|
I was thinking something along the lines of this would work, and it only needs a few components: Of course you are correct that two PFETs would be needed if you couldn't tolerate the Schottky drop on the 5V USB. Also, if there are any switching time or protection requirements it'd probably be best to look into a more specialized switch or switch management IC, like the LTC4412.
|
#
?
Apr 13, 2013 13:39
|
|
|
|
| # ? Jun 3, 2024 21:06 |
|
|
Boy With Stick posted:I was thinking something along the lines of this would work, and it only needs a few components: Yes I wanted to come in and post exactly this after you had me thinking last night. The switch only needs to "switch" in one direction - prevent current going from 5V to 3.3V so from that perspective you only need one. The diode is needed to get the control right. I'd actually consider using a second PFET simply as the diode to avoid buying a new part, although it's a bad diode, but that may or may not matter (I don't think the OP cared about voltgage drop from 5V). Note that because you're switching 3.3V you'd want a PFET with a low VGS threshold, often labeled "logic level".
|
|
#
?
Apr 13, 2013 15:35
|
|
|
asdf32 posted:Yes I wanted to come in and post exactly this after you had me thinking last night. The switch only needs to "switch" in one direction - prevent current going from 5V to 3.3V so from that perspective you only need one. The diode is needed to get the control right. Wouldn't hooking it up like this cause the battery to constantly slowly drain even when it's plugged in to the 5V though? I don't know if they care but I think that would suck personally.
|
|
#
?
Apr 13, 2013 18:46
|
|
|
Parallel Paraplegic posted:Wouldn't hooking it up like this cause the battery to constantly slowly drain even when it's plugged in to the 5V though? I don't know if they care but I think that would suck personally. The load it's attached too would be at a higher voltage when 5v is present. So it can't drain then. When 5v isn't present it's true that this circuit won't isolate the load. If you also want an on/off switch for the battery then a second fet or other solution is needed.
|
|
#
?
Apr 13, 2013 19:28
|
|
|
asdf32 posted:The load it's attached too would be at a higher voltage when 5v is present. So it can't drain then. When 5v isn't present it's true that this circuit won't isolate the load. If you also want an on/off switch for the battery then a second fet or other solution is needed. Wait so if the 5v is present the 3.3v wouldn't flow at all? I always thought if you had two power sources connected they'd both discharge current but you'd only see the "higher" voltage if you measured it, but then again I'm still fairly new to this  Sorry to kind of draw away from the core issue, but now I'm curious - what would happen if they were both 5V and the paths to the load had equal length and resistance etc?
|
|
#
?
Apr 13, 2013 19:41
|
|
|
Parallel Paraplegic posted:Wait so if the 5v is present the 3.3v wouldn't flow at all? I always thought if you had two power sources connected they'd both discharge current but you'd only see the "higher" voltage if you measured it, but then again I'm still fairly new to this Taking out the semiconductors, just look at the potentials at each node and the resistance between them (ie the 5v potential is going to be flowing towards the 3.3v)
|
|
#
?
Apr 13, 2013 20:14
|
|
|
ante posted:Taking out the semiconductors, just look at the potentials at each node and the resistance between them (ie the 5v potential is going to be flowing towards the 3.3v) Yeah but I didn't think that would "push back" the 3.3v, but now that I think about it I guess it doesn't really make sense to have current flowing in two different directions simultaneously, even if they are from different sources
|
|
#
?
Apr 13, 2013 21:25
|
|
|
It always flows from places of higher potential to lower, regardless of whether someone called that node a "source" or not. The one thing to worry about is whether a particular source can handle reverse current without exploding. E.g non-rechargeable batteries and some power supplies.
|
|
#
?
