Custom motor driver module
I am planning to make a high current motor driver module with IRLZ34NPBF mosfets. But I am wondering what components to add to make a nicely working system, like resistors, capacitors, and diodes. I know I need a 10k resistor for the gate to ground, and I know I need diodes from the drain to the vcc. But is there anything else I should have other than those?
443 Replies
Unknown User•7mo ago
Message Not Public
Sign In & Join Server To View
thank you. would the ceramic capacitors go across the would-be leads of the motors? and the electrolytic capacitors go from vcc to ground?
Unknown User•7mo ago
Message Not Public
Sign In & Join Server To View
okay thank you very much, I will come back with a schematic/pcb design sometime
I have seen some boards with 4.7uf and 0.1uf capacitors for decoupling, but wouldn't those be multiples of each other
Unknown User•7mo ago
Message Not Public
Sign In & Join Server To View
Here is my schematic so far:
I have 2 ceramics and a polarized capacitor at the bottom, a 10k pull-down resistor on the gate, and a flyback diode on the drain
I am also not sure what values I should be using for the poly cap
Unknown User•7mo ago
Message Not Public
Sign In & Join Server To View
Okay thank you very much, I did not know this 😅
Unknown User•7mo ago
Message Not Public
Sign In & Join Server To View
do you mean, a 10uf ceramic, a 22nf ceramic, and a 470nf polymer?
and no others?
Unknown User•7mo ago
Message Not Public
Sign In & Join Server To View
I think I will do this too haha
Could I just use 2 of those ceramic caps and the poly cap, instead of 3 of the ceramic caps and a poly?
Unknown User•7mo ago
Message Not Public
Sign In & Join Server To View
Okay thank you. I'll think about it
though if I do decide to use 2 instead of 3 ceramics, should I pick the 10uf and 22nf?
one very small and one a good bit larger?
Unknown User•7mo ago
Message Not Public
Sign In & Join Server To View
Alright thanks! I am planning to use the QWIIC system on the board too, that wouldn't make any unwanted noise right?
shared GND but not VCC
if that matters
Unknown User•7mo ago
Message Not Public
Sign In & Join Server To View
Alright sounds good
and how about the voltage for the polymer capacitor
the max voltage on the mosfet is 30v, so should the poly cap be rated for at least that?
incase I want to use 30 volts for the motor
Unknown User•7mo ago
Message Not Public
Sign In & Join Server To View
okay, I went with 40v instead of 35v haha
now for the temperature
I am hoping to maybe get a max current of 40 amps, and at a max copper trace temperature of 40C I would need 20mm wide traces
But I want to add a heat sink, im just not sure how I should implement it. Could I lay the mosfet down flat on the pcb (and solder the top tab to the board) then have a heat sink on the other side of the pcb?
Unknown User•7mo ago
Message Not Public
Sign In & Join Server To View
2oz of copper will cost way too much for me
I would like to just use a heatsink or two
Unknown User•7mo ago
Message Not Public
Sign In & Join Server To View
Would I be able to lay the mosfet down on the pcb with vias and have the heatsink on the other side?
Unknown User•7mo ago
Message Not Public
Sign In & Join Server To View
here is my current pcb layout:
I have that big grid of vias under the mosfet to the other side of the board
I also just realized, when I put the heatsink on the back of the board It is going to short circuit all of those bare power pads, is there anything I can do to stop that from happening
Unknown User•7mo ago
Message Not Public
Sign In & Join Server To View
hmm, I might just cut out the pairs of screw terminals I have and just leave the THT pads for wire too be soldered to, that way I can put the heatsink over more area, and hopefully not need to insulate it anymore
Unknown User•7mo ago
Message Not Public
Sign In & Join Server To View
I would do this, but the copper traces should be getting pretty hot too, so I think having a heatsink over those will help
I ended up just taking the THT pads out and adding large SMD pads
and if I cut the leads on the mosfet short, I should be able to stick a heatsink back there nicely
here is my final design:
If you have anymore suggestions I would be happy to hear them
Unknown User•7mo ago
Message Not Public
Sign In & Join Server To View
would I just do this so I could fit the heatsink in without trimming the pins?
