|
Malcolm XML posted:We don't see exotic non silicon MOSFETs because silicon is dirt cheap to make and is pretty drat good until you need something extreme This isn't really a response to ANIME ACKBAR's earlier post. There is no GaAs or GaN insulated gate FET outside of research demonstrations because the electronic surface property of GaAs and GaN and related materials is not as good as silicon and an excellent electronic surface property is a necessary technical requirement to consistently make a good insulated gate FET. The only commercial GaAs and GaN FETs are Schottky gate devices. silence_kit fucked around with this message at 23:54 on Feb 15, 2017 |
#
?
Feb 15, 2017 23:00
|
|
|
|
| # ? May 9, 2024 16:01 |
|
|
I assume all this Gas nonsense is related to the magic smoke theory?
|
|
#
?
Feb 15, 2017 23:32
|
|
|
Splode posted:I assume all this Gas nonsense is related to the magic smoke theory? Lol, the acronyms and abbreviations for III-V semi-conductor technology are all terrible. III-V is a good one though
|
|
#
?
Feb 15, 2017 23:49
|
|
|
That's a terrible acronym, if you can even call it that. I assume you say it 3 5?
|
|
#
?
Feb 15, 2017 23:53
|
|
|
3 5 is way better than Gas, Mimick, pee-hempt, other common acronyms
|
|
#
?
Feb 16, 2017 00:02
|
|
|
Jamsta posted:2.1mm barrel jacks would do, and are widely used. I've got a bunch with screw terminals on their ends for re-use. Personally don't like them as they fall out relatively easily. Agree on 2.1mm barrel jacks. You can get them in solderable pigtail, solderable panel mount, solderable modular, and screwdown, all for very cheap. If you want to separate positive and negative, then go with banana plugs.
|
|
#
?
Feb 16, 2017 20:58
|
|
|
So I bought a kinda weird probably broken second-hand LCR meter and lately it's been acting weirder than normal... Every coil I try to measure with it seems to be coming up as having negative inductance. Is that even possible? What the hell does that mean, that it creates a... negative... magnetic... field? 
|
|
#
?
Feb 18, 2017 09:04
|
|
|
ate all the Oreos posted:So I bought a kinda weird probably broken second-hand LCR meter and lately it's been acting weirder than normal... Every coil I try to measure with it seems to be coming up as having negative inductance. Is that even possible? What the hell does that mean, that it creates a... negative... magnetic... field? I have no idea if you are using your LCR meter correctly or if it is broken or not, but a capacitor looks like it has negative inductance, if you measure it at only one frequency. A capacitor has negative values of reactance, kind of like your fictitious negative inductance. Similarly, an inductor looks like it has 'negative capacitance' at one frequency. The shape of the reactance vs. frequency curve for a capacitor and a fictitious negative inductor is different though. The reactance of a negative inductor becomes more negative at higher frequencies, while the reactance of a capacitor approaches zero at higher frequencies. silence_kit fucked around with this message at 17:55 on Feb 18, 2017 |
|
#
?
Feb 18, 2017 12:44
|
|
|
silence_kit posted:I have no idea if you are using your LCR meter correctly or if it is broken or not, but a capacitor looks like it has negative inductance, if you measure it at only one frequency. A capacitor has negative values of reactance, kind of like your fictitious negative inductance. Similarly, an inductor looks like it has 'negative capacitance' at one frequency. I have it set to L and it's displaying using units of H and it's connected to an object that is most certainly an inductor so... yeah Also sometimes its measurements fly around wildly (like fluctuating between positive and negative, shooting up to hundreds of henrys, or never settling on even the leftmost digit) so I'm leaning towards "it's broken" but wanted to post and see if I was just horribly misinformed about how the universe functions  e: I should point out it does similarly nonsensical things when I have it set to capacitance and measure a capacitor, and the weird behavior is different between different frequencies but still present
|
|
#
?