Apr 13, 2013 23:03
|
|
|
Thanks for the feedback guys, the idea of throwing a diode on the 5V rail (where I can tolerate the drop I'm pretty sure) never occured to me for some reason  Found a pretty good 450mV diode I can throw on there, which should still allow the AVR to run fast enough to act as a USB device (4.5V - 5.5V is the range for 0-20MHz operation). Now, I'm planning on adding a male USB connector to this Digispark style (i.e., the PCB acts as part of the connector), anything I should know about doing this? I'm going to try to at least back-of-envelope go for impedance matching and stuff, but this is a dirt cheap learn-how-solder board, soo... e: or I guess the cost isn't /too/ bad for connectors, I could just throw a mini or micro B on it. movax fucked around with this message at 17:24 on Apr 14, 2013 |
|
#
?
Apr 14, 2013 17:17
|
|
|
I would go for the connector just for the "not having the cable fall off all the time" factor.
|
|
#
?
Apr 14, 2013 17:44
|
|
|
evensevenone posted:I would go for the connector just for the "not having the cable fall off all the time" factor. So I was thinking: 1. Have board have the connector as part of it, downside is that this may increase PCB cost / complexity, 2. Put a Mini-B connector on the board, needs to be through-hole and requires a cable (or people could just use their cell phone cable...) 3. Put a Male A connector on the board, through-hole, would look kind of weird, but you can just plug it into your PC, no issues
|
|
#
?
Apr 15, 2013 04:00
|
|
|
I don't think Mini-B through hole connectors exist*, but you'd want Micro-B anyway since that's the connector that has been adopted by the EU. A male A connector would probably be the easiest to find and use. *Feel free to prove me wrong, I want a through hole Micro-B connector or 5 pin breakout.
|
|
#
?
Apr 15, 2013 06:55
|
|
|
Ephphatha posted:I don't think Mini-B through hole connectors exist*, but you'd want Micro-B anyway since that's the connector that has been adopted by the EU. A male A connector would probably be the easiest to find and use. I don't know about through hole, but Sparkfun has MicroB breakout boards as well as 5-pin Mini-B.
|
|
#
?
Apr 15, 2013 07:01
|
|
|
Ephphatha posted:*Feel free to prove me wrong, I want a through hole Micro-B connector or 5 pin breakout. http://uk.farnell.com/jsp/displayPr...%7Bplacement%7D e: guess I didn't read well... oh well, have a micro-B thru hole too... http://uk.farnell.com/jsp/displayPr...%7Bplacement%7D (looks goofy as hell tho) SybilVimes fucked around with this message at 12:39 on Apr 15, 2013 |
|
#
?
Apr 15, 2013 12:36
|
|
|
I picked up some PCB mount Micro-B's the last time I ordered from Vetco and they look a lot like that e14 offering. The pins are teeny tiny and meant for surface mounting I suppose. I only needed power, so I just broke the middle pins off and soldered normal right angle header pins to Vcc and ground. A little HST to keep them from shorting and I had through-hole jacks. All right, so I've been doing a lot of research work on my Propeller Clock over the weekend. I've been thinking, is there a reason I can't use a D/A to control the brightness of the LED instead of PWM? I'd need to get something that had the ability to sink or source a decent amount of current (50mA/channel), but it might be a reasonable way to do it. I did a test last night with a white LED hooked up to my power supply, after some fiddling I found a current limit I liked that corresponded to a good top end voltage range. I found the LED came on at 2.35V and achieved maximum current usage (and therefor brightness) at 3.35V. Varying the voltage over this 1V range gave me 100 steps of brightness. To me, this seems like a good solution compared to PWM for my needs. Does anyone know of a chip that will fit the bill here? I found a few last night but they were either prhibitivly expensive or only 1 channel. At minimum I'd like a 3 channel version, but ideally it would be 8.
|
|
#
?
