Selecting a new H-bridge Driver!
HTML-код
- Опубликовано: 3 окт 2024
- In this video we discuss the SUPER tiny (1.5mmx1.5mm) SUPER cheap (0.14$ cheap (depending on the quantities)) PAM8016 driver and compare it to the DRV8837 (which was the other integrated h-bridge driver that i was using). Thank you PCBWay for sponsoring this project, check out there services: www.pcbway.com
PCB Design Software that I use: carlbugeja.com...
Free Version of Altium's Pro Version: www.altium.com...
PAM8016 - www.digikey.co...
DRV8837 - www.digikey.co...
Support me on Patreon: / carlbugeja
Cameras/Tools I use (affiliate links):
Main Camera: amzn.to/30tRgY7
Second Camera: amzn.to/2ZLm2wy
Light Setup: amzn.to/3jtqckA
Magnifying Lamp: amzn.to/3jrcaQv
Vernier Caliper: amzn.to/3eTju3I
Multi-meter: amzn.to/3hq6B2R
My Social Media:
/ carlbugeja
/ bugejacarl
/ bugeja.carl
/ bugeja.carl
/ carl-bugeja-0b922a135
Music:
Elevators Need Rock Too - Audio Library
Beach Walk - Audio Library
wow, I didn't realize so many people were into using chips that I would lose on the tip of my soldering iron; I'll have to up my game!
When engineering and shopping become the same thing...
If you have a safety razor on hand, the blade is a really good spatula for silk screening the paste onto the pads. The sharp edge of the blade makes a huge difference in print quality and works much better than conventional metal spatulas / credit cards with bigger corner radii. Do it in one go for best results. Cool video, keep it up!
These actuators would be perfect for ultra tiny rc airplanes.
No, because you need a massive magnet
@@kkuhn You'd be surprised. Electromagnetic actuators are already the standard for the smallest rc planes. It's not hard to find commercially available units that weigh only 300mg.
They have almost same part PA8013/15 as D class amplifier. On LCSC it cost US$0.08 @ 100pcs. I think these chips will work too.
nice! hmm the datasheets for it looks identical.. its just listed as a D class amplifier.. it is not listed on digikey though for low quantities
LCSC parts are always a coinflip on reliability and availablility. It's great for one offs and small batches, but you better not design a long term project without stocking up the parts in the first place. Also, sometimes there are complete batches of chips that simply doesn't work the way they're supposed to. I had this with a LED driver once - it was no problem because they managed to send me another batch of working ones for no cost, but on more complex projects this could end in costly bugfixing.
The power rating in the datasheet is not limited by the chip, but by the supply voltage minus drop across the switches and the load impedance (5 V supply with 0.5 Ohm switches allows max 3.1426 Vrms, into 4 Ohm it means the 3W when the sinewave is slightly clipped to 10% distortion).
So to judge the load capability, it is enough to compare the rated load impedance: The chip is designed for 4 Ohm, your coil has 17 Ohm, which alone means the driver won't be electrically overloaded in any way. To the contrary, it will have huge margin (barely 4x rated impedance means 1/4 current, so somewhere between 1/4..1/16 losses; exact number depends on the switching losses, but when driving it plain rail-rail they will be negligible)
Thermally it is a different story, because it is designed as a haptic driver, so to provide only short bursts and then stay idle and cool down, while your flapper is driven all the time without the cooling period. But here the margin acquired from the 4x higher impedance load will offset that, so this exact flapper should be safe.
結構前から英語字幕付けて見てるけど、3割程しか分からない。すごい高度な話をしてるんだろうな。
いつかはCarlさんみたいになりたいなと思うけど、一体どこで身につけたんだろう。
いつもいつも英語の勉強にさせてもらってます。有難うございます。
いつか、英語も回路も完全に理解できるようになれたらいいな。
これからも動画投稿頑張ってください。応援しています。
長文失礼しました。
Using Tj = Tc + °C/W*P , 0.19A flowing through 0.525Ω is about 0.019W of dissipated power. If Ta was 26 and Tc was 32 then the thermal coefficient would be around 317 °C/W very roughly from the one steady-state datapoint. It will most likely be lower at higher temperatures due to stronger natural convection.
So worst case with that 317 number, the chip could dissipate 0.33W at 125 °C, which is around 0.8A through both FETs and like 3.7W of 100% duty output power.
I love your work. interesting!!!
