2

I am trying to drive a 23-pole pair(i.e. 46 poles) BLDC motor using an LV8139 pre-drive, the output of which is fed to drive the 3 half bridge MOSFETs. I managed to connect the correct phase sequence(out of the 6) between the motor and driver.

The pre-drive IC basically provides a hardware solution for 120 degree and 180-degree commutation(sine wave). It requires a DC control signal to vary the speed of the motor, which I am doing via a potentiometer. Here is the link of the datasheet for the same

https://www.onsemi.com/pub/Collateral/LV8139JA-D.PDF .

Now there is literally zero material on the internet about this chip but I am able to use it to drive the motor. I do have a few problems though:

1.) Mostly I need to give a physical push (by hand) to the motor before it starts to rotate normally. I feel that this is an issue of the hall sequence. But once the external push is given, the motor starts normally.

2.) The specification says that motor will rotate in 120 degree mode till a certain speed and then it will jump to 180 degree sine wave. However I don't see any such transition i.e. the commutation is 120 degree always.

I searched about the first issue on the internet and found some suggestions to add a high value capacitance across the power supply to provide sufficient initial current, but the issue persists. Any suggestions are appreciated. Thanks & Regards

shwetank vishnu
  • 31
  • 3
  • What are the motor specs? What voltage are you running it on, and how much current does it draw? Can you show us the hall placement on the motor? Do you have an oscilloscope to show the timing? – Bruce Abbott Dec 02 '19 at 07:13
  • It is a 48 volt 3 phase motor. There are 23 pole pairs(rotor). It draws about 2.1Amps per phase. I didn't dissambled the motor but I assume the hall sensors are placed 120 degree apart( mechanical angle). Also I saw the three hall sensor output to check fo the sequence. And was able to confirm the 6 signal sequencing. – shwetank vishnu Dec 02 '19 at 07:18
  • Kv, Kt, rated current and power? It's possible you have a hall sensor configuration that isn't quite right. Make a chart of all possible combinations and try each one in turn. What bulk capacitance do you have now, and what is your power source? Can you show us the schematic and a photo of the controller? – Bruce Abbott Dec 02 '19 at 07:35
  • @BruceAbbott , I am not sure about the Kv and Kt however the motor has following specs:- 400-600 rpm at no load torque: 37NM Max torque: 72NM this is basically a scooter motor with a rim size of 10 inches 1Kw and 48volt as suggested above and I tried the combination of motor phases, do I need to do the same with hall sensors too?wouldn't it be redundant?I will add a link of the schematic... – shwetank vishnu Dec 02 '19 at 08:45
  • so there's 69 poles on the stator? – Jasen Слава Україні Dec 02 '19 at 09:09
  • @Jasen The rotor is a 23 pole pair i.e. stator would be 138 pole. – shwetank vishnu Dec 02 '19 at 09:21
  • May not be relevant. but note this datasheet note from [datasheet page 8](https://www.onsemi.com/pub/Collateral/LV8139JA-D.PDF): "The Hall IC to be used must be of open-collector or open-drain output type, and it must be pulled up by VREG5. The type of Hall IC incorporating a pull-up resistor cannot be used" | and also " Furthermore, when using a gate driver that cannot turn off the control power while VM is being applied, the control power must be supplied from the VCC pin rather than from the HB pin. An element with a short reverse recovery time must be selected as the output FET." – Russell McMahon Dec 02 '19 at 10:57
  • This **MAY** be useful. Or not. [Here](https://www.rlocman.ru/i/File/apnote/ON_Semiconductor/Power/LV8136V.pdf) is an application note for an LV8136 - the pinout seems to be the same. The 8136 does not have the same Hall drive requirement and has some differences in the basic functional description. || [ LV8139 data sheet](https://datasheet.lcsc.com/szlcsc/1910181032_ON-Semicon-LV8139JA-AH_C438946.pdf) & [LV8136 data sheet](https://www.onsemi.com/pub/Collateral/ENA2005-D.PDF) – Russell McMahon Dec 02 '19 at 11:19
  • _"I tried the combination of motor phases, do I need to do the same with hall sensors too?"_ there are 25 different combinations of phase and sensor connections, most of which are wrong. – Bruce Abbott Dec 02 '19 at 17:22
  • @RussellMcMahon , I am using a 5 volt single ended open drain hall IC. The negative inputs of the LV8139 have been set to a 2.5Volt reference in order to get the high and low hall signals. And yes I am not using the HB pin to power the external driver IC. I checked out the datasheet you recommended and it is similar to the one available for 8139, note that I am able to drive the motor however at certain times it requires an external push by hand. – shwetank vishnu Dec 03 '19 at 09:34
  • @BruceAbbott, I see, but I tried to work around with the phase combinations and found the one that is able to drive the motor. It's just that at certain times during the start it requires a little push to allign the hall sequence and then it works perfectly fine. There are 2 pages in the datasheet that describe about the hall sequence, it also says that if the sequence is not correct then the motor will work in 120 degree mode, which apparantly it is not doing and getting stuck to wait for the right sequence. – shwetank vishnu Dec 03 '19 at 09:38
  • @shwetankvishnu I cannot tell from your response if you have noticed or acted on my comment re "and it must be pulled up by Vreg5" as per page 7 & 8 on the datasheet of the IC you are using. So, again. "The Hall IC to be used must be of open-collector or open-drain output type, and **it must be pulled up by VREG5.** " – Russell McMahon Dec 03 '19 at 09:46
  • You could try positioning 3 Hall sensors yourself if there is any doubt about the suitability of the existing arrangement. – Russell McMahon Dec 03 '19 at 09:47
  • @RussellMcMahon , Hall IC outputs have been pulled high as described. I tried changing the hall configurations i.e. the 6 combinations but the motor still needs a push to start. – shwetank vishnu Dec 04 '19 at 07:46
  • Can you physically check the Hall locations and confirm that they make sense magnetically in your case. It's hard to imagine that they are "wrong", but checking may be instructional. – Russell McMahon Dec 04 '19 at 12:38

1 Answers1

0

two theory:

  • the motor requires a high starting torque because of the motor load. See if you can start the motor without a load or apply a higher CTL signal at startup.
  • The hall sensors are not correctly aligned with what the motor driver expects, check the "Timing chart" in the datasheet vs the motor's datasheet.
m-walker95
  • 255
  • 1
  • 10