Apr 15, 2013 18:21
|
|
|
HATE TROLL TIM posted:I picked up some PCB mount Micro-B's the last time I ordered from Vetco and they look a lot like that e14 offering. The pins are teeny tiny and meant for surface mounting I suppose. PWM is likely going to be cheaper and easier, though. At the PWM frequencies you'd need to use, it is essentially acting as a D/A converter, using your visual response as the LPF. You have all the same problems with D/A that you have with PWM, and then more on top. You just need to choose a micro with adequate PWM peripherals in hardware. I think some of TI's Piccolo series has good PWM hardware, but I haven't used them personally. D/A is just way too expensive for the same effective resolution. It's way less work to get muliple 32 bit PWMs going with like 10 MHz or whatever base clocks and going from there.
|
|
#
?
Apr 15, 2013 18:42
|
|
|
If you're adding USB connectivity to a device, do everything you can to avoid the mini connectors. The fullsize ones are fine. The micros are also fine. The mini connectors are _horrid pieces of shit_. I've made good money resoldering miniB's back on to 2.5" externals right before finals time.
|
|
#
?
Apr 15, 2013 19:56
|
|
|
Slanderer posted:PWM is likely going to be cheaper and easier, though. At the PWM frequencies you'd need to use, it is essentially acting as a D/A converter, using your visual response as the LPF. You have all the same problems with D/A that you have with PWM, and then more on top. You just need to choose a micro with adequate PWM peripherals in hardware. I think some of TI's Piccolo series has good PWM hardware, but I haven't used them personally. Hmm, interesting. The only reason I was thinking the D/A might work is because I did end up hooking a LED to an MCP4802 I had on hand (the output of which I fed into a transistor to raise the output voltage). I had put together a simple test rig last week consisting of a wheel attached to a motor with a tiny breadboard attached and a 3.7 lipo underneath, with thin lead weights on the opposite side to counter-balance the whole thing. When I ran this D/A setup on it, I didn't seem to get any smearing like I got with PWM. But it would be complex and expensive to implement 48 D/A chips (in a 16 LED setup), plus the support hardware (transistors, resistors, etc)... I guess a PWM solution it is. I'll check out the TI stuff, thanks for that. I mean, I'd be happy with 16 steps of control per channel (which would be like 4096 colors).
|
|
#
?
Apr 15, 2013 23:01
|
|
|
 Heat Question! This is more of a communications/general electronics thing: I have a Motorola SB6180 cable modem. It works great, but I've noticed it gets pretty hot at the top of the case. The metal back section where the plugs are are lukewarm, the cables are cool, the wall-wart is warm (but not excessively so), but there's a considerable amount of heat coming out of the cable modem. It is NOT hot enough so that it cannot be uncomfortably held. But it's a lot more heat than my DSL modem (the piece of poo poo that it was) put out. I'm not sure yet if the heat level is constant or depends on how much data is being received and transmitted. I've read reports of people with Motorola Surfboards running the gambit from "cool" to "very hot". There are no obvious signs of a malfunction, like the modem locking up, connection problems, or anything like that. No melting plastic or magic smoke smells. I was talking to my dad, and he thought the DSP in the modem might be generating the heat. I've seen elsewhere mentioned that the cable tuner circuit can run hot. Any ideas as to what specifically will be creating all the heat in the modem? There is a reasonably-sized heatsink near the top, so I've deduced that's where the heat is coming from. It's about one square inch, and the fins are arranged so there should be a nice chimney-effect as far as cooling goes when the unit is placed vertically.
|
|
#
?
Apr 16, 2013 00:36
|
|
|
HATE TROLL TIM posted:Hmm, interesting. The only reason I was thinking the D/A might work is because I did end up hooking a LED to an MCP4802 I had on hand (the output of which I fed into a transistor to raise the output voltage). I had put together a simple test rig last week consisting of a wheel attached to a motor with a tiny breadboard attached and a 3.7 lipo underneath, with thin lead weights on the opposite side to counter-balance the whole thing. When you did your earlier test, do you know what frequency PWM you were using?
|
|
#
?
