Jump to content
Search In
  • More options...
Find results that contain...
Find results in...

Hackentosher

Member
  • Content Count

    4,612
  • Joined

  • Last visited

Everything posted by Hackentosher

  1. it's possible there are transients on the 12v rail greater than the +/- 200mV you recorded that are occurring faster than whatever voltage monitor or voltmeter you're using can sample at. Larger transients may cause the hardware to freak out and shut down, but if you're reading fairly consistent 12V (usually +/- 200mV is fine for most electronics) with an external voltmeter I'm not so sure. To confirm, the power supply from your dad's computer passed the same stress test your power supply failed?
  2. The wires on each cell junction are balance wires. When you charge a series battery such as this one, you charge the whole pack to its full charge voltage, then drain the cells that are too high, bring up the whole pack back to full, drain the high ones, and repeat until all cells have the same voltage across them. You'll need to add balance wires to the additional two cells you add and get a 5s BMS or balance charger in order to charge properly.
  3. Well yes, but also no. The resistance of a conductor is given by the legnth of said conductor multiplied by the restively of the material divided by the cross sectional area of the conductor. So the lowest value resistor has a short length and a large cross sectional area (gauge). The voltage determines how much current can flow through a given conductor. If you have a 1 ohm wire and throw 120v down it's throat, by ohm's law the current through that wire will be 120 amps. The wire will try to pull that, but there may be other factors in the system that prevent this from happening. Ohm's law and the power equation can be rearranged into P=V^2/R, so that 1 ohm wire would dissipate 14,400W across it if given the opportunity. That's the equivalent of almost 10 space heaters! There are other parts of a circuit that limit the current, and thus power dissipated across wires and other conductors otherwise shit wouldn't work. Generally, you want your power wires to be large gauge so they are low resistance so that they don't drop very much voltage across them when delivering power to your circuit. Probably not, 2400W is a shit load of power, especially to dissipate in a wire. That heat probably can't go any where so it'll just raise the temperature of the wire higher and higher until it either melts, or auto-ignites something around it. Here you run into a different problem. Wire insulation has a property known as a dielectric strength, aka maximum electric field. A definition of how large of a voltage potential can be placed through a length of the material before the potential is too high and the material starts conducting. This is why you see some lengths of wire specify a voltage rating, or why electrical tape is only good to 600v. You would need a wire that has good enough insulation to withstand that much potential.
  4. I think I'm confusing the CH341 with the CH340 which I know to be a USB to UART adapter https://www.sparkfun.com/products/15096
  5. "SPI chip" could be anything, even an SD card. Though, a lot of microcontrollers are programmed over the SPI protocol. You pretty much need a programmer and/or arduino to talk to an SPI bus. I'd buy a 3.3v logic arduino, perhaps a SAMD21 based board from adafruit, or a cheapass STM32 board from ebay. Do your research on what libraries you may need to get either of these to act as a programmer, so you're probably best off buying an ISP or some other debugger.
  6. Silicon Labs cp2102 or an FTDI uart adapter from digikey, sparkfun, or adafruit is the way to go.
  7. It's really hard to tell with just that one poorly lit image. If I were you, I would scour my aliexpress purchase history to see if I could find it. Was there a packing list included with the package?
  8. Might be able to find some info on r/mk or geekhack.org. Probably just need to find a fab house on Alibaba that does good work. No offense to you or your work, but I like to use my own photographs for my wallpapers. A deskpad would be sick though...
  9. Yo the Art and Color Punch series would make good desk mats... I'd be interested if you ran a group buy.
  10. Wow, Fusion 360's animation work space really uses the GPU. That's exciting, most of all CAD workflows I've encountered barely even look at the GPU, this is the most utilization I've seen in the tool in the 6 or so years I've been using it image.thumb.png.d46491745d9541c2a93da7f736ae4dd1.png

    1. Show previous comments  1 more
    2. TVwazhere

      TVwazhere

      That's assuming you're us and are using Autodesk products that havent been updated to use more than a single core in twenty years lol. 

       

      Doesnt surprise that an animation workspace utilizes the GPU a lot. Finite Element analysis  and other forms of simulation are fully multi threaded which is good, but most things still only care about PIC and single core speed. 

    3. Hackentosher

      Hackentosher

      As far as I can tell modeling is also single threaded as well (which makes sense considering the linearity of the calculations that need to be performed, can't really work on a series of calculations out of order), which gets annoying when I start working with assemblies with a shit load of components or trying to manipulate a shitload of features at once. At least it's more intuitive than SW lol.

    4. TVwazhere

      TVwazhere

      I havent had an issue going between SW and Inventor but I've never had to use Solidworks for a job; just school. 