Unknown User•7mo ago
Message Not Public
Sign In & Join Server To View
okay so you are talking about that
originally I was planning on maybe soldering header pins in there, but I decided that I could just temporarily stick pins into those pads to program the microchip, instead of soldering them in permanently. so really in the end those pads wont really "be there"
and they aren't exposed at the back either so I shouldn't have any short circuit problems
Unknown User•7mo ago
Message Not Public
Sign In & Join Server To View
no sorry, i am leaving the THT pads as-is, just not having any solder mask on the back and not soldering anything to them
if that makes sense
Unknown User•7mo ago
Message Not Public
Sign In & Join Server To View
yeah, I just figured since I would only need to program with them once or twice, I don't need to permanently solder pins in
I also have an soic-8 clip I can use to program the attiny85 if I do really hate the alternate way
Unknown User•7mo ago
Message Not Public
Sign In & Join Server To View
well I haven't really messed around with them before
maybe this can be a project to try them out on
Unknown User•7mo ago
Message Not Public
Sign In & Join Server To View
alright thanks, i can make the pogo pin board tomorrow haha
I noticed sparkfun uses small THT pads for the pogo pins to sit, I guess that helps with keeping them aligned
Unknown User•7mo ago
Message Not Public
Sign In & Join Server To View
Well I could have the pads covered up on that side and leave the other side bare
Unknown User•7mo ago
Message Not Public
Sign In & Join Server To View
Jlcpcb
Unknown User•7mo ago
Message Not Public
Sign In & Join Server To View
I mean that's how I've designed it
they just don't put any solder mask there like any other part of the board
Unknown User•7mo ago
Message Not Public
Sign In & Join Server To View
well im sure I can find a way to cover them if they don't put solder mask there
here is the now final design:
I am planning to solder on these to connect wires easier
Might need a welder for that 😂
Unknown User•5mo ago
Message Not Public
Sign In & Join Server To View
The ones for 8 gauge or smaller are really really big seems
I may not be looking for the right thing though
Unknown User•5mo ago
Message Not Public
Sign In & Join Server To View
well, the board should get pretty hot with the even with the heatsink of course, but might as well keep the wires from getting that hot if possible
Unknown User•5mo ago
Message Not Public
Sign In & Join Server To View
that's correct, I could figure out a way to connect it, but I though the tab would be enough, do you think I should connect it too?
that's right
Unknown User•5mo ago
Message Not Public
Sign In & Join Server To View
I will double check
one lead of the motor is directly connected to VCC, and the other lead is connected to the mosfet's drain
and everything as a common GND
Unknown User•5mo ago
Message Not Public
Sign In & Join Server To View
That's right
Unknown User•5mo ago
Message Not Public
Sign In & Join Server To View
I can only find them on the chinese sites so I don't have much information on them, I do know their dimensions, they should be copper, and there is this
(The model I picked does not have that tab on the bottom)
Unknown User•5mo ago
Message Not Public
Sign In & Join Server To View
I think the flux I have is this. I will definitely use a lot of flux
i like it has a "One-word" screw 😉
Unknown User•5mo ago
Message Not Public
Sign In & Join Server To View
sounds good
Here is the motor driver with just the mosfet circuitry
I might clean it off now
lol
I preheated the lugs before I put the rest of the board on the hot plate, and that worked great
They immediately melted the solder paste when I plopped them on
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
That 1st pin is a bit dry, I didn’t really have a good angle to reach it with the iron but it’s good enough for now, the 3rd pin took 2 soldering irons to heat up enough to get a real bond, but a little blob was hanging off which is why it looks a bit weird
I do this already, but i needed to clip the pins flush for the heat sink and I didn’t put solder paste on those pads anyway, just the big drain tab
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
Seems like you need to heat those planes up better :D. Even when conventional soldering (more like desoldering, thats the harder part IMO), I sometimes like to hot air preheat the planes to around 100C because they just suck the heat away otherwise making soldering hard.