Apr 16, 2013 00:55
|
|
|
Slanderer posted:When you did your earlier test, do you know what frequency PWM you were using? Just found it, 100Hz. For kicks I tried it with a new, high-speed hardware PWM driver for the RPi at 1kHz and surprise surprise, it didn't smear. Honestly, I thought the initial test was at 1kHz (1000Hz), not 100Hz. I think I can get Pi-Blaster up to 10k if I really lower the resolution, but I can't do a rotation test with this (because I can't really mount the Pi on the wheel). I've got a couple of samples coming from NXP, Maxim and TI; I think two of the chips claim a 97kHz frequency. I wish there was a way to just easily calculate how high of a frequency I need (and how low of a duty cycle) based on RPM and resolution.
|
|
#
?
Apr 16, 2013 18:35
|
|
|
HATE TROLL TIM posted:Just found it, 100Hz. For kicks I tried it with a new, high-speed hardware PWM driver for the RPi at 1kHz and surprise surprise, it didn't smear. Honestly, I thought the initial test was at 1kHz (1000Hz), not 100Hz. The calculations for this stuff might be vague, or at least governed by complex models, due to weirdness in vision. We can't perceive high speed variations in brightness if the "off" time of the PWM is sufficiently small---but our eyes can perceive extremely short bursts of light pretty well, which is the effect you get with slow PWMs and low duty cycles (short bursts followed by long dark periods). In order to get sufficient brightness resolution, I'd think at least 100kHz would work. But if you need good spatial resolution as well, you'd need to increase that by possibly orders of magnitude. Can you remind me what the size and speed of the rotor you'd like to use, and what angular resolution you're trying to achieve (or at least just approximately what sort of resolution, if you wanted to draw a raster).
|
|
#
?
Apr 16, 2013 19:25
|
|
|
HATE TROLL TIM posted:Just found it, 100Hz. For kicks I tried it with a new, high-speed hardware PWM driver for the RPi at 1kHz and surprise surprise, it didn't smear. Honestly, I thought the initial test was at 1kHz (1000Hz), not 100Hz. As a heads up, if you're trying to do PWM at 100Hz (especially if you're rotating it), you're going to end up with very non-linear luminance adjustment due to the Broca-Sulzer Effect and the Brücke-Bartley Effect (and probably some other stuff I'm too tired to think about right now). Whether or not that matters for your application, I have no clue.
|
|
#
?
Apr 17, 2013 06:26
|
|
|
Three-Phase posted:
DSP, line drivers for the cable output (basically a medium size radio transmitter), internal power supply. Really the only way to tell is to pop it open and try feeling around! Is there a reason you want to know or is it just curiosity?
|
|
#
?
Apr 18, 2013 09:32
|
|
|
Slanderer posted:The calculations for this stuff might be vague, or at least governed by complex models, due to weirdness in vision. We can't perceive high speed variations in brightness if the "off" time of the PWM is sufficiently small---but our eyes can perceive extremely short bursts of light pretty well, which is the effect you get with slow PWMs and low duty cycles (short bursts followed by long dark periods). This is all still on paper, but I'm thinking 5" for the total rotor length. There will be 16 piranha RGB LEDs, each taking up 1/4" for a total of 4" length. I'm not sure on angular resolution, I was thinking 5 degrees? As for the PWM for the LEDs, I'd be happy with a resolution of 4-bits per channel; 16^3 would be 4096 colors, which is a hell of a lot better than the 8 colors I'd get without PWM. I'm also not sure on rotational speed, I suppose slower would be better? I could deal with 15Hz I suppose, but 30Hz would be a lot better. I don't think I can go above 3500RPM (60Hz) without vibration and balance issues. I just need to find a solution that lets me get into the hundreds of KHz for PWM.
|
|
#
?
