|
ullerrm posted:If I wanted to get a one-off PCB made for this, is KiCad the way to go? I used Eagle years ago, but it sounds like Autodesk acquired it and ruined it. I make boards for personal projects about once every year or other year or so. My claim about KiCAD is that if you're used to Eagle from the bad old days, KiCAD has all the same usability as I remember from Eagle in 2010, but at a price that fits! But no, in seriousness, I like KiCAD, and I think it'll be perfectly fine for this scale of thing that you want to build.
|
#
?
Sep 11, 2023 19:46
|
|
|
|
| # ? Jun 9, 2024 00:15 |
|
|
I'm actively enjoying KiCAD for my own one-off board project. Saying that as someone who last did this back in ~2014 via Eagle.
|
|
#
?
Sep 11, 2023 23:24
|
|
|
kicad has gotten good enough that people I work with (who have access to altium) are choosing to use it for prototypes and hobby boards.
|
|
#
?
Sep 11, 2023 23:46
|
|
|
I forget who said “just use an opto-isolator” but WOOOO https://i.imgur.com/zhZ5QZo.gifv Ignore the righteous mess on the desk, I was working on this while trapped in a late meeting, just crammed all the needed poo poo in
|
|
#
?
Sep 12, 2023 00:02
|
|
|
ryanrs posted:I'm going to guess 244 Hz. Out of curiosity, did they ever get back to you? I’m currently running it at 1000 Hz and it seems fine but I’m curious what the official spec is.
|
|
#
?
Sep 12, 2023 00:33
|
|
|
I need to stop trying to aim for perfection and just draft a PCB. I *always* do this when I get into a technical hobby. I’m drafting a power distribution PCB whose literaly only purpose is to take low DC voltage from one connector and route it to a bunch of other connectors, and instead I’ve been researching PCB design best practices and naming conventions and a hundred other things instead of just placing the components and sending it off for manufacture like five days ago, becasuse I’m worried I named something wrong or I violated some general best practice somewhere 
|
|
#
?
Sep 12, 2023 00:39
|
|
|
Bad Munki posted:I forget who said “just use an opto-isolator” but WOOOO Hell yeah Bad Munki posted:Out of curiosity, did they ever get back to you? I’m currently running it at 1000 Hz and it seems fine but I’m curious what the official spec is. I think you might be misunderstanding what he was trying to do some kinda jackal posted:I need to stop trying to aim for perfection and just draft a PCB. I *always* do this when I get into a technical hobby. Reading only gets you so far, the only way to really improve is to 
|
|
#
?
Sep 12, 2023 00:40
|
|
|
ullerrm posted:If I wanted to get a one-off PCB made for this, is KiCad the way to go? I used Eagle years ago, but it sounds like Autodesk acquired it and ruined it. This project is cool as heck, I hope you continue to post updates about it. KiCAD will work great for you, there's a ton of learning videos out there plus this thread to ping if you get stuck on something. OSHPark.com is my go-to for prototype boards when you're ready, you can literally drag-and-drop your KiCAD board file onto their website and they'll quote you a price for 3 copies. It's $5/in2 delivered. The Linux Fairy posted:I make boards for personal projects about once every year or other year or so. My claim about KiCAD is that if you're used to Eagle from the bad old days, KiCAD has all the same usability as I remember from Eagle in 2010, but at a price that fits! KiCAD has gotten a ton better in the last couple of years, I have free access to Eagle through my employer and still use KiCAD as my main go-to. They have really removed a lot of the UI jank and improved performance in versions 6+.
|
|
#
?
Sep 12, 2023 02:26
|
|
|
some kinda jackal posted:I need to stop trying to aim for perfection and just draft a PCB. I *always* do this when I get into a technical hobby. PCB design best practice is constantly changing, often a Matter of Some Debate, and often not particularly important. When it matters it matters but if you're not dealing with high speed signals, you don't have to pass EMC, and you don't have any high voltages, you have to try extremely hard to gently caress it up. If you're worried you can post it here and we'll give it a quick review but also Yolo. Also it's rare to get a design correct in the first revision, even for professionals, so if it comes back and it doesn't work, don't stress too much about it.
|
|
#
?