Feb 19, 2017 02:43
|
|
|
It's broken for sure
|
|
#
?
Feb 19, 2017 02:46
|
|
|
Oh man I just tightened one of the screws that connects the probe/clamp wire connector to the body and it started working fine again now I feel like a complete dumbass, sorry thread  e: For some reason if it's cable connector is not connected all the way it will display that bizarre behavior but if it is connected (but the probes aren't connected to anything) it will realize it's disconnected and just display ------, weird Shame Boy fucked around with this message at 02:56 on Feb 19, 2017 |
|
#
?
Feb 19, 2017 02:52
|
|
|
It's ALWAYS the connectors.
|
|
#
?
Feb 19, 2017 03:56
|
|
|
ate all the Oreos posted:Oh man I just tightened one of the screws that connects the probe/clamp wire connector to the body and it started working fine again now I feel like a complete dumbass, sorry thread Then I realized that there were plastic tips covering the leads.. I just thought it was a fancy inductance-based one! 
|
|
#
?
Feb 19, 2017 05:25
|
|
|
Since the discussion of SiC devices has come up, would a SiC MOSFET be better for linear regime operation than a comparable Si device? My naive thinking is that since MOSFETS running in linear mode run hot and SiC devices handle heat well they could be better but I haven't begun to really dig into the topic. I have a far future, maybe might happen, kind of project idea floating around that would involve trying to make a low-noise, high current DC power supply and perhaps SiC would be worth looking into if we could build a smaller thing that just ran a bit hotter. Efficiency would be a nice-to-have but not critical design concern in this case.Splode posted:That's a terrible acronym, if you can even call it that. I assume you say it 3 5? Yep, it's pronounced 3-5. The underlying idea is that silicon devices are made of (duh) Si which is in group 4 on the periodic table of elements, while things like GaAs are made of gallium (group 3) and arsenic (group 5) so they're 3-5 or using roman numerals III-V devices.
|
|
#
?
Feb 19, 2017 06:01
|
|
|
ate all the Oreos posted:Oh man I just tightened one of the screws that connects the probe/clamp wire connector to the body and it started working fine again now I feel like a complete dumbass, sorry thread Hey, don't feel bad at all. I once spent nearly half a day trying to troubleshoot a particular section of wiring I'd installed in a plane, wondering why even the basic continuity checks were showing wildly fluctuating values instead of the 0.1-ish ohms they should have. Turns out that an expensive, calibrated Fluke meter still doesn't work all that reliably when the battery gets too low.
|
|
#
?
Feb 19, 2017 15:46
|
|
|
ESP8266 testing not going to plan, can't even get an upload onto it despite the tutorials. So I decided to abandon the female to female Dupont cables I was using with the USB converter thing and get this on a bread board. I soldered it's plug to a perf board, word the copper side down, but now I need to solder the other side so it can plug into the breadboard with male headers. How do you actually make that solder though? It only makes a solder cap and doesn't melt down to the copper underside, which means it won't connect electronically. How do you solder pins on both sides of a perf board. This is getting a long time to get to the actually hard stuff!
|
|
#
?
Feb 20, 2017 01:35
|
|
|
Fat Turkey posted:ESP8266 testing not going to plan, can't even get an upload onto it despite the tutorials. So I decided to abandon the female to female Dupont cables I was using with the USB converter thing and get this on a bread board. Getting things set up is the hard part
|
|
#
?
Feb 20, 2017 05:16
|
|
|
Fat Turkey posted:ESP8266 testing not going to plan, can't even get an upload onto it despite the tutorials. So I decided to abandon the female to female Dupont cables I was using with the USB converter thing and get this on a bread board. I'm not sure I fully understand but it sounds like you're trying to solder veroboard on the non-copper side? That won't work, you put the solder on the copper side.
|
|
#
?
Feb 20, 2017 05:19
|
|
|
They sell veroboard with pads on both sides, which you should probably have instead of whatever it is you actually have.
|
|
#
?