Apr 19, 2013 03:50
|
|
|
I've been drooling over this "Fuel" line of Milwaukee lithium-ion cordless power tools - the "M12" (12 volt), "M18" (18 volt), and "M28" (28v) series. Which brought up a couple of questions I've been pondering over the years. 1) What's the advantage of a higher voltage cordless tool? I always figured the amount of current, not voltage, would be the determining factor when it comes to power (power in terms of, like, torque or speed). Or should I be thinking about it with Ohm's Law in mind? In other words, a cordless tool with a 28v battery can draw up to (theoretically, of course) -> 28 amps (of course A LOT less current because of all the resistance, but the point I'm trying to make is "Is it true that a 28v battery is naturally capable of delivering more current at any one moment than an 18v battery? And is that why a higher voltage power tool would be preferable?") Batteries have ALWAYS confused me; sometimes they don't seem to follow Ohm's Law. For instance, I've seen a lot of 12v lead-acid car batteries that say something like "500 cold cranking amps". But how is that possible? It says it right on the battery, which implies that there's no kind of voltage increasing circuitry, and even if there was a transformer, you can't transform DC, can you? I'm utterly baffled by this. 2) It looks like the battery packs (there's a cutaway diagram in the brochure) are made up of AA sized Li-Ion cells. So one of these Milwaukee "Fuel" M12 12v battery packs would have less cells, in series, to provide the 12v the tool runs on; whereas a M28 28v has either more of those cells, or bigger, higher voltage ones. But does this mean a 28v tool battery necessarily has more AH than a 12v battery? Although, of course a 28v tool surely uses more charge, more quickly. 3) Would it safe/ok/a good idea to use, like a 12v lead-acid rider lawnmower battery wired (with the correct polarity of course) to the terminals of a 12v cordless power tool where the tool's normal battery's own terminals usually connect? Or would I need to build some kind of current or overloading protection circuit? When I was in Radio Shack one time asking about chargers, the guy told me that as long as the voltage was the same, you could use a charger rated for a higher current on a battery with a lesser current charge/discharge capacity. This make sense if my understanding is correct, in that while over-voltage could cause problems like frying components not rated for that voltage or arcing across PCB traces or something, but the device only ever draws as much current as it needs if the voltage source is supplying the correct, safe voltage. Is that the right way to think about it? One last unrelated question: 4) What type of circuit do you need to "convert" (if that's even the right term) a dc voltage from a dc voltage source to a higher voltage, especially if you don't know what the resistance of the rest of the circuit is gonna be (or if you know that the resistance is going to be varying)? If you just give me the name of a type of circuit or system or concept, I'll do the requisite research on my own. Thanks everyone. Sorry for  
|
|
#
?
Apr 21, 2013 20:36
|
|
|
The Scientist posted:I've been drooling over this "Fuel" line of Milwaukee lithium-ion cordless power tools - the "M12" (12 volt), "M18" (18 volt), and "M28" (28v) series. Which brought up a couple of questions I've been pondering over the years. Current and voltage are two different things. Multiply them together and you get the total power. So, a 12V battery giving you 3A of current and an 18V battery giving you 2A are producing the same power: 36 watts. However, a 12V battery supplying 2A is only giving you 24 watts. So the first thing is that if you compare batteries with the same mAH capacity, the higher voltage one actually stores more watt-hours. As for current capacity, calling a battery "12V" is under ideal circumstances. The more you load a battery (the more amps you demand of it) the less the actual voltage is. One way to think of this is as an internal resistance, the battery actually behaves as if it were an ideal voltage source in series with a resistance that limits how much current you can actually get. Car batteries are engineered to produce very high currents for very short periods of time to turn starters, and they trade off other qualities (such as reliability and ability to hold a charge) to get it. If you stack up more cells to get a higher voltage, the internal resistance will be nearly the same at reasonable currents, so if you can get 2A out of a cell, and you stack them into 12V and 18V batteries, as mentioned above the 18V battery is giving you 50% more power (watts). There are such things as DC-to-DC converters but they work by changing DC to AC and back to DC at a different voltage again. Note that if you are changing to a higher DC voltage, you will be demanding higher currents from your source, because you have to supply the power (watts). If you produce 18V/2A using a DC-to-DC converter connected to a 12V battery, the converter will draw 3A (actually more because converters are not 100% efficient, more like 80-85%). Ain't no free lunch.
|
|
#
?