Sep 12, 2023 05:30
|
|
|
Thirding/etc. KiCad. I started using it for a four layer PCB I wanted to design (BGA chip, no way it'd be possible two layer with Eagle, which I'd tried previously) and it was similar enough to Eagle that I could figure it out but also free so I didn't feel so bad about any jankiness, and it's only gotten better since anyway. I also find it mildly cool to think about CERN having contributed fairly significantly to the project, makes me feel like my PCB is extra advanced. I've had PCBs work first try and ones that didn't too, thankfully it's always been something I could work around for prototyping. My number one fear applying power is that it'll burst into flames but that's yet to happen!
|
|
#
?
Sep 12, 2023 06:47
|
|
|
The thing that I find really freeing, is that if I have an off the shelf product that bursts into flames, I don't know specifically what went wrong, and I just have to eat the cost of the device usually. But, with things I designed, if I plug it in and it bursts into flames, I can go back to my design and figure out exactly what went wrong, work around it, and replace just the parts that broke - Often one $3 IC. Or, in the case of me, last week, looking at a board I designed in May and never got around to testing - Looking at my 12V silkscreens, looking up the main couple ICs datasheets and seeing that they're rated to above 24V, so plugging that in, instead. And then discovering, in an exciting fashion, that all my tantalums were 16V rated
|
|
#
?
Sep 12, 2023 17:33
|
|
|
I have a really dumb/fun electronic-adjacent question for the group. I'm building a grill temperature thing and needed some thermocouples. I got one of these from mouser (made by sparkfun) https://www.sparkfun.com/products/13715 in the description, they explicitly say: Note: This Type-K Thermocouple probe is not intended to be used for food. Can anyone think of why they explicitly say that? Is it a certification and/or compliance thing? I know that the K-type thermocouple dissimilar metals are most likely Nickel-Chromium and Alumel (Nickel-Aluminum?) neither of which would I want to put into my food. But those metals and wires are inside a casing is made out of stainless steel and that's pretty food-safe right? I'm not asking anyone to sign off on anything... I'm just curious what peoples thoughts are. I found this reference that says no you can't (although not because of the SS casing)... and gives links to a couple of food-safe ones that they say aren't expensive, but the link no longer works. https://cooking.stackexchange.com/questions/82691/are-hardware-store-themocouples-food-safe I can't find ANY that are less than $80 each...
|
|
#
?
Sep 15, 2023 03:04
|
|
|
Replacement Weber probes are considered food safe and while they aren’t billed as thermocouples I bet that’s all they are. Plus, if you build your thing with that stereo micro plug thing you don’t even have to modify them. Stainless probe with stainless braided cord and a little color coded silicon sheath thing. I dunno, just a thought. https://a.co/d/3wyLvvG
|
|
#
?
Sep 15, 2023 03:08
|
|
|
When the sous vide cooking (basically cooking food in a bag immersed in temperature controlled hot water) thing was a craze and commercial systems cost hundreds of dollars I built my own setup by plugging my already owned crock pot into a doohickey I made with a food-grade thermocouple in stainless steel housing, a PID controller, and an SSR to turn the heating element on and off. I got the thermocouple off Amazon for like $10 and the PID/SSR as a set for another $10. Thermocouple probe goes in the water, reports temperature to the PID, then the PID turns the SSR on/of as needed to keep the water at a set temperature. I guess I'm taking a random Amazon seller's word for it that it's really food grade, but it's encased in stainless steel and doesn't actually touch the food so likely not a problem?
|
|
#
?