I swear that 1 blob looks a lot better in real life 😭
Maybe this is a better angle haha
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
that's what was happening before i starting using 2 soldering irons
but the shading there is from the flux
i swear 😂
i am about to start testing it
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
From the other angle we can see that the solder is not molten into a single piece. Like when you add blobs from your iron.
Not trying to criticize your skills, just trying to help. As Rob said, it has happened to us as well
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
I was talking about this angle. When I'm only prototyping and it's not meant to hold up for any longer than just testing, I don't care about these on my work. It should hold up anyway
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
When I was soldering that joint, and I got a nice blob on there, a little bit was sticking off bridging to an untended via or something, so a bit later I just tried to smooth that little imperfection out and it made it look a lot worse
Anyway the testing was going well until I tried checking the voltage, but I may have hit something accidentally with the multimeter probes and then this happened 😭
luckily my laptop that was connected to the arduino still works, and the uno still turns on
but it isn't connecting to the computer anymore 💀
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
it was an n-channel
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
😭
better than anything else
any ideas why my uno won't connect to my computer anymore?
the USB ports still work and I am using a different cable now
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
ive got a multimeter
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
about 20 minutes
also, the only connections from the Arduino to the motor driver were GND and 5V
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
yes it was
when it happened, the arduino and the computer turned off too
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
should i check its resistance?
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
@Taire Look at this, we like seeing that. Some juicy naked FETs 🤪
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
it's at .7 ohms right now
not plugged in
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
it still isn't unfortunately
also the L and ON leds light up but not any others
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
this one's got an atmega16u2, and I recently bought a fake ftdi serial adapter or whatever, so could I use that?
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
im crossing my fingers
i got that error, but is it because i only have tx and rx connected to the board and no power?
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
i connected the adapter's 5v and gnd to vin and gnd on the arduino, but now the uploading message just stays there forever
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
😭
but don't i need to connect dtr and cts too?
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
lmao what
this is what I am using
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
what about trying to use the DTR pin too
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
this is a piece of a schematic for a ch340g chip to an atmega328pb, and it uses RX, TX, VCC, GND, and DTR, but not RTS
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
resp=0x30
resp=0x20
Seems like somethings happening. Not just 0x00
maybe some hope
i tried connecting dtr to the reset pin, but i might try somewhere else
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
so i am going to try and find somehere else on the board that would actually work as dtr
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
yeah, that's why i figured the reset pin wouldnt work since it's on the other side of the cap
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
well i found a pad i could hold the dtr wire on to, but it still didn't upload
though when I touch the dtr wire to that pad the L led blinks for a second
and when I take it off
well time to get some fake nanos 😭
do you think I would be able to buy a new atmega16u2 and replace it with the old one?
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
Afaik on the arduinos the 16u2 comes in a leadless package. So only hot air could be used
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
it probably wouldn't be too hard if i could use the hot plate, but there are too many tht leads sticking out the bottom for other components
and I don't have a hot air station unfortunately
but I do have a hot air gun....
well, wouldn't I also need to program the atmega16u2?
Try it on some old electronics and report back about the amount of smoke and delamination because of it
I wasn't contradicting you by any means. Just adding more problems to the one you listed
😂
I meant seriously haha
I believe you
but, i will need to program the atmega16u2 right?
so it can work as the converter?