Apr 21, 2013 20:49
|
|
|
The Scientist posted:
A couple things that you should know: A battery isn't modeled as a direct source. Think of it as a source with a resistance attached. That's the internal resistance, which increases under non-ideal conditions (like cold or dying batteries). With Ohm's Law, V=IR, resistance can be less than 1. So that 12v battery that can supply 500A, R=V/I = R=12/500 = 0.024 ohms of internal resistance Base Emitter posted:
This is wrong. Some DC-DC converters, like SMPS, can be up to 98% efficient, without any kind of conversion to AC.
|
|
#
?
Apr 21, 2013 21:04
|
|
|
The Scientist posted:
ICs that do this are generally called boost converters. The switching type can be >95% efficient.
|
|
#
?
Apr 21, 2013 21:12
|
|
|
The Scientist posted:I've been drooling over this "Fuel" line of Milwaukee lithium-ion cordless power tools - the "M12" (12 volt), "M18" (18 volt), and "M28" (28v) series. Which brought up a couple of questions I've been pondering over the years. These are awesome tools. As to point 1) specifically. All the cores are wound about the same size (motors have the same physical dimensions). So more voltage means more torque. All the other electrical stuff comes into play, too, but that's the main difference in the product lines. The 12v line is going to be homeowner-type stuff. Normal screws and holes. 18V is contractor-level. 3" screws without pilots, 1 1/2" spade bits, 1" auger bits, with the expectation that a charge is going to last ~4 hours or so. The 28 (and 36) volt series are for heavy duty stuff; 2" and 3" augers and spades, 4" - 6" hole saws, 1" hammer drills, that kind of thing. The batteries don't have to last all that long (though they do), it's more for portability and convenience not having to drag a cord out to somewhere for one or two holes that you'd normally use a corded drill for. I have an 18v milwaukee of the older generation. It'll spin a 6" hole saw through 3/4" plywood for four or five holes, but that's about it. The 28V series will do about 20 holes. The 12v series usually bogs down on 6" hole saws; just not enough torque or power to do the job.
|
|
#
?
Apr 21, 2013 21:41
|
|
|
The Scientist posted:1) What's the advantage of a higher voltage cordless tool? I always figured the amount of current, not voltage, would be the determining factor when it comes to power (power in terms of, like, torque or speed). Or should I be thinking about it with Ohm's Law in mind? In other words, a cordless tool with a 28v battery can draw up to (theoretically, of course) -> 28 amps (of course A LOT less current because of all the resistance, but the point I'm trying to make is "Is it true that a 28v battery is naturally capable of delivering more current at any one moment than an 18v battery? And is that why a higher voltage power tool would be preferable?") Batteries have ALWAYS confused me; sometimes they don't seem to follow Ohm's Law. For instance, I've seen a lot of 12v lead-acid car batteries that say something like "500 cold cranking amps". But how is that possible? It says it right on the battery, which implies that there's no kind of voltage increasing circuitry, and even if there was a transformer, you can't transform DC, can you? I'm utterly baffled by this. As far as how a car battery can supply 500 amps? If the load is less than .024 ohms, a 12 volt could supply 500 amps. Car batteries are designed for very low internal resistances, such as high surface area plates. This is because that .024 ohm resistance has it include the battery's internal resistance. quote:2) It looks like the battery packs (there's a cutaway diagram in the brochure) are made up of AA sized Li-Ion cells. So one of these Milwaukee "Fuel" M12 12v battery packs would have less cells, in series, to provide the 12v the tool runs on; whereas a M28 28v has either more of those cells, or bigger, higher voltage ones. But does this mean a 28v tool battery necessarily has more AH than a 12v battery? Although, of course a 28v tool surely uses more charge, more quickly. So if you have one 1.5 volt battery with 1000mah, or 2 half size 1.5 volt batteries at 500mah. You'd have the same total watt hours. 