These types of videos always make me wish I was better at the ee side of this stuff
Carl, missed this video.. Not sure why. Pretty damn awesome component. I love going through data sheets and finding new components to spec out for projects. I was telling my wife how that is actually my fav aspect of electronics. I might need be able to make use of those little guys. Thanks for sharing. Crazy that thing can do all that with such high power and stay so cool.
Nope turns out I didn't miss it and I just am losing my memory. I had posted I think the day you had the video up.
Lucky 100th commenter! I'm glad I discovered your channel, or I would've never considered using flex PCBs for anything, and I wouldn't have known about this H-bridge.
Check mate ! I have extech too ! Put cardboard in the battery compartment to hold the batterys from falling
Cool, thanks for sharing your thought process and testing. Also seeing you succeed definitely motivates me to try some SMD soldering on my own ❤️ I am thinking about getting a hot air station now 🙃
Holly shit, holly shit, dude! Thank You! I was looking for something like that myself for a while.
Nice to see a video on the tiny chip you've been posting glamour shots of on twitter. Can't wait to see some new builds based off the new 'haptic driver' chip. At $0.24/chip, I might actually look into fully realizing my my pcb stir plate. Probably need to nail that down before I look into adding heat.
Cool! TI has a VCM driver with ringing compensation and closed loop current control but it is a low side switching one with a really low power handling capability.
Nice find on the part.
The datasheet is really strange tho, they don't give any thermal data for this package, so there is no way to calculate how much it would heat up itself from the power losses. I guess it was specifically designed for audio applications, hence the 4 Ohm and 8 Ohm load ratings in the datasheet. As long as you make sure your loads impedance is somewhere around there, i guess it would be okay.
i totally agree! the datasheet is very poorly documented which is a really shame considering its potential
If you use it pwm the drop on the transistors is quite low, so is the power dissipation. You should be fine.
It's fun to explore all those data sheets for better components, isn't it?
What led you to find this one? Just curious.
mainly just trying to find a cheaper option for some of my next projects :)
great! hope use in future
Question; instead of using magnets with your felxi pcb actuator, Have you tried another coil on the same pcb, ie 2 sets of opposing coils on one pcb?
You cant put 2 coils on the same pcb. He needs separate pcb because if the 2 coils are fixed on same material how do you expect them to move?
Edit: adding another pcb with coil could work but magnetic flux would be way to weak compared to neodymium magnets.
Here comes heat problems
the magnet-pcbcoil configuration works because of the strong magnetic field of the n52 magnet.. the coil's magnetic field is much smaller when compared to the one of the magnet so two weak magnetic fields won't oppose the coils .. this can work by maybe adding a core inside the coil but then it may get to heavy for the flexing :)
@@CarlBugeja There's a magnetic camera made from hall effect sensors floating around the web can't remember if it's open source or not but it'd be a nifty way to "see" that magnetic field if you ever felt like making one. Might be interesting to see if sharp zigzags or switch back curves help or hinder the field if you did make one. If you could get the cricut to make a halbach array well hello free lunch right? hackaday.io/project/18518-iteration-8/log/64376-a-magnetic-imager-tile
@@ransombot Sir, interesting..thanks for the heads-up.
This would be great for the control surfaces on very small RC airplanes. You could make the whole wing surface out of flexible PCB material. Flight controllers and VTX systems are down to a 16x16mm (hole distance) form factor now, but even a small microcontroller would do the job if combined with a barometer and inertial sensor. RC receivers are also very small today.
The problem is the energy needed. And think of it, it means as long as you hold the new shape you need that power.
Also this sort of motor isn't very strong.
For flapping a little foil around it's ok but for anything "work" that's useless.
The problem is the coil gets to far from the magnet.
First he would need to change the design from loose foil floating over a magnet to something common things.
Like the actuator in a hardrive moving the arm.
Or like a voicecoilmotor (laudspeaker)
Or look at CD-ROM drives or smartphone cameras. In booth cases the lenses are also actuated by coils and magnets
Awesome video Carl !
For ease of soldering and more flex strength, have you tried any of the cheap sot23-6 motor driver ICs like mx116L? Slightly poorer performance (higher RDSON) than the device you used but only $0.05 on LCSC. I've used the FM116 version on a number of DC motor products and they seem fairly robust. Keep up the good work by the way, it's fascinating seeing some of the novel things you're doing with PCBs.