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
well that is definitely out of the question
i might just put the atmega328p on a breadboard with the serial converter and just use the chip that way
i think all I would need is a tht crystal with tht caps, or a tht resonator
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
the chances were high 😂
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
I think I would feel better using an external crystal instead of flashing a new bootloader, keep it so the chip is still "arduinoy"
there are multiple internal oscillators? 😲
in the 328P
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
plus I haven't even figured out how to flash a bootloader on a different atmega328 chip
Interesting
yep, I will look it up, don't worry 😄
and also try them. I only knew about the watchdog timer being separate
once I get some new boards, how can I protect them if this happens again
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
I kind of missed what happened tbh
the atmega328pb in the leadless package
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
mosfet blew up, killed something with the serial converter on an arduino uno
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
no thank you 😂
Yeah but I saw no extra components. Did you have an Arduino connected to its gate?
i had 5v from the arduino connected to the gate
and gnd connected to the gnd pin
8'd add programming pins on my PCB
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
Resistors and transistors go a long way. Or if you want real isolation then,... that. Optocoupler or digital isolator
This is the board that I am trying to program an atmega328 on
Aren't those programming pins? Haha
well I wouldn't be too sure how I could wire in restistors or transistors
yeah, but I can't figure out how to use them lmao
You burn a bootloader using ICSP connection first
You need to figure the pin outs out
i had done this, connected to the arduino uno I killed today, and tried following a tutorial to use a minicore bootloader, but I kept getting errors
It seemed like i need a usbasp instead of another arduino?
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
yes, just for the 328pb
it would 😫
im thinking i will just use a different power supply to control the gate
Using a power supply to... control the gate? How does that work?
well i was just using 5v from the arduino before, so if I find a 5v power supply i could use that
I think i have some old barrel jack wall supplies that are 5v
Isn't a mosfet made for switching things on and off? Whats the purpose of that?
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
the 5v would be to power the gate of the mosfet in the motor driver
And what exactly does the mosfet do? Does it ever get switched or something?
If i had an optoisolator i would use it, but I don't want to wait the time it takes to get some
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
i was thinking about this, but for the test I wasn't planning to use an pwm or anything, just keep the gate open
im hoping the dc wall supply could take it 😂
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
if i want that fast delivery though, i would have to spend $8 on opto isolators on amazon
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
well it could blow up regardless
A resistor to gate would go a long way
i might do that
but it also wouldn't be 5v going to the gate anymore
Sure about that?
no haha
what resistance would I want?
im not sure how I would calculate that
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
i couldn't tell ya
i think I have been convinced to get an opto isolator
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
Your problem lies in bridging things that shouldn't be bridged. Like VCC to gate/GND/....
The opto isolator is not the solution to the problem, but in case of a problem it would save your Arduino and PC
yes and this is what im looking for
im trying to firgure out how I would use one though
for this case
Preventing the problem from happening would be a potentially better solution
would the output of the optocoupler have a significant voltage drop? NVM
also im not even sure how to wire this thing
the collector would typically go to a pullup resistor, the emitter to ground
when light hits the phototransistor the collector goes to ground, when off the collector goes to the pullup voltage, the collector voltage drives the MOSFET gate
What PWM frequency are you looking for? Above some, you will have troubles with a simple optocoupler connected to GND
i wasn't planning to use a pwm frequency
Yes, but after the testing is done?
well i haven't though about that
i haven't seen people use optocouplers when controlling mosfets
is that a normal thing?
very common between a low voltage controller side and a high voltage output side, the light beam provides high "galvanic isolation"
maybe I should add an optocoupler to this board if i make another revision?
Have you thought about switching speeds?
ah I guess that wouldn't work well with super high frequencies then
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
A transistor to amplify the output of the optocoupler could go a long way, couldn't it?
But as long as we don't know the requirements,we are just shooting in the air with our suggestions 😂
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
I thought that those have quite an internal resistance and that they can only supply in the range of 10mA-s limiting switching speed greatly
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
well thanks for all the help guys
hopefully i can restart testing this again
(and be more careful with my multimeter probes)
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
i was sticking the probes into the heads of the screws on the lugs which were very stable, just very close to things i could short
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
more the NPN emitter to 5v (wasn't thinking of MCU failure leaving the FET on)
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
for my next revision of this motor driver, I was thinking I could use a reverse voltage protection circuit like the one shown below to protect the microcontroller or whatever else is controlling the mosfet
but I am not totally sure how I would implement it
I'm working on similar circuits right now. Whats the issue?
well, i guess i would make the "load" the gate, and the "30V" the microcontroller pin output?