1 * 1.5 * 1000 = 2 * 1.5 * 500 = 1500 mWh. quote:3) Would it safe/ok/a good idea to use, like a 12v lead-acid rider lawnmower battery wired (with the correct polarity of course) to the terminals of a 12v cordless power tool where the tool's normal battery's own terminals usually connect? Or would I need to build some kind of current or overloading protection circuit? When I was in Radio Shack one time asking about chargers, the guy told me that as long as the voltage was the same, you could use a charger rated for a higher current on a battery with a lesser current charge/discharge capacity. This make sense if my understanding is correct, in that while over-voltage could cause problems like frying components not rated for that voltage or arcing across PCB traces or something, but the device only ever draws as much current as it needs if the voltage source is supplying the correct, safe voltage. Is that the right way to think about it? Lead acid batteries are pretty much always safe to replace a different battery with. They are very safe and hard to hurt. The original charger may not work. Different batteries have different charging requirements, and if the charger may not recognize the new one.
|
|
#
?
Apr 21, 2013 22:25
|
|
|
I can think of a special case where the battery type is important though: LED flashlight drivers; even some fairly expensive LED lamps rely on the alkaline battery (or reduced voltage from the ni-mh) to limit current it seems. They run very hot when powered from the AA lithium batteries you can get these days (which are excellent, easily 4x battery life in the right conditions), though I've never seen one fail, there are some special products that rely on alkaline batteries high ESR to work properly.
|
|
#
?
Apr 21, 2013 22:59
|
|
|
longview posted:AA lithium batteries you can get these days (which are excellent, easily 4x battery life in the right conditions) What conditions are these? I was looking at Energizer's datasheet for their Lithium AAs and they were only rated about 10% higher in terms of lifespan at 50-ish mA loads; are they better at very low load currents maybe? I know at high load currents they're a stiffer source, but I don't think they have longer absolute life unless you're calling 1.3v/cell dead or something like that.
|
|
#
?
Apr 21, 2013 23:07
|
|
|
At 50mA and below there's not much to be gained at all, but in LED flashlights that pull an amp or two they'll give significantly longer run times in my experience. They are also far better at extreme temperatures. The internal impedance is very low yet they still have very high shelf life, making them great for flashlights, radios and similar high drain devices. I also use them in my Agilent U1272A (not known for super long battery life, I've been using the same AAA lithiums since late 2011) and HP 50g calculator which had the same batteries from summer 2011 until just recently when they started to die (it went through a pair of alkalines in about three months, for reference). An anecdote: a few weeks ago I was at 30% charge on my iPhone and had a 14 hour day of traveling ahead of me, I bought a 4xAA charger that normally is rated for about a half full charge with alkalines (user reports), I was able to do a full charge and then keep the screen on playing video and internet use for 10 hours before the batteries died, at which point I still had a full charge. I suspect that charger was also designed for the alkaline's high impedance because it did get alarmingly warm during the initial charging. I think as a general rule, they're best used for things where you're otherwise replacing alkalines more often than once every few months. E: flashlights are a special case though, the flatter discharge curve and high output voltage means they deliver 100% brightness for longer than alkalines, Ni-Mh cells are hopeless in my Led Lensers since their low output voltage means a dimmer light even when fully charged longview fucked around with this message at 23:38 on Apr 21, 2013 |
|
#
?
