the current regulation would be implemented?

Hi Ke Wang,

DRV8801 may look similar to DRV8802 and in essence, until the ITRIP event is registered, their behaviors are fairly similar. Once the ITRIP takes place, however, what happens to the inductive load is quite different.

On DRV8800/01, the H Bridge will be disabled for 1.2 ms. In this long period of time, the winding current will most certainly decay to zero. If you try yto regulate current, you really won't be able to as the current control is discontinuous.

On DRV8802, we have an internal current chopping engine with a switching frequency of about 50 KHz. What this trasnfers to is an internal 20 us period which will define how long the H Bridge is disabled. When the ITRIP occurs is asynchronous to the internal timer, so it is not possible to determine when in the cycle the ITRIP will take place. However, once the ITRIP takes place then the remainder of the cycle will be used to regulate current according to the DECAY pin state.

For example, lets assume the H Bridge is disabled and an ITRIP is engaged 9 us into the current chop cycle. In this case, the H Bridge will be placed into the decay mode specified by DECAY (slow, mixed or fast) for about 11 us. The process repeats on a continuous basis, but because the H Bridge is disabled for so little time, you can see current being regulated to one particular value. With some current ripple, of course, as defined by the decay mode.

To control speed, you will want to either control current by modulating the analog VREF input, which will give you a PWM with a duty cycle directly proportional to the programmed current, or apply a PWM of your own. Notice you do not want to engage both PWM's at the same time as this will represent a loss of control.

If you apply a PWM which leads the device to source a current higher than the programmed ITRIP, then the device will start chopping current in between your PWM cycles. The results are in essence catastrophic to the control structure and this should be avoided.

If you use a DAC output to modulate VREF, you can get the best of both worlds. The device will generate the PWM to control speed, and the motor will be running at the programmed current. In this scenario, however, you loose control of the PWM itself. For example, you will not be able to change the PWM frequency, and the TBLANK will limit the smallest duty cycle possible. It is my impression It should be fine for most DC motor driving operations.

Hope this info helps. Best regards,

Jose Quinones

Thanks for your reply!

as I expected DRV8801 without current chopping engine can be called an voltage control chip, is this correct? because I can't directly control the current through the winding, whereas the speed can be modified with duty cycle of voltage.

one more question, if  I feed an PWM signal into DRV8801, and set ITRIP to a certain value, say 1.5A. during the ON time of PWM, H-bridge enabled, and the current will be approached to ITRIP value, in this case what if the value can't be reached due to the OFF time being coming? or I should consider that the 1.5A is just an MAX current value for protection purpose, any value greater than this would cause the H-Bridge being disabled. during the disabled interval(current recirculation) of 1.2ms , I can't control anything until current is  decayed to zero.

in the case the current through winding is less than this value, the H-bridge would be turned on and off according to the PWM signal on ENALBE pin. is this correct?

Devices like DRV8802 has internal current chopping engine, which has an internal PWM signal with frequency of 50Khz, such signal is different to the PWM signal tha I should feed into ENABLE pin for speed control, right?

Hi Ke Wang,

DRV8801 is actually not a voltage controlled device. You can make it so by regulating voltage, but by itself, DRV8801 is plainly an H Bridge with some protection blocks. It is flexible enough that you could add your own current regulation or your own voltage regulation, for example,  by adding some code in a microcontroller that generates a PWM input after reading current information from the VPROPI output ,or sampling voltage information from the power outputs.

in the case the current through winding is less than this value, the H-bridge would be turned on and off according to the PWM signal on ENALBE pin. is this correct?

Yes! When you feed a PWM into any of these devices (DRV8801/02), the H Bridge will be ON until an ITRIP event is acknowledged (because current went above the ITRIP level) or until the PWM TIME OFF transition comes up. If there is no ITRIP, then the outputs simply obey the inputs.

In either one of these devices, you do not want the internal ITRIP generation to take place if you are applying your own PWM. The reason is when you apply a PWM you are assuming the motor will move at one speed. For example, if the application voltage is 24V and you are applying a 50% duty cycle PWM, then you are expecting the motor to move at the speed you would get with 12V. However, if an ITRIP event comes up while the TIME ON portion of your PWM, then the duty cycle is no longer 50% and the voltage the load sees is not 12V. Hence it is crucial to operate without ITRIPs when dealing with DC motors.