Sep 15, 2023 03:20
|
|
|
Bad Munki posted:Replacement Weber probes are considered food safe and while they aren’t billed as thermocouples I bet that’s all they are. Plus, if you build your thing with that stereo micro plug thing you don’t even have to modify them. Stainless probe with stainless braided cord and a little color coded silicon sheath thing. hmmmm... I really wanted to build from scratch, but this is a really good idea. So is that plug really called a "Stereo micro plug"? I'll look into it, thanks! I have found some cheaper ones by searching for "NSF certified thermocouple" as directed by the link in my post. but that connector is brilliant.
|
|
#
?
Sep 15, 2023 04:15
|
|
|
namlosh posted:Can anyone think of why they explicitly say that? Is it a certification and/or compliance thing? Yes. It's not specified as food safe, so they say it's not intended for food. It's probably made exactly the same way as a food-safe probe, but isn't specifically tracked to ENSURE it's food-safe. It's probably fine.
|
|
#
?
Sep 15, 2023 04:41
|
|
|
I've been using a regular thermocouple for food for months, didn't even occur to me that it should be food safe. I haven't died yet!
|
|
#
?
Sep 15, 2023 07:37
|
|
|
Weird. It said there was a new post in this thread, but I had already read babyeatingpsychopath's post. Almost like a ghost posted or something.
|
|
#
?
Sep 15, 2023 07:45
|
|
|
namlosh posted:hmmmm... I really wanted to build from scratch, but this is a really good idea. Weber probes with the micro stereo plug are normally 10k NTC thermistors, not thermocouples.
|
|
#
?
Sep 15, 2023 08:00
|
|
|
I've started working on creating a DIY replacement for a "Flo by Moen" smart water meter and I've been obsessing over making sure the water-contacting bits (pressure sensor, electrically operated ball valve) were NSF61/food safe. I did consider YOLOing it but I know I'd lie awake at night worrying about it. I did eventually find a $70 pressure sensor from Honeywell on Mouser/Digikey that was NSF61, but it has an automotive connector and getting *that* seems to be a nightmare too. Digikey implied it came with the terminals but they marked it "obsolete" as well as "specifically for this sensor we still sell", and Mouser had the connector but had a vague implication it had no terminals, so I guess we'll see when it arrives! Still haven't found a valve, though one company insists their stainless steel one is when I asked them (but their site just says both "it's safe for drinking water!" and "beware the lead").
|
|
#
?
Sep 15, 2023 09:23
|
|
|
Charles Ford posted:I've started working on creating a DIY replacement for a "Flo by Moen" smart water meter and I've been obsessing over making sure the water-contacting bits (pressure sensor, electrically operated ball valve) were NSF61/food safe. I did consider YOLOing it but I know I'd lie awake at night worrying about it. What kind of valve? I use Grainger and McMaster for a lot of food-safe stuff. I also found out (by looking at my order history) that we buy type J thermocouples for our stuff. I also found an ancient box of type E, I assume for the liquid nitrogen blast freezer.
|
|
#
?
Sep 15, 2023 13:01
|
|
|
namlosh posted:hmmmm... I really wanted to build from scratch, but this is a really good idea. Rat Poisson posted:Weber probes with the micro stereo plug are normally 10k NTC thermistors, not thermocouples. Aww, well, maybe still useful with some adjustments to the build.
|
|
#
?
Sep 15, 2023 13:46
|
|
|
Bad Munki posted:Aww, well, maybe still useful with some adjustments to the build.
|
|
#
?
Sep 15, 2023 15:47
|
|
|
babyeatingpsychopath posted:What kind of valve? I use Grainger and McMaster for a lot of food-safe stuff. Basically something with NPT-style 3/4" threads and a ball valve, similar to the manual main valve on the house. That's how the Moen Flo works too (I could take it apart and use it, it also has a pressure sensor, but no idea how it's set up and I'd probably save money on parts for the project if I sold it, they're expensive enough that even used it'd cover the parts I'm buying). This is the one I found, their "sort by materials" page mentions the stainless steel version is "food safe", so it's probably fine. I'd just be happier if their actual product page actually said "NSF61" on it.
|
|
#
?