Feb 20, 2017 06:34
|
|
|
why would you need copper on both sides?
|
|
#
?
Feb 20, 2017 07:13
|
|
|
Splode posted:why would you need copper on both sides? Mechanical strength.
|
|
#
?
Feb 20, 2017 07:44
|
|
|
Splode posted:why would you need copper on both sides? Because I need to solder something on both sides that would be connected. Basically making a plug so something can plug into the breadboard. Or not as the case may be!
|
|
#
?
Feb 20, 2017 09:57
|
|
|
Fat Turkey posted:Because I need to solder something on both sides that would be connected. Basically making a plug so something can plug into the breadboard. Normally, this would be done by soldering both parts to the same side and connecting with wire underneath. Otherwise, your perf board isn't doing anything. A short length of wire that's mechanically secure isn't going to have the same (if any) issues as a cable harness, if that's what you're worried about. Edit: if I'm grossly misinterpreting you, a picture or sketch would go a long way
|
|
#
?
Feb 20, 2017 14:01
|
|
|
Hey guys, Solar charging questions. I have two 12 V AGM batteries - one 105 Ah (single battery), the other 190 Ah (two 6 V batteries in series). I use them for astronomy (at night) and charge them the next day at home where I have AC power. I never deplete them to less than 70% of their original capacity. I want to go to more remote locations where it is darker (= better for astronomy). But before I can do that, I need a way to charge my batteries in a non-AC environment. Because of the noise and fumes, I have ruled out gasoline-based generators, so I am thinking of buying a solar panel+charger. Numbers: Total discharge: 75 Ah, @ 12V = 900 Wh Total charger efficiency: 50% Required energy: 1800 Wh Assumed sunlight: 6 h (summer and early fall in CA, NV) Required power: 300 W 1. Is it reasonable to assume an efficiency of 50%? 2. Would two of these (connected in parallel) cover my solar panel needs? 3. Is this charger ok? I can't find a dual channel (charge two batteries simultaneously) MPPT charger that can take a 320 W input. Thanks.
|
|
#
?
Feb 20, 2017 23:57
|
|
|
PDP-1 posted:Since the discussion of SiC devices has come up, would a SiC MOSFET be better for linear regime operation than a comparable Si device? My naive thinking is that since MOSFETS running in linear mode run hot and SiC devices handle heat well they could be better but I haven't begun to really dig into the topic. I have a far future, maybe might happen, kind of project idea floating around that would involve trying to make a low-noise, high current DC power supply and perhaps SiC would be worth looking into if we could build a smaller thing that just ran a bit hotter. Efficiency would be a nice-to-have but not critical design concern in this case. I don't think so because I don't think real life high temp is there yet. See if you can find a device rated for >175C, I havn't noticed any. But what it does seem is that SiC have terrible thermal coefficients so you can't get the heat out. In a switching application where they may generate less heat than a comparable device or where efficiency is important SiC can make up for that problem. But linear is inefficient by nature so the entire problem is removing the heat. In general for high power in the linear region make sure to check the safe operating area tables. Most devices have limits that are below just their wattage limitations alone.
|
|
#
?
Feb 21, 2017 02:33
|
|
|
theHUNGERian posted:Hey guys, I see AGM listed as having 99% charge efficiency (compared to ~90 for flooded) so 50% seems quite low which combined with the 6 hours, which is also conservative I think, makes me think you may have more than enough power. Note that charging multiple banks from one output is should be fine assuming they're the same type. This is quite common in the marine world where boats have multiple, often quite asymmetric banks charged from one alternator or solar system. The thing you don't want to do is leave them connected when they're not charging, as in that case an imbalance causes current transfer which is just wasteful.
|
|
#
?