  11. I'm not trying to judge what you do with your money, but this seems horribly inefficient. Also, your numbers don't add up. You want to spend $20k on a rig, but have been saving $1k a month for a year. So you're $8k short? I think what I really have a hard time understanding is why you're so set on using these Titans when they are now two generations behind. You can get more performance for less money.. I just don't get it.
  12. It's allowed That enforces their point. If you're spending $10k on graphics cards (why tho, there are better cards for less money), proper cooling for them should not be an issue. You don't buy a luxury car if you don't have luxury car money. If you can find them, the blocks will probably cost you about $150 a piece, then you'll probably need a 4x120 rad at least to cool them all. I hope you haven't already bought the cards because you could probably get more, and more consistent, performance with a single 3090 for $1500 a card. edit: I hope you can flip them and get some 3090s, that will offer a much better experience in both the build and gameplay.
  13. The SG pin is for USB OTG. I think you're supposed to ground it or throw a pull down resistor on it, but I'm not sure. VCC is the power pin of USB, often also called VUSB or VBUS. VCC is a common term for the main power rail for circuits, especially when the power rail is common to several devices. The term VCC itself comes from transistors, it's supposed to stand for Voltage Common Collector, or the voltage you connect to several collectors in a multi-transistor circuit.
  14. Most workloads I face are primarily CPU bound (not games), and primarily single threaded as well. So, I generally choose the machine with the best single thread performance. Judging by Passmark, that appears to be the 4750U. https://www.cpubenchmark.net/cpu.php?cpu=Intel+Core+i7-1065G7+%40+1.30GHz&id=3466 https://www.cpubenchmark.net/cpu.php?cpu=AMD+Ryzen+7+PRO+4750U&id=3740. Games generally need more GPU horsepower, so the intel machine will probably be better for games. Personally, I use a Razer Core and a GTX 970 for games when I have time so GPU performance isn't that important to me. For Adobe suite, again I believe this is primarily CPU and memory based, so probably the 4750U machine would be better here. Battery life might be a toss up. Both chips are rated at the same TDP, but that's basically meaningless. I would look at reviews of both machines to see if you can get some real data on actual battery performance. If it was my money, I would get the Ryzen machine, but I get the sense that my needs and workflow are pretty different than yours.
  15. Oh duh, I've made forms for carbon layups before lol. here's some example of forged carbon fiber, it's common in car interiors and I believe they usually use an injection molding process to form the carbon-resin mix into a mold.
  16. woah that's interesting, but again what gives composites like carbon fiber their strength is the woven fibers that can distribute forces through the material from a point. I'm assuming that molded thing is similar to a "forged" carbon fiber, which from my understanding is chopped fibers mixed into the resin and then cast or injection molded. It retains the rigidity of normal carbon fiber because it uses the same resin, but it doesn't have the fibers so it'll be even more brittle.
  17. LOL I should have known, of course it's that easy.
  18. Structrual rigidity is pretty important for flight dynamics iirc, but you're right that a high speed impact is likely to crack carbon. However, you can mitigate the damage done by reducing the mass. There's less energy absorbed the arm if the system weighs less, meaning less likelihood of exceeding the material's limits. Frame design has also gotten pretty good, engineering the frames in such a way to distribute an impact to an arm across more parts of the frame, reducing the strain at a point. Maybe at a different scale, different materials would make more sense. But for 5" aircraft, I believe in CF.
  19. I beg to differ for a couple reasons. CF's strength to weight ratio is insane. It's also incredibly rigid, but can distribute the stresses across the mesh of fibers. Metals can't exactly do that. They're also heavier and bend, which isn't great for multirotors. Carbon will crack, but it takes a lot more of a beating before it does.
  20. CNCMadness is great! incredibly affordable, and the quality is great. They cut my Mad Hatter, that poor frame has been through so much and I've only broken a couple of bits on it.
  21. The issue I had was huge voltage spikes (probably on the order of 50-100v transients) would fuck with the flight controller and or ESCs, not really sure. All I know is sometimes I would death spiral after punching out and cutting throttle. Just something to be ready for, but I have a feeling the tech these days has caught up. 6S was a pretty new thing in FPV at the time.
  22. Oh damn, 6s is a lot of fun, but was hard to manage. The back EMF when you cut throttle is pretty insane, hopefully more modern electronics handle that better than when I last tried 6s.
  23. I think you'll have to get a donor, but it's possible Cooler Master support still has some spare parts for that case kicking around an office somewhere. You should send them an email, it wont cost anything.
  24. iirc the MX350 GPUs are barely better than intel integrated graphics. If it was my money, I'd prefer to have a laptop without a discrete GPU to save power.
×