I'm not really following you there
i want to protect the microcontroller from anything bad on the high power side, for example a short circuit sending reverse voltage back to the microcontroller
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
I think it was a "testing accident" where they multimeter probed slipped
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
Bruh 😂
Would digital isolators be able to “switch” fast enough with a high frequency pwm signal
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
I don’t know lol, I just hope @robherc KV5ROB gets what I mean since it would be for a mosfet
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
So that is probably something people would almost never hit?
With a pwm frequency that high
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
great then, thanks!
i began testing the motor driver again today, and it was going great
but the 50A bridge rectifier on my power supply keeps over heating, and when it does it starts to smoke and makes the transformer get a lot louder. Im only at 20A too
it does still work, maybe i can run it underwater?
the bridge rectifier in question: https://www.aliexpress.us/item/3256801251045663.html?spm=a2g0o.order_list.order_list_main.50.d30e18027Y65lX&gatewayAdapt=glo2usa#nav-description
saw this lol
Weird that yours get smoking smoking hot
i am able to see a bit of smoke while its running, then when it gets too hot it immediately lets out a bunch of smoke, and keeps going until i unplug the transformer
even if there is no load anymore
it like shuts down into a thermal runaway or something lmao
Im guessing the epoxy or whatever the black stuff is, is what is burning or melting
do you think running it submerged in tap water will work?
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
im just letting the built in heat sink handle it, but its not doing enough
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
ive got the perfect fan for the job
@robherc KV5ROB i was thinking of trying to use optoisolators instead of digital isolators for safely controlling the mosfet gate, it seems like they support a wider range of voltages.
and It also seems like they can support a pretty high frequency pwm signal
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
so I am not totally sure how to incorporate optocouplers into an h-bridge...
Im not really sure what to do
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
If you were going optocouplers, why not separate the grounds as well?
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
Even if they are shorting VCC and GND accidentally?
Or having a ground break and the current could maybe even travel through their Arduino? Which won't be caught by the polyfuse
I know these should be avoided but in the first place he is still trying to avoid these "accidents" and the consequences of user/testing errors. In normal operation they wouldn't really need anything, right? 😂
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
Lmfao
I’m going to try something else
well it definitely wasn't worth the $30 for a real arduino
and came close to killing my laptop (it seemed like)
so I think It would be worth it lol
plus the ones on LCSC are way cheaper than that
Did you completely kill it?
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
Well it doesn’t connect to my computer anymore
Not completely killed though
About 20v is my vcc, don’t know anything else though
I’m pretty sure it uses diodes
well i got everything setup to restart testing the motor driver and it turns out yesterday's incident with the full bridge rectifier finally killed it
hopefully i can get my money back for that lol
I think this will be my replacement: https://www.amazon.com/Baomain-Bridge-Rectifier-MDQ-100A-Module/dp/B01JIKSHCA/ref=cm_cr_arp_d_product_top?ie=UTF8
Baomain Bridge Rectifier MDQ-100A 100A 1600V Full Wave Diode Module...
Baomain Bridge Rectifier MDQ-100A 100A 1600V Full Wave Diode Module One Phase
I could use this
and then find a way to either use a digital isolator or an opto isolator on the gates
actually, could I use a gate driver for the gate like this one? they have some hella fast speeds https://www.digikey.com/en/products/detail/infineon-technologies/1EDB7275FXUMA1/14309451
DigiKey Electronics
1EDB7275FXUMA1 | DigiKey Electronics
Order today, ships today. 1EDB7275FXUMA1 – 5.4A, 9.8A Gate Driver Magnetic Coupling 3000Vrms 1 Channel PG-DSO-8 from Infineon Technologies. Pricing and Availability on millions of electronic components from Digi-Key Electronics.