Thanks! That part you've sent looks interesting but its a bit larger when compared to this one
Would you be able to make a galvo to control a laser?
Happy for you!)
5:00 Wrt pdf, the inputs are analog.. You can feed DAC output to them.. I would pull them to the ground instead of VCC..
Wouldn't the field be stronger with just one magnet? Or perhaps I should say more extended.
Its amazing idea man thanks
Wait, this IC should be soldered like BGA chip, not SMD...
Hello, what is that nice art on your wall? The black triangles art sculpture thing. Really nice!
great work ♥
Ah yes Greg Davill, such wonderful projects! 🙂
Nice video!
First of all good job on your vids! Quick question, what courses did you take to get this awesome knowledge?
_Hi Carl, how can you be sure that you're getting good reflow at _*_4:01_*_ on the center pad of the PAM8016?_
He should have pressed it down gently
Have you tried using a sound amplifier to drive a motor?
This is awesome bro.
I recently found your channel it's awesome
Just hats off🤗
did you design your wall of component drawers/containers above your monitor or did you purchase them? I really like your setup and wanted to see what you did.
DRV8837 is $0.18 for 1 and $0.10 for 1000 at LCSC with 3 day delivery and guaranteed to be genuine. Not available at TME, but highly recommend you search there for any components first, as they are in EU and charge very little for 3- day DHL shipping.
You have no reason to go to Digikey for anything in 2020.
I just hate BGAs. Can't see if they have solder problems, cant repair them without lifting the whole chip and just trying again, hoping for a better result.
Im just ranting. Great work on getting your costs and size down!
haha thanks :) its package is actually an lga though not bga (which i think is worse :P)
Thats small... and super cheap. Good job dude!
The only problem i see with it is maybe the difficulty to solder it reliably ? But you already did it, so i don't think it will be a big deal.
By the way, did you do it on your first attempt or no ? l
I did 🙂 (I also wasn't that confident with soldering it haha but the stencil makes it easier) but the soldering footage I showed was actually not the first attempt as the recorded video from the microscope got corrupted 😣
why no 4k quality
You are genial. The people genial as you are making evolution
Carl, love your work. Dead bug that little guy. lol
I am also want to learn these things. What kind of engineering department or any specific subject for this magnetic magic's, where did you learn this things I like this magnet matter. I need your suggestions sir!....😀
Electronics engineering will do I guess haha
PAM8016 stock is estimated in 1/2024
Cries in L293D
Does that H-bridge actually use N and P channel FETs? Usually they just use N channels for both, with a bootstrap circuit to power the high-side FET. But going by the differing voltage drops I can only assume it does, maybe that’s the difference between the high-power stuff I’m looking at and this low-power microactuator stuff.
It has an N and a P FET. It's giving you the different Rds for the high- and lowside FETs in the datasheet.
It does.. Using a pmos and nmos can basically allow you to use a simpler gate driver without the bootstrap circuit - eliminating the extra diode and cap
How did you generate the silkscreen?
Cool! you might find Dialog Semiconductor or GreenPAK products useful for some of your projects. Recently came across their products and am implementing them to make 36W COB UV LED curing lights with built in CV/CC boost converter, temperature monitoring, and controlling some led indicators.
Man I wish I was smart enough to tinker with this stuff!
The great thing about electronics is you choose the difficulty (at the cost of price).
Start out with breakout boards, where a one usually costs $1 to $5, and an Arduino Uno. These are great way to ease into the tech.
Pretty cool, it would be cool if you attempted to recreate some logic gates
Oof, 325mOhm Rds_on is pretty high :(
Has anyone seen a cheap half-bridge? Preferably with a bit more grunt.
Half-bridge?
how did you make that stencil O_O ?
The H-BIETCH
nice
I am trying to make my own lcd display but they only show how to make the screen could you make a video on the electronics please.
I want your robotic knowledge, what did you study ? mechanical engineering ?
Electronics engineering
Try making a flying PCB robot that uses a series of broad light weight PCB that flies similarly to a butterfly similar to the swarm bots using a lightweight 3D printed frame
Do you have any screen recorder software? If you don't, then _do not buy one._ Get OBS instead. It's free and it works.
Hello pro
So, you built a penile contraction simulator
Italiano?
Maltese 🙂
Physics for ALL peoples, Free!!!
This is a 5 million subscribers channel, not 51k.
Nnnooooooo not a pic !!!