|
Shelvocke posted:My intention is to regulate and rectify the non-cdi coils to 12v with a 110w max reg/rec and help smooth out variations with a large capacitor, as battery space is limited and these bikes tend to kill them. I'm also going to install a fuse box which was missing in the original harness. Before you get into any of your numbered questions: what voltage is the coil putting out? If you're 100% replacing everything, why not build a new lighting system from scratch in something more friendly? You could probably have your R/R output 5V or 3.3V and just use hobby RC stuff off-the-shelf. This seems to be a perfect use case for pre-designed modules you plug together. A bridge rectifier to get noisy, ripply DC, then a pre-built DC-DC converter that already has all of these design considerations implemented for you, so now you just need to plug in pre-built 6V or 12V LED lights. I built an arduino-controlled lighting system for my motorcycle that ended up being the size of a deck of cards (not including the DC-DC converter), used the existing switchgear, and consumed basically no power, using only "conversion" type LED lamps that fit in the factory housings. To directly answer your questions now: 1: skip this question with a pre-built regulator that gives cleanish power. Don't build your own R/R. 2: Always use a current-limiting resistor, and know that you may need heatsinking of some kind with high-power LED modules. Low-side switching is a good idea; it's what I used, because I could use easy and cheap power MOSFETs to do it. 3: In a box with vent holes is probably fine. The commercial DC-DC stuff is good up to 80C operating temp; any airflow or even decent fins is good enough. 4: If all that's on it is lights, you won't even notice 20% ripple current. Depending on the frequency of the alternator, you may not even notice 90% ripple current. One thing that comes to mind is maybe your coil system is going to be super unhappy with massively unbalanced currents in each of the coil windings.
|
#
?
Nov 9, 2019 16:22
|
|
|
|
| # ? May 9, 2024 18:03 |
|
|
Thanks for comprehensive reply. I've not tested it myself, but I believe the coil puts out anywhere from 9-15v. The capacitor was going to be in addition to the R/R, because the original loom had one and it seemed like a good idea. I like the look of the stepdown module - the harness needs to power horn, lights, and possibly phone charger, so having lower voltage could make sense. On resistors - I still need one if voltage drop matches supply?
|
|
#
?
Nov 9, 2019 16:47
|
|
|
Please do not buy an extremely bottom barrel dc/dc converter for something like running your headlights and taillights. You probably want something that is potted or covered with silastic so vibrations dont cause all of the component leads to metal fatigue and fall off. What else is on this harness that makes you need it to be 110w? If this is a bike that goes on roads or is used at night you are going to want way more than 9 watts of headlight unless you have predator vision. You cannot run series LEDs with no resistor (for much more than a second anyway) just by making the voltage close to the combined drop, the resistor is for limiting the current not for limiting the voltage. You should not run series LEDs for safety critical lights, because if one fails open circuit then all of them turn off, and if one fails closed circuit then the others will be on, but way over current and will likely fail because of it, so you should run them parallel with each having their own resistor.
|
|
#
?
Nov 9, 2019 16:54
|
|
|
The 110w is the maximum rating of the r/r unit I already have, not my intended draw. Sorry for not elaborating - the headlight unit I already have has lenses that will focus rows of 3 LEDs into a legal light image. There's room for 4 X 9w rows, 18w each for main beam and high beam. I admit I've struggled to grasp the resistor, it makes sense that it limits current.
|
|
#
?
Nov 9, 2019 17:02
|
|
|
You need to be able to withstand something like 80V in the event of a load dump. Get a DC/DC that can survive high input voltages or put a big honking zener diode in front of the DC/DC.
|
|
#
?
Nov 9, 2019 17:42
|
|
|
So this is probably a dumb idea and way beyond my electronics design capabilities, but I'm wondering what it would take to simulate a very simple neural network(just a handful of "neurons") in analog hardware. I wouldn't need to it to be "trainable" with back-propagation, and I'm picturing adjusting weights manually with potentiometers or swapping resistor values. But I'm wondering what is the simplest circuit that could be used to sum multiple analog inputs. I don't suppose it would be feasible to implement with just discrete components? I guess I was originally thinking that it would sum voltages, but maybe summing currents as inputs would make a lot more sense and simplify things to the point that I could use a single transistor for each neuron? peepsalot fucked around with this message at 21:24 on Nov 9, 2019 |
|
#
?