I feel like using a gate driver would be the way to go?
nevermind that they all have too high supply voltage ranges
I am looking at this digital isolator mainly: https://www.lcsc.com/product-detail/Digital-Isolators_Texas-Instruments-ISO7720DR_C486037.html
LCSC Electronics
ISO7720DR Texas Instruments Signal Isolation Devices | SOIC-8 Digit...
ISO7720DR Texas Instruments - SOIC-8 Digital Isolators Digital Isolators. Check out the in-stock pricing and datasheet for electronic components from LCSC Electronics.
i am wondering if this would work, @robherc KV5ROB
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
I wired the low side N mos the same way (5V drive voltage) in my design but I looked up the datasheet and it should be fine? It has a 20V+- Vgs so that's why. This is my MOSFET's graph, your has one as well that has to be considered Patrick
Also the 1k resistor on the gate slows down switching time by a lot. Completely hindering the ability of a gate driver IC. That is usually under 100ohms. Usually you also place them for each Mosfet and not 1 for the two
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
Well, for paralleling MOSFETs I thought it was also serving a purpose of matching them together by some few ohm resistors in 'parallel'. I'd really need to read up on that one tho
But that doesn't even matter this time
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
I think we'd all rather not know that smell 😂
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
I wasn’t too sure about that resistor, I’ll take that off
I’m not really sure what either of you mean @KLevi
What is the input and output?
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
Rob is saying that feeding 5V to the driver might not be enough and that it might not turn the MOSFET completely
Ohhhh
Don’t I just need to use a voltage above the VGS(th)
Mine also have a VGS of +- 20v
this is for the n-channel
Vgs(th) is where it barely starts conducting.
yikes looking at that graph, that isn't really good lmao
Rob just saved your design. It would've blown
the p-channel one looks fine, i'll look for a different n-channel 😬
You can find a different MOSFET or get a higher drive voltage
Do you have a level shifter for the P channel?
I don't
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
most development boards don't have a 10v output
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
yeah, but I still need to supply to it
im trying to get up to 48V
48V max
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
yes they are
the new ones are both 60v
So thats the max voltage on the S of the PMOS?
yes
So to get VGS = 0 (to turn the MOSFET off), you would need a voltage on Gate to be.... ?
i wouldn't know how to find that
im guessing this is something i can't do with my 5v
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
wouldn't they blow up anyway like my 1 did?
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
For the voltage across Vgs to be 0V, you need both of them to be 48V (or whatever voltage you want to use to run your motor)
same with like 12v?
Yes
And then if you pull the gate of your PMOS lower, let's say to 48V -10V so 38V, it will turn on
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
so pretty much what your saying is i need to use a level shifter for any drive voltage over 5v?
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
well I won't need to do that for the n channel ones right?
i mean I didn't have to do it for my current motor driver board
Yes but you will still need to delay/speed up the switching times to match with the P mos...
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
5a max is multitudes too low
Well I guess I’ll do that
So I would need 48v on the gate of all 4 mosfets?
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
That depends on your Vgs max
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
Switching that much power is not an easy task to accomplish tbh
well if i was using 48v i definitely wouldn't be trying to pull 30 or 40 amps
vgs max is +-20V for both of them, so should I just use like 18?
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
so the gate drivers don't just feed whatever vcc you gave it to the gate?
i wouldn't call myself that much of a beginner anymore
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
i'll go back to looking for those
If you don't want separate gate drivers for the MOSFETs, then that means that 20V is the max (unsafely) that you can switch. And less with the safety margin
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
i am not really sure how these work
sorry not that one
something like this one https://www.lcsc.com/product-detail/Gate-Drivers_Infineon-Technologies-IR2104STRPBF_C2960.html
LCSC Electronics
IR2104STRPBF Infineon Technologies Motor Driver ICs | Half Bridge 2...