Feb 21, 2017 02:46
|
|
|
asdf32 posted:I see AGM listed as having 99% charge efficiency (compared to ~90 for flooded) so 50% seems quite low which combined with the 6 hours, which is also conservative I think, makes me think you may have more than enough power. The efficiency I list was referring to the sun being at different angles with respect to the panel (the panel does not track the sun), the solar spectrum shifting, the temperature changing, the charger consuming some finite power, partial overcast, joule heating, ... . But if things are still conservative, I'm guessing I can just buy the panel and charger and connect them with simple cables and be done with it, right. No extra gizmos needed? The panel and charger I list are of a high enough quality that they will last a few years?
|
|
#
?
Feb 21, 2017 02:56
|
|
|
The modules you posted are 14.5% efficient, you should be getting panels with Sunpower cells in them like this: https://www.amazon.com/gp/aw/d/B01B3TEM1E/ Sunpower cells are 22% efficient, they are like the Intel of solar cells where no one else is currently competitive on efficiency/value with them unless you can find some smoking deal on used stuff.
|
|
#
?
Feb 21, 2017 07:10
|
|
|
Can someone tell me if this is a stupid idea: I have some lithium ion cells that almost certainly don't have protection circuits to avoid over-discharging them. Would I be able to make a cheap, simple protection circuit using a resistor, zener diode and a couple of transistors- the zener sets the cut-off voltage, below which it doesn't allow any current to flow, thus shutting off the transistors. I mean I realize it would be horribly inefficient and existing solutions would do a much better job but these are parts I already have and I don't feel like making another mouser order since I just did that last week Something like this:  Is there a better way I'm overlooking?
|
|
#
?
Feb 21, 2017 08:19
|
|
|
KnifeWrench posted:Normally, this would be done by soldering both parts to the same side and connecting with wire underneath. Otherwise, your perf board isn't doing anything. A short length of wire that's mechanically secure isn't going to have the same (if any) issues as a cable harness, if that's what you're worried about. Heh, I may be grossly misusing the perf board. I have an ESP8266 which has 2x4 pins, but putting that on a breadboard would have 2 pins in each row, no good! So building a plug that takes those 8 pins via a 2x4 socket into one side of the perf board, and then to solder male headers below so the 8 pins are now wide apart enough that they sit either side of a breadboard, each pin having its one row. That way I can make sure every pin is connected properly to 3.3/GND/etc I have a workaround which might work. I just want to start using ESP8266 properly without the messing around getting everything connected!
|
|
#
?
Feb 21, 2017 13:48
|
|
|
Zero VGS posted:The modules you posted are 14.5% efficient, you should be getting panels with Sunpower cells in them like this: https://www.amazon.com/gp/aw/d/B01B3TEM1E/ Are you commenting on my efficiency assumption? If so, isn't the panel efficiency already taken into account when the vendor specs the output power of the cell? In other words, the area of the cell is chosen to give 160 W. Or are you simply saying that the panels I posted aren't the best I could have?
|
|
#
?
Feb 21, 2017 17:09
|
|
|
theHUNGERian posted:Are you commenting on my efficiency assumption? If so, isn't the panel efficiency already taken into account when the vendor specs the output power of the cell? In other words, the area of the cell is chosen to give 160 W. Sunpower cells are much more energy for the area and more durable. I don't even know what's going on with your link because the same 160w Cynergy panel on Amazon appears to be much much larger: https://www.amazon.com/DOLSS-12vdc-Monocrystalline-Solar-Panel/dp/B01HS7BPO2 That's how big I'd expect it to be to produce 160w with only ~15% efficiency.
|
|
#
?
Feb 21, 2017 17:37
|
|
|
Fat Turkey posted:Heh, I may be grossly misusing the perf board. Use two-sided protoboard, or even better, stripboard (which has conductors running the same way), then use a bandsaw, dremel, or nibbler to cut it to size. Install female header topside in the middle two rows and male header bottomside on the outer two rows. Then the ESP can span the ridge in the middle of the breadboard. Or, ditch the 8-pin ESP8266 and just get a NodeMCU or similar host board with onboard USB, and everything will be 10 times easier.
|
|
#
?