Nov 9, 2019 21:21
|
|
|
I remember a write-up from the Neurobytes guys talking about how the excitation current of (real) neurons works, and it seemed really do-able and fun in analog. They did it in code, but I think your idea has legs.
ante fucked around with this message at 03:37 on Nov 10, 2019 |
|
#
?
Nov 9, 2019 21:36
|
|
|
Yeah that sounds workable and fun.
|
|
#
?
Nov 9, 2019 22:31
|
|
|
I read some other article that was discussing doing it with memristors so all you need to do is get yourself some memristors. Easy!
|
|
#
?
Nov 10, 2019 03:34
|
|
|
BattleMaster posted:Kilowatts and whole amps are a different world from what I'm used to in radiation detection. I live in a land of gigaohms and kilovolts and picoamps. From a few pages back, but is the metal box itself connected to earth ground in any way? If not, you just have a big floating node that only has its own internal charges to work with to block passing EM noise. If the box is tied to earth ground it would have a much better source of charges to pull from to make shielding currents / arrange internal charges to counter external E fields. My guess is that your box is not earth grounded, then you add one 'ground' wire to an SHV line that is also not earth grounded so it doesn't help improve shielding much. Then you add the second ground wire to a line that is actually connected to earth ground and shielding gets a ton better.
|
|
#
?
Nov 10, 2019 16:08
|
|
|
The second SHV connector I added the ground to was to the detector, which isn't grounded. However, the other SHV connector and one BNC connector were going to a NIM crate (for an HV supply and tail pulse generator) which was definitely earthed, and the output (which had the first ground lug around it) was connected to a scope which should also be earthed but I'm not as certain as the NIM crate. Yeah that makes a couple of big ground loops, but all the radiation detector setups I've ever seen are a big mess as far as ground loops go and that wasn't even the problem anyway. The next time I build equipment I'm going to consider using isolated coax connectors on one side of each coax cable and grounding through only one wire, like in this setup here:  but I don't think I'm good enough at the ground thing yet to do advanced stuff like that, if I'm having trouble understanding wtf is going on here BattleMaster fucked around with this message at 06:38 on Nov 11, 2019 |
|
#
?
Nov 11, 2019 06:35
|
|
|
I've seen high end audio amplifiers that have a big switch to connect or disconnect ground. The idea is that if you have ground loop problems, flip the switch and see if that works, because grounding is mysterious. My first audio project started working better all of a sudden, and I realised that my audio negative cable had fallen out, fixing an accidental ground loop. poo poo is wack.
|
|
#
?
Nov 11, 2019 07:29
|
|
|
The high end audio gear probably has transformer coupled inputs if it has a ground lift function. DI boxes are often used for this function as well, especially when interfacing with things like instruments. There shouldn't really be a need for that feature if the transmission cable is balanced+shielded and grounded to PE at the source and receiver. The fact that it's common to need to float the ground on equipment used for live sound reproduction is pretty scary from an electrical safety perspective, since it indicates that it's common to have equipment with potentially hazardous ground/isolation faults.
|
|
#
?
Nov 11, 2019 13:44
|
|
|
Are those $50 long wei k3010d 30V10A regulated power supplies (for ex https://www.aliexpress.com/item/33030880652.html?spm=2114.12010610.8148356.2.60706f77U08ScH ) well-suited to electroplating? I wanna gently caress around with copper electroforming (like plating but different) for the purposes of making experimental short-run sheet metal 'rapid tooling' dies for use in the rubber-pad sheet metal forming process. For context, this is an old manufacturing technique that replaces the conventional two-part mating die (v expensive and finicky to make properly) with a single die and a pad/block of polyurethane rubber. The rubber acts as a universal conforming die that adapts to whatever tooling you throw in there. The technique is fantastic for prototyping and small production runs because it hugely simplifies the basic tooling requirements to start turning out parts. anyways, electroformed tooling is an established rapid tooling production technique in industry, but I see no reason why it shouldn't scale down to my needs. I want to see how electroformed tools stack up to my other rapid tooling approaches already in use, stuff like like building up dies from laser-cut acrylic slices or casting and recasting them using reusable low-melt fusible alloys. the process I'm planning on using is: - produce die model 'mandrel' using combination of stereolithography resin printing for the high-detail art/geometry parts and laser cut acrylic/hardboard for the base/lower-detail stuff - paint mandrel face with sealant and then with an unbroken coating of conductive graphite paint - hang conductive model from copper wire in electroforming bath, chuck a copper sheet anode in there as well - use regulated power supply to plate a heavy copper layer over the mandrel face - carefully remove mandrel from copper shell - fill shell with incompressible mineral/metal-filled epoxy - use completed composite tool to produce up to several hundred or thousand parts before the die wears out It's not actually too hard to get started with jewellery-scale work, the only expensive part is the regulated power supply. As I understand it, electroforming has very forgiving power supply requirements irt precision, you dial in the supply based on the surface area of the part and then fiddle with it during operation based on the plating appearance/colour. I will need more than 5A to do the parts I want, though, and most 10A supplies are out of my budget. the k3010d seems to meet my needs p well, and I dug up enough non-shill youtube reviews where they ran the thing at 9A draw for a week straight to be comfortable in at least some of these supplies working for hours without exploding. Ambrose Burnside fucked around with this message at 03:16 on Nov 13, 2019 |
|
#
?