Sep 15, 2023 19:04
|
|
|
babyeatingpsychopath posted:Yes. It's not specified as food safe, so they say it's not intended for food. It's probably made exactly the same way as a food-safe probe, but isn't specifically tracked to ENSURE it's food-safe. It's probably fine. I assumed this, but wanted to get opinions. in this application, there's obviously no risk of bacteria or other contamination... my only worry is that it'll leech some type of metal into the food or something. But again, stainless steel is stainless steel I assume. Rat Poisson posted:Weber probes with the micro stereo plug are normally 10k NTC thermistors, not thermocouples. Oh thanks for this... I borrowed my neighbor's to use at one point and assumed it was a thermocouple. If it's not, then that changes stuff a little. From the looking around I did before, most thermistors have a top end that would be juuuuust below what I'm hoping for. In my charcoal grill, I'd like to read temps right at the grill surface (for searing) and that crap gets HOT. Like 600F hot by my IR thermometer... probably hotter. I'd want a safety margin too since I wouldn't want to melt it if the thing accidently fell into the coals or something. That's why I went with the Thermocouple to begin with... although I have to admit, I do think it sounds cooler and that might have been a factor as well. Might as well keep the "cool" factor since this is just a fun learning project, lol. Anyway, I'm going to assume it's safe and move forward. Thanks thread! e: anyone who's looked at this space has probably seen these things before: https://store-us.meater.com/products/meater-block I wonder what power source those little probes use... they have some strict rules about how to use them (they have to be inserted at least half-way into whatever meat you're cooking) which I assume is because the components inside can't take the heat. I've looked for a teardown but couldn't find one. It's a really neat implementation, but not really what I'm looking for and also $$$$ namlosh fucked around with this message at 20:11 on Sep 15, 2023 |
|
#
?
Sep 15, 2023 20:07
|
|
|
namlosh posted:I assumed this, but wanted to get opinions. I know with those Weber probes, they make a clip that holds the probe in the air right above the grill surface, so they're at least intended for that exact use.
|
|
#
?
Sep 15, 2023 20:13
|
|
|
Bad Munki posted:I know with those Weber probes, they make a clip that holds the probe in the air right above the grill surface, so they're at least intended for that exact use. Stay away from my “cool” factor!!!! But for real, Thx for the info though 
|
|
#
?
Sep 15, 2023 22:22
|
|
|
Ok, so you're looking to make a device that does something then shuts a valve off or turns it back on based on a condition? You don't necessarily need "NSF rated" things; potable water contact is sufficient. NSF is a whole set of things for bacteria resistance and washdown and crevices and all this other stuff that starts doubling price every time you add a condition. A motorized valve is probably overkill; they're typically servovalves that are designed to drive to a specified position based on a control signal. If you just want on/off control, then there are a couple of options. A motor and a couple of limit switches and attach that to a normal 3/4 valve like this, for example. Something like a window regulator out of a car that has torque sensors at end of travel could be a $free option. Depends on how professional you want the final product to look, I guess. Alternatively, a normal solenoid valve would work, either NO or NC depending on whether you want it to be open or closed if the power goes off. You could use a sprinkler valve for that: it's a potable-water rated, normally closed, full port, on/off valve. An industrial "thread this into a port and read voltage/current/i2c/rs485" sensor is going to be expensive, but a surface-mount transducer with a port on top can be in the $10 range. Do you need a transducer, even? Would a pressure switch work? I like the idea of this project, and the thingy that Moen has put out seems to be well-engineered to hit a price point. However, Moen has enough in-house manufacturing and engineering that they can say "we need X for a valve" and their valve department makes them one. Duplicating their product directly is probably not going to get you the same thing for even triple the cost, but using their idea as inspiration and making your specific device to do the exact thing you want in the exact way you want (safely) is well within pretty much anyone's abilities.
|
|
#
?
