DRV8434: loosing steps at 1/4 step

Hi Zeinab,

I think VM of 48V may be a bit too high for DRV8434 which VMmax is 50V, you have only 2V margin for any overvaltages and ringing in the circuit.

Second picture shows current waveform at 250Hz what for 1.8deg motor gives speed 250x4x60/200 = 300 rpm, that waveform looks quite OK.

The first current waveform is at 500Hz what gives 600rpm, that waveform is distorted by motor BEMF. At 600rpm motor has reduced torque but with low load and proper acceleration/deceleration ramp should not loose steps yet. If you increase speed further you will probably go into midrange resonance and some troubles with motor stability.

If you want to compare motor currents at different microstep resolutions, please do that at the same motor speeds (the same current waveform frequencies).

If you want to run Nema23 motors at around 600rpm with sinusoidal current and with nominal motor torque (or close to nominal) you need a motor with low inductance of around 1mH or lower and probably nominal current of around 4-6A what would be too much for DRV8434. In this case the better choice would be DRV8462 or DRV8711.

I guess with present setup motor current would stay sinusoidal up to around 400-500 rpm. If you want to run motor as smoothly as possible at low speeds you should go for 1/16 microstepping or higher. High microstepping modes are quite effective means of reducing low speed resonances and possible step loss caused by that phenomena.

Regards,

Grzegorz

Thank you very much for your reply. I had not noticed the role which speed is playing here. it's true, if I limit the cruise velocity, I can get a fairly good current graph even in low steps. as for the VM, 48V is what I have on my board and I have found the smart tuning drivers from TI very promising, solving the problem of overheating at high currents while following the sinusoidal waveform and providing the desired torque (1.7 for this motor). we have been waiting for DRV8462 and now I see its available. since the Eval for DRV8462 is not available, does it work if I get the DRV8462DDWR and replace it in any other Eval kit from DRV84XX family?

Thank you Grzegorz for your post .

Hi Zeinab,

The DRV8462EVM has been released but out of stock at the moment. We expect it to restock within a few weeks. The lower current version (up to 3A) is the DRV8461DDW https://www.ti.com/lit/ds/symlink/drv8461.pdf. The EVM for this is available to order now, https://www.ti.com/tool/DRV8461EVM, 

Regarding your question on DRV8434 STEP pulse timing please refer to page-8 of the datasheet, snippet below. The indexer can support up to 500kHz, so limitation would be on the motor + drive combo. Our EVM uses approx. 50% duty cycle but it is not necessary as long as tWH and tWL specs are satisfied. 

You mentioned "same motor can give a very smooth 1/4 motor current graph when i run it using DRV8434 Eval kit". Could you please share the current waveform for this? What I'm noticed in the 1/4 step current waveform you shared previously was the current was not rising fast enough to exceed Itrip for that microstep and did not result with tOFF chopping until it hit the peak amplitude. Did you observe the stepper physically lose steps? Usually such scenario results with reduced torque not necessarily missing steps. But the spacing of the microsteps may not be what was intended. I hope this helps.  

Regards, Murugavel

Hi and Thanks for your reply, 

the motor current graph with 1/4 step was at faster cruise velocity which resulted in the distorted graph. while FW was sending the same velocity and acceleration, you see the measured frequency is different for 1/4 step Vs. 1/16 step. (both previously attached)

using the feedback i got here, if I reduce the cruise velocity- with 48V, 1.7 Imax and NEMA-23,1.8deg, at around 3000pps, and 8000pps2, I can get same current graphs from both DRV8434Eval and my control board for 1/4 step.

however at the same configuration if i go for 1/2 step, the graph will be distorted.

In conclusion, it appears there is a tradeoff between the step, the cruise velocity and VM, which is apparently close to BEMF for this motor.

now this motor is in series configurations, can i run it faster with 1/4 speed while maintain the good current graph if i switch to parallel? will the driver maintain the same torque then?

- I am designing my board with DRV8434PWPR, HTSSOP (28). so far, the current and temperature has been fine, although 2V margin is not ideal. I have been waiting for DRV8462 with the same package since i have some space limitation concerns and have no idea if it will be available sometime soon.

Hi Zenab,

"can i run it faster with 1/4 speed while maintain the good current graph if i switch to parallel?" - Yes, you will be able to run it around 2x faster in all microstep resolutions but you will have to deliver 2x more current to get the same torque. 

"will the driver maintain the same torque then?" - you will need driver that is able to deliver 2x higher current and DRV8434 probably will not be able to do that.

If you can afford to lose one or two DRV8434s you can try to push it to its limits, maybe lower torque in parallel connection would be sufficient for your application.

I would do it as follow:

- lower VM to lets say 45V,

- add large electrolytic cap between VM and GND, something around 2200uF,

- set some small speed and run the motor,

- slowly increase driver current while monitoring its temperature,

- once I know maximum current I would play with speed and all other settings,

- stay below lets say 1500rpm (probably it will be possible to go much faster but uncontrolled stop can lead to driver destruction)

Regards,

Grzegorz

Hi Zenab,

The DRV8462 comes in DDW-44 and DDV-44 packages only. The DRV8461 has an HTSSOP-28 which also would be a drop-in pin-2-pin with the DRV8434 with up to 65V operating supply voltage range. This is not released yet.

You mentioned "now this motor is in series configurations, can i run it faster with 1/4 speed while maintain the good current graph if i switch to parallel". Does this mean you're using an unipolar stepper motor configured with bipolar connection? Please elaborate. A datasheet for the motor will be helpful. Thanks.

Regards, Murugavel

HT23-ST_speed-torque.pdf

 I am using a hybrid stepper motor, in bipolar and series configurations.

Hi Zeinab,

Do you have anymore questions with this topic? thanks.

Regards, Murugavel

Yes, Thank you. I will try DRV8452 PWPR-28.