Nov 13, 2019 03:12
|
|
|
Report back if you get something working I've done some gold electroplating and it's pretty easy, but not the same thing really
|
|
#
?
Nov 13, 2019 04:47
|
|
|
yeah electroforming is easy to "do" but very hard to do to a professional standard. getting an even plate thickness across a part w highly-variable geometry is apparently a nightmare, you gotta tailor the sacrificial anode's geometry to follow the mandrel contours closely + a specified offset. once the forming bath gets saturated good results become really hard to lock in, etc etc i'm gonna start with some artsy freeform stuff, i've got a background in silversmithing which dovetails nicely w the usual first exploratory project, forming an organic/freeform copper bezel onto a mostly-finished piece of jewellery with the stone set in place but not actually mechanically-retained by anything
|
|
#
?
Nov 13, 2019 06:01
|
|
|
I'm looking for sensor recommendation for measuring level of a liquid in a container. The distance from sensor to liquid will be less than 2m. If like accuracy of +/- 1cm or better. The environment will essentially be without light (does this matter for infrared type sensors)? I've played around with the HCSR04 sensor but measurements seem to float s fair amount. Would the Garmin Lidar be something to look at?
|
|
#
?
Nov 13, 2019 17:31
|
|
|
Mr. Bubbles posted:I'm looking for sensor recommendation for measuring level of a liquid in a container. The distance from sensor to liquid will be less than 2m. If like accuracy of +/- 1cm or better. The environment will essentially be without light (does this matter for infrared type sensors)? I've played around with the HCSR04 sensor but measurements seem to float s fair amount. Would the Garmin Lidar be something to look at?
|
|
#
?
Nov 13, 2019 17:52
|
|
|
Mr. Bubbles posted:I'm looking for sensor recommendation for measuring level of a liquid in a container. The distance from sensor to liquid will be less than 2m. If like accuracy of +/- 1cm or better. The environment will essentially be without light (does this matter for infrared type sensors)? I've played around with the HCSR04 sensor but measurements seem to float s fair amount. Would the Garmin Lidar be something to look at? They make ultrasonic tank level sensors, I'd give those a shot if a float sensor isn't viable.
|
|
#
?
Nov 13, 2019 18:02
|
|
|
CopperHound posted:How deep of a range are you trying to measure? +/-1cm is a lot easier when you are dealing with 30cm total range variation as opposed to 2m. Is sticking a float sensor in the liquid possible? I'd prefer not to use a float sensor. Range appx 5cm to 1m, now that I think more about it shouldn't need 2m.
|
|
#
?
Nov 13, 2019 18:28
|
|
|
you shouldn't have to spend more than idk $5-10 for this sensor, lots of budget options. given that you require short(ish) range and a decently-high precision as far as bargain-bin rangefinding goes, a ToF laser sounds like the best fit- except for the pesky fact that water doesn't reflect well enough to get an accurate reading. if you pointed it at a flat-topped float in a tube in the tank it'd work fine, but then you're introducing a moving part, albeit the type that should only be an issue if the water is allowed to freeze. cheap ultrasonic rangefinders will be pure non-contact, but they're also a lot noisier signal-wise and that +/- 1cm might be hard to hit with a $2 ebay ultrasonic rangefinder. theyre commonly-used for this purpose when the fluid is more inconvenient than water, though, so a sensor that meets your specs should do just fine. You could even DIY a conductance-based sensor with a resistor ladder and dozens of open connections stacked at 1cm or w/e intervals in a tube- the output voltage values will all correlate to a specific number of circuits being closed by water, giving you your water level. it can be as accurate as you have patience to put together. if you printed the ladder contacts onto PCBs and designed them to tile end-to-end while maintaining the needed spacing it probably wouldnt be too painful to put together. a float sensor will also probably do just fine for most applications, assuming this is in a climate-controlled location where freezing isn't an issue. they're not foolproof but they're very reliable, sump pumps all over the place already rely on them for emergency dewatering, they wouldn't use em if they weren't fairly reliable in that application. (search back a few pages, i'm doing water level sensing for a pump project rn so i've been wrangling with this) Ambrose Burnside fucked around with this message at 18:32 on Nov 13, 2019 |
|
#
?
