Developing a board with DRV8332, need help

Uhh… Ding ding! Lets put in on a bike :rofl:

that looks intersting but max switching frequency is 20khz, hopfully thats right outside the herable range for motor whine. thats the only problem i see with this. There is also one for 20$ that is 30A wich would be ideal for my application.

I am running at 20khz and I can’t hear anything. Please note that the actual switching frequency gets “smeared” by the motor coils and unless your coils are loose there is no noise at all. Unless your hearing is really good. However, in a bike or board application outside on the street you can’t hear a thing.

They have lower ones which are cheaper, that’s correct, and the circuitry is the same for all of them. Are you talking about the NFAM3065L4BT? Or a completely different MOSFET one?

You are so right. I never understood why would someone pay $100+ for a vesc or odrive when they can use one of these for a fraction of the price at much higher wattage.

VESC on a e-board does have a place in this world. What I dont get, Is why the international community does so little, when it comes to developing next gen coated iron powder stators for small stator design. We should create a fund and let Musk invest a few mill on tooling :wink: WTF is up with sending a car into space/vacuum ?

did you already do somthing with the NFAM5065L4BT? that on semi system seems nice, 20 to 50 amp in the same package. Also seams to be sufficient supply in these trying times of the semi shortage.

I am not in this business, and know nothing about it.

Next on my list. Need to finish 8332, I need a low current small footprint design.

im probably going to use this part ( https://www.mouser.de/datasheet/2/308/1/FNB43060T2_D-2313697.pdf ), its 15$, could you please have a look over the datasheet for any problems, as far as i can tell this should work for my aplication.

EU first… Another matter, the danish ministry of agriculture has just been exposed in payin 3x fold the EU support for individual farm support, by letting them willingly create 3x different companies. How does this this leave the African farm´s competing with EU products. Why is this not a story? you tell me… There is something rotten in Denmark :eye: :eye: Its all connected…

As I said I’m not an expert and haven’t worked with this. It seems fine at first glance, you may want to ask others on this board if they have any input.

To get you started these are the design notes:

https://www.semiee.com/file/ON/ON-FNB43060T2-NOTES.pdf

This looks like a nice driver… but comparing it to the DRV8332 is like comparing apples and oranges?

Are you intending to use it at low voltage? I don’t know exactly why, but low voltage IGBTs are kind of rare, and generally IGBTs aren’t recommended over MOSFETs for designs <250V - though I haven’t read a satisfying explanation yet as to why. On paper it seems they would have some advantages!

If you’re planning to run at 400V, then this chip is for you. At 30V, I’m not so sure…

On the one side, I’d love to hear about someone running SimpleFOC on a 400V driver and getting 10kW of power… wow! On the other hand, I really would hate for anyone to hurt themselves!

Anyone planning to work with self-made high voltage drivers and power supplies, they should probably know what they’re doing. That’s a bit of a different game to playing with the low voltages usually used with SimpleFOC motors, and is more than enough to seriously injure or even kill you.

I plan on using this thing for max of 50 volts but, in the future 230v vfd design interests me quite a bit, and a simplefoc inverter with an ac motor and an encoder could do some serious shit.
Anyway the datasheet for the thing says typical operating voltage is 300 volts but I don’t see why it wouldn’t work with 5v 20v or whatever voltage. Anyway I’m probably going to copy the reference design more or less. And see if I get something working.
I don’t want to spam this topic anymore and going to open my own for this thing.

My next board is an igbt with the driver I took a pic above.

According to specs the source/sink voltage is around 3V which I gues means it could be ran as low as a single LIPO cell. I wonder too so I’ll test drive it on a breadboard.

One way to tell… hur dur…

Agree. Heavy duty metal grid cage. Proverbial “don’t do this at home, kids…” However these drivers can be ran at less than 1A for development and test. Problem is selling those, I can’t imagine the liabilities.

Im just going to order 3 of them, what’s 50$ if it works hey, great, if not, then they are going in the parts bin and probably be newer seen again.

:crossed_fingers: :crossed_fingers: :crossed_fingers:

I don’t see why it wouldn’t, it’s like a drv gate driver and 6 mosfets in a package.

I’d be really interested to hear your results - if you don’t mind sharing!

I just put a topic up in Wich I intend to share my progress.

Hi @Valentine,

Very interesting. Your words very much echo my understanding and experience about the servo industry. It is ridiculously hard to get information on industrial servos, let alone get in-depth performance figures, and pricing information. I suspect even if you were prepared to pay say >$50k for a servo it would take a lot of work reaching the relevant person within a company (Siemens/Kollmorgan etc) to actually sell you a unit (<5) pieces. If anyone reading this is thinking…“nah, that’s wrong” - please let me know!

Back to your question - you’re after some suggestions for tests/motors to stress test your planned driver which will control a servo (one of three) in a joint for moving a payload (how large? what’s the smallest/largest payload?) in the most efficient (how will you know/test this? - I suspect as long as your the servo does what your algorithm tells it to do you’ll be satisfied?) trajectory from one point to another. Right?

Cool. So, functional tests and motors:

Stress tests:
a) Board functional tests: I am no expert. As you already have, I would start at the datasheet for the driver chip and design tests that confirm that the board you make behaves as described there. There are a bunch of over/under current/voltage and temperature expected behaviors listed. For tests beyond these - I defer to the hardware experts behind SimpleFOC boards!
b) System tests: these are hard without further insight into what your system aims to do. What stresses do you expect your system to fall under? Raising a load that is too heavy? Generally, I would list all the things that could go wrong, and then understand how will these events look like from your driver perspective? Then determine how I want the system to act when under that stress, and then what the driver has to do when it recognizes that it’s under that stress so that you don’t get some weird behavior and the whole thing falls apart due to some positive feedback loop.

Motors:
Hah, I feel like an idiot - but I have to say again, it depends :slight_smile:. For a robot arm would you be better served by steppers? If not, why not?
Some more random thoughts: if your max current is 10A (13A peak) let’s say your voltage rail is 24V, that’s a maximum of 240W delivered from your driver to motor. Let’s say your motor/mechanical gearing/coupling system converts 80% of that into useful mechanical work so delivered useful work output is ~200W, i.e. torque x speed = 200W (max). Say maximum mass is 10kg, lever arm 0.1m, max speed is: 200/(10100.1) = 20 (rad/sec) or 190RPM. That sounds like a gimbal motor - which can produce torque of 10Nm (direct drive setup).

Anyway, clearly, these are all the things I would be thinking about, but I may have strayed into some rabbit hole that doesn’t apply to you at all. These numbers all become much easier to figure out when your application is know. My approach would be to select a motor (given your known constraints for $$/current/voltage/RPM/torque/efficiency etc) and then design the driver board to drive that motor really well (bonus if it drives others). Hope some of these thoughts are helpful.

Best,
Nat