Sep 16, 2023 17:02
|
|
|
namlosh posted:I have a really dumb/fun electronic-adjacent question for the group. namlosh posted:I wonder what power source those little probes use... So I managed to luck into 30m of K-type thermocouple wire (cut into 3m lengths). So I built a thermopile with it to power a microcontroller. That's where I ran out of steam for the project, because while an MSP430 can run on nanoamps, I couldn't find a thermocouple amplifier that could, so I just have this thingy that you can stick in the grill flame and make an LED blink (dimly).
|
|
#
?
Sep 16, 2023 17:09
|
|
|
babyeatingpsychopath posted:So I managed to luck into 30m of K-type thermocouple wire (cut into 3m lengths). So I built a thermopile with it to power a microcontroller. That's where I ran out of steam for the project, because while an MSP430 can run on nanoamps, I couldn't find a thermocouple amplifier that could, so I just have this thingy that you can stick in the grill flame and make an LED blink (dimly). Wow, that’s ambitious and cool as hell… Reminds me of the RTGs the soviets made for radio repeaters and such for remote locations. These huge things, generate about 10 watts, lol. I assume they used the Peltier effect rather than seabeck though, but idk I’m pretty sure the meater probes have to be charged in the base station. I was just wondering what kind of rechargeable battery could deal with those temps. Id think that lithium would be completely out
|
|
#
?
Sep 16, 2023 18:05
|
|
|
babyeatingpsychopath posted:Ok, so you're looking to make a device that does something then shuts a valve off or turns it back on based on a condition? You don't necessarily need "NSF rated" things; potable water contact is sufficient. NSF is a whole set of things for bacteria resistance and washdown and crevices and all this other stuff that starts doubling price every time you add a condition. Yup - the reason I was looking for NSF61 is basically it's an easy shortcut to potable, but I realised that at least one of the product pages for the valve I linked earlier actually says it's NSF rated so I just ordered it. It was also listed as 5 wire - no control electronics, just limit switches and a motor, so easy to control with a microcontroller and an H-bridge (their other models have more going on inside and expect you to constantly apply power because it's easy, but I'd prefer being able to power off the motor when I'm not actively moving the valve), and it was just $40. I already got the $70 pressure sensor also NSF61, which I'm fine with, and the Chinese ultrasonic flow sensor is the most expensive part at $140 or so from aliexpress (absolutely doesn't need to be NSF61 as it's not in contact with the liquid). The latter actually also has analog inputs so I might just use that for the pressure sensor, which is a linear 0.5->4.5V output, if it's accurate enough, since it saves a bit of work as I need to read the RS485 anyway. Regarding duplicating their product, the cleverest thing it does is the slow leak detection - once per day (or per configured period, except their app doesn't work right for me so I can't change it) it turns the valve to turn off your water, and watches the pressure. Over a ~2 minute period, if it detects a drop, it reports a leak (and guesses how fast it is based on the change in pressure), then turns the water back on, so I'm certain I can reproduce that even if I don't have the maths they use to guess the size of a leak, but it does require the real, actual pressure sensor as opposed to a switch. It has a bunch of stuff that I don't have access to as I don't pay them like "detecting your water appliances" similar to the Sense product that monitors your home's mains current and guesses what appliances are, but I assume it's terrible (the Sense is). I actually have the Moen but dislike the app and cloud-only ness so my plan is actually to sell it if my DIY one works, and since they're $400 new hopefully actually cover the parts. Obviously won't cover the time put into it but since it's also for fun it's "free". The funny thing is I actually have a broken Moen someone gave me as they got a warranty replacement, so I could have used that valve, but I wanted to source everything from scratch. The part that was broken was its pressure sensor so it just constantly thought there were problems (they do live pressure monitoring as well), but the valve part was fine. I already checked and it'd be easy to control, the motor (and pressure) parts are on a separate PCB so it'd have been easy to control. I had had the intention of ripping the firmware from that one to figure out the ssh password and gain local access, but I think my electronics/embedded software skills exceed my firmware extraction and hacking skills.
|
|
#
?