Nov 13, 2019 18:28
|
|
|
What kind of liquid is it? There are options available for water that aren't suitable for stuff that's flammable or corrosive, for example.
|
|
#
?
Nov 13, 2019 21:22
|
|
|
Sorry, should have mentioned not water. It's a fuel, so definitely flammable if that changes my options.
|
|
#
?
Nov 13, 2019 21:23
|
|
|
Ambrose Burnside posted:you shouldn't have to spend more than idk $5-10 for this sensor, lots of budget options. given that you require short(ish) range and a decently-high precision as far as bargain-bin rangefinding goes, a ToF laser sounds like the best fit- except for the pesky fact that water doesn't reflect well enough to get an accurate reading. if you pointed it at a flat-topped float in a tube in the tank it'd work fine, but then you're introducing a moving part, albeit the type that should only be an issue if the water is allowed to freeze. cheap ultrasonic rangefinders will be pure non-contact, but they're also a lot noisier signal-wise and that +/- 1cm might be hard to hit with a $2 ebay ultrasonic rangefinder. theyre commonly-used for this purpose when the fluid is more inconvenient than water, though, so a sensor that meets your specs should do just fine. Good points. Maybe ToF laser would be an option I can look into.
|
|
#
?
Nov 13, 2019 21:25
|
|
|
Be careful with fuel and anything that isn't rated intrinsically safe.
|
|
#
?
Nov 13, 2019 21:27
|
|
|
You're the same guy who is trying to make something kind of like a sump pump but not really and it's definitely not for drugs or anything but you can't tell us what it is, right? I am not sure I really want to know what you're doing with a tank full of liquid fuel and a bunch of secret custom electronics
|
|
#
?
Nov 13, 2019 21:31
|
|
|
Sagebrush posted:You're the same guy who is trying to make something kind of like a sump pump but not really and it's definitely not for drugs or anything but you can't tell us what it is, right? I think you're confusing me for another poster? I've never tried to make a sump pump. Link me to what you're referring to. This is a project for measuring a fuel tank that is otherwise measured using a calibrated wooden stick. No secrets here.
|
|
#
?
Nov 13, 2019 21:38
|
|
|
If you can guarantee that it's only going to have fuel in it, you could measure the weight of the fuel tank. The normal solution for this sort of thing (eg measuring how full a bin is) is to use an ultrasonic distance sensor. For your range and accuracy, IR may be better.
|
|
#
?
Nov 13, 2019 21:46
|
|
|
Mr. Bubbles posted:I think you're confusing me for another poster? I've never tried to make a sump pump. Link me to what you're referring to. Okay, yeah, no hard feelings then. There's some other guy doing something mysterious with fluid level sensors that he's very cagey about. Anyway, they actually make fluid sensors that you just put into the liquid and they do their thing and that's it: https://www.adafruit.com/product/464 Sounds like you need more than 12 inches  but that company might make a variant that works for you.
Sagebrush fucked around with this message at 22:02 on Nov 13, 2019 |
|
#
?
Nov 13, 2019 21:52
|
|
|
I'm a bit confused about what I'm working with here. I bought some switches for a project I'm working on. The actual button is removable and there is a light below it. When I bought it on Aliexpress, it said it is an LED switch. When I took out my multimeter to check which direction the diode goes, it lights up in either polarity. Both directions have a forward voltage of 2.9 volts and light up the light. Is there some sort of magic going on inside this thing? The switch says 12V which I am assuming is the max voltage of the light or something. The switch itself is rated for 250V 5A, so I have to assume the 12V refers to the LED. So is there some sort of circuitry in there that allows it to be bidirectional? This is what the actual light looks like  Is this a normal LED? Do I need a resistor in series with this?
|
|
#
?