IR2104STRPBF Infineon Technologies - Half Bridge 2 210mA 100ns 50ns 10V~20V 360mA SOIC-8 Gate Drivers Gate Drivers. Check out the in-stock pricing and datasheet for electronic components from LCSC Electronics.
i would use 2 of them for each side, and connect the high and low mosfets to HO and LO
one of the main thing's i am wondering about is why is there only 1 input for controlling when they are on or off
do they mean "in phase" as in HO and LO get the same signals and power?
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
and it turns one on/off when i give it a signal?
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
thanks, how about the vcc and vb, im not really sure how those work
vcc is rated for 10-20v
and vb is a max of 600 ish
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
im looking at one now with a vcc rated for 5-20v
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
it's 0-VCC
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
whoops
i'll just send the whole table because I wasn't sure which one you are talking about
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
so that's gonna work
would I be able to make the whole h-bridge with just n channel mosfets now?
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
hooray!
i just realized, you wouldn't be able to use pwm to control the motor speed right?
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
how would I do it?
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
but the single input I have only turns the one mosfet on and the other off right?
if i send a pwm signal through that, it would just reverse the motor super fast
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
but wouldn't the 2 pwm signals need to be synced up perfectly?
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
yeah i did
im not sure how it fully corelates
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
ohhh thanks
it's like a riddle lmao
here is the new schematic
here's the datasheet's reference design
Im guessing all the resistors on my schematic would be 10k pull downs?
i just realized that there are 2 inputs on this driver rather than 1, the HIN and LIN
does this mean I can control each mosfet by itself unlike the other one that just switches one on and the other off?
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
I have the HIN from the first gate driver connected to the LIN of the second, and the LIN from the first driver connected to the HIN of the second, this way I can just send a pwm signal down one of those wires to get the motor going one way, and vice versa
that would work right?
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
Some TVS diodes would look great on there IMO
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
i would keep that, but i found that those drivers are a lot more expensive than these
also I think people are more used to using motor drivers in this fashion
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
i can do that
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
i am using "LIN"
LIN with a line over it is the same as normal LIN, but needs a low signal instead of high (active low)
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
no sorry
the lin and hin from seperate drivers are wired together
not the lin and hin from the same driver
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
well it will work though?
it's cheaper, uses 2 inputs the same way other motor drivers do, and in my opinion easier to use
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
now I have a TVS diode accross the motor supply VCC and GND
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
Isn’t that how you normal control the speed of a motor?
Turn the mosfet on and off really fast
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
i see, what would be the better way to operate the mosfets
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
It is like if you cannot turn the motor OFF, you can only reverse it's direction if I understand Rob correctly. (Didn't look into the whole thing, correct me if I'm wrong.
but to make the motor spin with all the mosfets controlled independently, I would still need to sync both diagnol pairs of mosfets to make the h-bridge work
Or just forcing GND or VCC on both sides of the motor instead of letting it float so it coasts nicely?
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
But in the coasting state the motor acts like a generator and at that point a very large load is connected to it. Which would slow it down faster, right?
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
Switching speed is something we haven't really discussed IIRC.
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
Thanks! Also for the explanation, kind of had a gotcha' moment in there.
We don't know the application field but coasting and/or active breaking could be a design choice
it looks like almost every manufacture has their half bridge drivers always have the H mosfet opposite from the L mosfet
except for the one making the driver i was going to use
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
gonna take more than Kapton tape to fix that...
Bruh 😂
I tested the driver again today with the new full bridge rectifier in the power supply, but it was still getting very hot even with a fan blowing on it, so I had to make a makeshift heat sink instead 😂
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
It worked pretty well
I couldn’t tell that it was getting very hot
anyway after that one of the lugs on the motor driver fell off because the board melted the solder
At 15A the mosfet got to 117F but at 20A it jumped to 250F
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
the whole back of the pcb is flat and I bought heatsinks specifically for it
but im surprised the temperature jumped up so high with just a relatively small increase in current
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
im not really sure how you are supposed to read that first graph
Is that 0.5 to 2 ohms on the second one?
not sure if it is ohms or milliohms
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
oh, so if the case is 150c I can only pass 25A
back to the h-bridge for a second, do any of these pins need to suspport up to 48V?
for example, DRVH can only output up to 38v
when it may need to go up to 48v?