Sep 16, 2023 19:05
|
|
|
Charles Ford posted:I actually have the Moen but dislike the app and cloud-only ness so my plan is actually to sell it if my DIY one works, and since they're $400 new hopefully actually cover the parts. This is all incredibly cool and good! Keep us posted! I'd be interested in an Instructables with BOM kind of thing.
|
|
#
?
Sep 16, 2023 19:11
|
|
|
babyeatingpsychopath posted:This is all incredibly cool and good! Keep us posted! I'd be interested in an Instructables with BOM kind of thing. I was actually just thinking it might make a good article, or YouTube video, or something, so I'll definitely make some notes on the way, and hopefully post any source code to Github/etc.
|
|
#
?
Sep 16, 2023 19:30
|
|
|
namlosh posted:I assumed this, but wanted to get opinions.
|
|
#
?
Sep 16, 2023 23:33
|
|
|
I'm designing a power supply that uses the LV2862, which comes in two flavors - 770kHz and 2.1MHz. I think I wanna go with the higher frequency one to get smaller components, and since higher switching frequencies are easier to smooth afaik. The datasheet has this to say:quote:8.2.2 Detailed Design Procedure It then doesn't actually tell you how the switching frequency is limited by the input and output voltages, or what the "frequency shift limitation" is. I'm looking at making three separate power rails of 3.3V, 5V and 12V from an input of 24V using 3 of these chips, how do I know what frequency to use? e: Oh also I think the higher frequency one is more efficient if I'm reading the charts right, and a goal of this project is to waste as little power as possible since it's going to run 24/7 and be tucked out of the way somewhere I'd rather not have it cooking itself with waste heat. Shame Boy fucked around with this message at 15:05 on Sep 18, 2023 |
|
#
?
Sep 18, 2023 14:56
|
|
|
Shame Boy posted:It then doesn't actually tell you how the switching frequency is limited by the input and output voltages, or what the "frequency shift limitation" is. I'm looking at making three separate power rails of 3.3V, 5V and 12V from an input of 24V using 3 of these chips, how do I know what frequency to use? I can't see any other way to interpret the datasheet. The example in 8.2.2 doesn't say that the 770khz part was necessary for those parameters.
|
|
#
?
Sep 18, 2023 15:48
|
|
|
Captain Cool posted:The datasheet gives tON-min as 80ns. At 2.1mhz the period is 476ns. So your minimum duty cycle is 17%. That means you'll have to use the 770khz part for your 3.3v rail. Okay I think that made it click, thanks. Captain Cool posted:The example in 8.2.2 doesn't say that the 770khz part was necessary for those parameters. I guess it's confusing because it makes a point of saying "you probably want to use the higher frequency one if you can" and then uses the lower frequency one
|
|
#
?
Sep 18, 2023 15:51
|
|
|
Shame Boy posted:I guess it's confusing because it makes a point of saying "you probably want to use the higher frequency one if you can" and then uses the lower frequency one
|
|
#
?
Sep 18, 2023 16:24
|
|
|
Shame Boy posted:e: Oh also I think the higher frequency one is more efficient if I'm reading the charts right, and a goal of this project is to waste as little power as possible since it's going to run 24/7 and be tucked out of the way somewhere I'd rather not have it cooking itself with waste heat. The datasheet shows better efficiency for 770kHz at vin=24V/vout=12V. Lower switching frequencies help with efficiency by reducing switching and gate drive losses. I wonder if they recommend the 2.1MHz version because it's the one that's not right in the middle of the AM broadcast band while showing more measurements from the 770kHz version because it tends to perform better.
|
|
#
?
Sep 18, 2023 16:42
|
|
|
|
| # ? Jun 9, 2024 00:15 |
|
|
Texas Instruments has the absolute worst datasheets. That kind of confusing gotcha is rampant. I've started avoiding using their parts, particularly their switchmodes, because I'm sick of playing "find the trap" when every other manufacturer is up front about what their chips can and cannot do. Especially when I'm generally choosing parts in a parametric search like Digi-Key - TI lying in their parameters is infuriating.
|
|
#
?
Sep 19, 2023 01:14
|
|