Nov 13, 2019 21:55
|
|
|
Mr. Bubbles posted:Sorry, should have mentioned not water. It's a fuel, so definitely flammable if that changes my options. Go buy something here: https://www.sick.com/us/en/fluid-sensors/level-sensors/c/g98155 Find a sensor that suits your needed depth and desired output with an ATEX approval. If a real sensor is out of your budget, walk away. Don’t DIY something in a fuel vapor atmosphere.
|
|
#
?
Nov 13, 2019 22:06
|
|
|
Alternate real-deal sensors: http://www.mtssensors.com/products/liquid-level-transmitters/index.html
|
|
#
?
Nov 13, 2019 22:07
|
|
|
What is the capacity of the fuel tank, and what is it mounted/installed on, and what is the fuel? A tank full of gasoline in the ground will yield a very different answer than a tank of heavy fuel oil on a ship.
|
|
#
?
Nov 13, 2019 22:16
|
|
|
Sagebrush posted:You're the same guy who is trying to make something kind of like a sump pump but not really and it's definitely not for drugs or anything but you can't tell us what it is, right? nah that was me, i just dont like to talk in an identifying way about jobs ive signed NDAs for but didn't obfuscate the details competently for the record, its literally a sump pump, yes Ambrose Burnside fucked around with this message at 22:22 on Nov 13, 2019 |
|
#
?
Nov 13, 2019 22:20
|
|
|
I'm fixing up an old VTVM and I was thinking that this thing could really use a fuse. How do I size which fuse to use?
|
|
#
?
Nov 14, 2019 01:37
|
|
|
Cojawfee posted:I'm a bit confused about what I'm working with here. I bought some switches for a project I'm working on. The actual button is removable and there is a light below it. When I bought it on Aliexpress, it said it is an LED switch. When I took out my multimeter to check which direction the diode goes, it lights up in either polarity. Both directions have a forward voltage of 2.9 volts and light up the light. Is there some sort of magic going on inside this thing? The switch says 12V which I am assuming is the max voltage of the light or something. The switch itself is rated for 250V 5A, so I have to assume the 12V refers to the LED. So is there some sort of circuitry in there that allows it to be bidirectional? They make bipolar LEDs that are just 2 diodes in parallel pointing in opposite directions. If it's intended for AC that makes a lot of sense.
|
|
#
?
Nov 14, 2019 19:09
|
|
|
hifi posted:They make bipolar LEDs that are just 2 diodes in parallel pointing in opposite directions. If it's intended for AC that makes a lot of sense. Ah, ok that makes sense. I did a bit more investigating and it does seem that one of those things lights up in one direction, and the other lights up in the other direction. Thanks.
|
|
#
?
Nov 14, 2019 20:08
|
|
|
So the 1961 radar sentry I picked up on ebay arrived. Some of you guys are into vintage stuff right? There's some sort of yellow (fluoro yellow, not at all orange) crust on the screws, it looks like a metal oxide from the screws corroding. Any idea what kind of metal corrodes like that? I'd like to clean the screws off, but I'd like to know what I'm dealing with first. I'm a bit wary because this thing ran off mercury batteries, which while not near the screws, is a nice reminder that poo poo from the 60s could be made out of all sorts of weird and wonderful poisons. Splode fucked around with this message at 09:55 on Nov 18, 2019 |
|
#
?
Nov 18, 2019 09:51
|
|
|
Splode posted:So the 1961 radar sentry I picked up on ebay arrived. Some of you guys are into vintage stuff right? I'm very interested in what you've got. Depending on the storage conditions, ferric oxide can be one of two colors: red (as rust) or bright yellow. So it's just a different flavor of rust. If you're worried, heat it gently with a torch. If the yellow stuff starts turning red or black, it's just rust; if not, then it's probably chromate or some other anti-sieze goop.
|
|
#
?
Nov 18, 2019 12:38
|
|
|
|
| # ? May 9, 2024 18:03 |
|
|
Splode posted:There's some sort of yellow (fluoro yellow, not at all orange) crust on the screws, it looks like a metal oxide from the screws corroding. Any idea what kind of metal corrodes like that? Uranium  
|
|
#
?
Nov 18, 2019 15:16
|
|