Feb 21, 2017 17:41
|
|
|
ate all the Oreos posted:Can someone tell me if this is a stupid idea: That's a good circuit in theory but neither the zener nor the turn on voltage of the transistor is very accurate. You'll have on the order of 0.5V of variation at least which isn't too good (low voltage zeners are particularly bad). Look at replacing the NPN with a LMV431 to add accuracy (google for TL431 comparator to see how to use it this way). You'll need the LMV431 or another 1.25V TL431 variation to get it to work at this low voltage. You should also try to add hysteresis in this application with a resistor between the PNP output and the comparator input.
|
|
#
?
Feb 21, 2017 20:49
|
|
|
Does anyone know where to get the tool-side connector for the Dewalt brand battery packs? I have a project that could really use an already engineered battery pack, and I have some of the Dewalt 40v packs, but no nice way to use them without taking apart a charger to get the connector for them.
|
|
#
?
Feb 21, 2017 22:39
|
|
|
asdf32 posted:That's a good circuit in theory but neither the zener nor the turn on voltage of the transistor is very accurate. You'll have on the order of 0.5V of variation at least which isn't too good (low voltage zeners are particularly bad). I appreciate the tips but since I'm gonna have to order something anyway I just wound up buying a cheap little USB charge board that has discharge protection built in, oh well
|
|
#
?
Feb 22, 2017 00:14
|
|
|
Methylethylaldehyde posted:Does anyone know where to get the tool-side connector for the Dewalt brand battery packs? I have a project that could really use an already engineered battery pack, and I have some of the Dewalt 40v packs, but no nice way to use them without taking apart a charger to get the connector for them. You could also look for cheap/broken/used tools of ebay. Though 40V won't be too common so it might be hard. ate all the Oreos posted:I appreciate the tips but since I'm gonna have to order something anyway I just wound up buying a cheap little USB charge board that has discharge protection built in, oh well Makes sense. Consider adding the TL431 to your repertoire though. It's possibly the most mast produced and cheapest IC in existence and works as a reference/zener/comparator/op-amp from 2.5V (or 1.24V in some variants) to 30V.
|
|
#
?
Feb 22, 2017 01:04
|
|
|
Methylethylaldehyde posted:Does anyone know where to get the tool-side connector for the Dewalt brand battery packs? I have a project that could really use an already engineered battery pack, and I have some of the Dewalt 40v packs, but no nice way to use them without taking apart a charger to get the connector for them. There's a very good chance that it's on Digikey, or through one of the major connector manufacturers' catalogs. However, it won't be labelled as such, you'll have to take careful measurements and examine the engineering drawings very carefully. Fat Turkey posted:Heh, I may be grossly misusing the perf board. Yeah, I also recommend picking up a NodeMCU instead. Also use PlatformIO to program it, it owns.
|
|
#
?
Feb 22, 2017 03:21
|
|
|
Fat Turkey posted:Heh, I may be grossly misusing the perf board. Just so you know I have the same ESP8266 (-01 I think it's called?) with the dual-header and while I can get it to run and talk to it over serial I have never been able to successfully get it to run a program and I have no idea why not, even the example programs delivered in the exact standard way don't run on it. Maybe it's a counterfeit one or something idk
|
|
#
?
Feb 22, 2017 03:37
|
|
|
|
| # ? May 9, 2024 16:01 |
|
|
theHUNGERian posted:Are you commenting on my efficiency assumption? If so, isn't the panel efficiency already taken into account when the vendor specs the output power of the cell? In other words, the area of the cell is chosen to give 160 W. Yes, panel efficiency is taken into account wrt the power rating of the panel. A panel composed of more efficient cells will be smaller for the same power rating. If the area of the array is a concern then you want to look for panels with higher efficiency. Otherwise you don't care.
|
|
#
?
Feb 22, 2017 04:42
|
|