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
well all of these datasheets i have looked at are all rated around 30v for those pins
maybe i am reading them wrong?
it's either ~30v ones or ~600v ones with seperate inputs for HIn and LIN
Especially if you want to use N channel high side switching.
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
Yep, I just that 60-48V = 12V is kind of an ideal gate drive voltage usually
i guess i will need to use the seperate input ones?
since I can't find any of the "good" kind with the correct parameters
should I use more transistors to make the inputs go on when the other is off and vise versa?
or find a different way
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
this would still give me the same result though right?
one input and output for each mosfet
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
i think the auto switch on/off for one of the drivers would be a good idea
how would you do that
i should try to figure it out myself first
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
ooOOooo that's pretty cool
thanks
Well i tried making a circuit with a n channel and p channel mosfet, which didn't work out too well. but i did some researching and it seems like i can use an XOR gate to get what i need
here is the link https://falstad.com/circuit/circuitjs.html?ctz=CQAgjCAMB0l3BWc0FwCwCY0HYEA4cEMElURTJyBTAWjDACgwE0QBmNgNnclbU9ZteUEBhEImLdl3YEQ-QXMpjSDALLyBPVgjBihrZSnWbFOvbMOjjAd2nc2GShwec8UBgBl77J6LzujpSUEABmAIYANgDOVOQeAEo+GAHJCNzB8pRImTASAOY+aHBFYBkeQA
is there a name for these 2 xor gates stacked on top of each other?
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
do you think the logic gates are a better idea?
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
alright well thanks
i guess i will do it with the XOR gates then
I've noticed on all these 4 channel XOR logic gates (not sure if it's with others too) that there is a low level output voltage of 50mV
does it mean the output of one of the gates at a low state would be outputing 50mV?
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
ah okay, so I shouldn't have to worry about that
thank you
i just realized, do I even need to use these logic gates, if i instead use the gate drivers with the HIN and LIN with a line on top?
If I connect both inputs to the same wire, when I send a high signal the bottom will turn off and the top will turn on, and when I send a low signal the top will turn off and the bottom will turn on, right?
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
but LO should only turn on if the voltage is 0, right?
or is that somewhat what I should be checking for
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
I couldn't find anything for it, but i'll just post this chart here if I missed something
Unknown User•4mo ago
Message Not Public
Sign In & Join Server To View
Alright thank you
Here is my current design
i am using open traces on the front to add solder on top of, or solder solid core wire to, for more current
and I have a bunch of vias under the mosfets to transfer heat through the board better
I also made sure to leave the bottom side under the mosfets blank so I could put a large heatsink there and/or ontop of the mosfet bodies
Unknown User•3mo ago
Message Not Public
Sign In & Join Server To View
thank you, I'll remove those vias on the output terminals, but I might still keep the open tracks on the top
maybe I can put some thin soldermask "barriers" to make soldering easier
Soldering on the wires will be near impossible
I’m reflow soldering lugs like the old design so as long as they don’t slide around while soldering I should be fine
Ah I see what you mean
I could put some extra wire on all the open tracks while reflowing if I did want to do that
I had an idea
Do you think when the 12v back feeeded through the arduino, it put the REAL 😂 ATmega16u2 into high voltage programming mode and erased the code that was on it?
If that was the case I could try re-programming it with whatever serial to usb code it had?
apparently the ide comes with a hex file for the code on it, so I may try flashing it
Unknown User•3mo ago
Message Not Public
Sign In & Join Server To View
well i wouldn't know, im not even sure if that could happen
but I might try it anyway
I am double checking that I am using the right bootstrap capacitor
I took the input capacitance which was 4130pF and multiplied it by 10 to get 41.3nF, and I rounded up to the nearest capacitor value which was 47nF. Does this seem like it would work fine?