In fact, a good percentage of DC motor users will just disable the current regulation by making VREF very high, or sometimes going to the extreme of wiring the SENSE pin to GND (You don't want to do this with DRV8801 as this will render VPROPI useless). The only time in which it makes perfect sense to regulate current while driving a DC motor is if you want to regulate motor torque. This is not the most typical of applications but it is out there. Curiously, in BLDC lore, speed is controlled by regulating torque. We must realize all of these aspects of motion control are intertwined. Clearly, if you have higher torque you should develop higher speeds, and so on.

The internal PWM on the DRV8802 is not like the PWM that you would apply at the ENABLE pin because the internal PWM is trying to regulate current, while the PWM you will apply will most likely be used to regulate speed. The 50 KHz internal timer will apply chopping chopping cycles (H Bridge enablement/disablement) only if an ITRIP is met. Hence, the 50 KHz is for the timer, not necessarily for the resulting PWM itself, as depending on motor inductance, application voltage and motor speed, there might be cycles in which the H Bridge will not be disabled. For example, if all of these variables are such that you skip every other cycle, then you get a PWM with a 25 KHz frequency.

Hope the info helps! Best regards,

Jose Quinones

Thanks again for your info. now I understood the basic concept of these device.

In the case of current regulating, should I apply higher voltage than rated voltage of a DC motor? because as a PWM signal feed into ENABLE pin, the current will be approached to ITRIP point, if the voltage is not high enough, current will be increased slowly, so that during the ON time of my PWM signal, there isn't enough time for it to reach ITRIP point. as result current will never be regulated to the target value. alternatively I have to make the duty cycle of PWM signal enough long to make sure the current will have enough time to rise. 

is the above description correct?

Regards

Hi Ke Wang,

Current will reach ITRIP only if torque requirement are high enough. This happens during startup because breaking the motor from rest requires some major form of energy. There is also the capacitive parasitic component. Once the motor is running, however, application voltage will have little to do with whether the motor current is close to ITRIP or not. Current is torque, and voltage is speed.

With that being said, you are welcome to use higher than rated voltage, but this poses a threat. In the event the application fails and aplies 100% duty cycle to the H Bridge, the motor will rotate at a speed larger than its rated speed. Whether this will destroy the motor or not depends on how much higher is the voltage. We can feel certain, however, that the motor's life will be adversely affected.

Best regards,

Jose Quinones

Hi Jose

still with DRV8801, how should I determine the Itrip value in order to choose sense resistor for stalling control? for example if I have a motor with rated voltage 12V and rated current 2A, should I use the 2A for the calculation? and I also find on datasheet of DC motor that there is one term called stall current which is much greater than the rated one.

once I set ITRIP, the current will draw as required by the load, but the peak value would be ITRIP, is it correct? if also I apply this configuration for another motor with smaller  rated current, would it be drive fine? will it be overheated? 

finally can you give me some hits if I'd like to apply current regulate using DRV8801, after the voltage on VPROPI pin is feed into DSP/MCU's ADC module, how should I change the PWM signal which is proportional to the current. and which pin is used to send this PWM signal? ENABLE OR PHASE. hope I make this clear/

Regards

Hi Ke Wang,

Stalling current should definitely be the largest current the motor will see under normal operating conditions.

It is hard for me to just say how to select the SENSE resistor as this will depend on the application which in this case you know better than me. A good rule of thumb is that you would choose the SENSE resistor to stay away from the stalling condition if you are not planing on operating the motor under staling conditions.

For example, if you are articulating a fan, stalling is a rare scenario. If you are articulating a linear actuator that pushes on something heavy, stalling may be commonplace and the application may need to go into this current level to operate properly.

To properly select a SENSE resistor you need to know how much current you truly want to allow sourced. Usually this information is given by the mechanical engineer designing the application. They know how much current the motor needs as they know how much torque their application needs. They also know for how long the torque needs to be applied.

Sometimes this information is unknown, however. In this case, you can measure the current on the application and determine how much current you see under normal operation.

Do note that if you want to do torque control, then DRV8800 is not the best device. Once you disable the H Bridge and current goes to 0A, then you lose all torque. For torque control you need a device with real current regulation such as DRV8802/14/40/41/42 or you would need to implement current regulation yourself. Another option, of course, is to close the loop and add a shaft encoder and a PID loop.

Best regards,

Jose Quinones