Apr 21, 2013 23:34
|
|
|
longview posted:... AA lithium batteries you can get these days (which are excellent, easily 4x battery life in the right conditions), though I've never seen one fail, there are some special products that rely on alkaline batteries high ESR to work properly. ... I can see how this test could be rigged (well, depending on how you look it at you might find it misleading). I remember reading in my "Intro to Electronics" course textbook that one Li-Ion cell is 1.2v as opposed to the standard 1.5v for normal alkaline AA's. so if you were doing a study comparing the relative "life" of some Li-Ion batteries - especially if the study is for your company or research sponsor - you wouldn't consider a 1.2v Li-ion vs. a 1.5 alkaline a fair match. Perhaps you'd be inclined to test a set of batteries in series whose voltages are equal. Maybe 12v. To achieve 12v's with the Alk.'s, you'd need 8. But 12 volts worth of Li-Ion cells @ 1.2 each equals 10 batteries. So right there, you've got a potential unfair advantage. Add to that the fact (which someone alluded to in one of the posts immediately preceding this one) that lithium ion batteries have that unique discharge curve where they maintain their 1.2 volts steadily until they're basically completely discharged, when the voltage across the battery plummets abruptly. A researcher could use this characteristic to be like "Hey look! The Li-ion is STILL showing 1.2 v meanwhile the alkaline has discharged all the way to 1.4v! Clearly the bLi-ion has monumentally more life from a single charge." When in reality they have both been discharged some. I hope this make sense and is correct, I'm super tired.
|
|
#
?
Apr 22, 2013 04:31
|
|
|
It's Ni-Mh and Ni-Cad that's 1.2V, li-ion is like 3.7V but the AA replacement lithium are not li-ion, they have 1.5V nominal and maintain it pretty well during discharge. The point is the amount of energy you can get from a lithium AA is higher when pulling large current and especially with cold batteries because they have very low ESR. Alkalines have around the same amount of energy stored but the high ESR means the batteries warm up like crazy at high currents, reducing the useful energy they can deliver. If you want to start counting the energy dissipated as heat internal to the cell, then I've got a class A amplifier that's 100% efficient, every single J put into it does something  Using data from: http://data.energizer.com/PDFs/E91.pdf and http://data.energizer.com/PDFs/l91.pdf Even at 500mA the effectice capacity (capacity delivered to the load) of an alkaline is less than 1500mAh, where a lithium provides around twice that. At 1A the effective capacity of an alkaline is only about 1000mAh, where the lithium is almost 3x better using the same discharge technique to 0.9V. At 0 degrees 1W load (typical for outdoors use where I'm from) the alkaline is about as useful as a coin cell, they switch to constant power for those graphs but the difference is somewhere around 10x better performance for a lithium, even at 50mW the lithium is twice as good. My 4x figure is from the room temperature 1W constant power graph, where an alkaline provides around 1Wh and a lithium has almost 4Wh. Additionally the higher nominal voltage and flatter discharge curve is advantageous for simpler LED flashlights.
|
|
#
?
Apr 22, 2013 17:49
|
|
|
Quick question. I want to rig a small camera to an rca jack. I think the diagram below is how i need to wire it but i want to check before i blow anything up.
|
|
#
?
Apr 25, 2013 14:00
|
|
|
If that - is really - and not ground this is defiantly wrong. Otherwise it looks OK. Also the red wire is usually + and black is usually ground.
|
|
#
?
Apr 25, 2013 15:47
|
|
|
I've used an MPY634 to make a nice little DSB modulator, but I'm having a problem generating a nice clean stable carrier for it, the frequency stability is critical for SSB/DSB so it needs to be crystal controlled. Basically the stronger and cleaner the carrier is the more modulation I can achieve so it should ideally be ~12Vp_p. My current solution is to take a 10 MHz oscillator (I'll order a TCXO or used OCXO later) and divide by 4 to get a 2.5 MHz square wave, then I run it first through a 2.5 MHz crystal, a 2N7000 to provide some voltage gain, then a second crystal at the same frequency and another 2N7000 amplifier, the end result is a ~6Vp-p 2.5 MHz sine but it's still not super clean. I have two of these crystals so I was wondering what's the best to apply them for filtering harmonics? I have on order some ~200 MHz GBW opamps so I was thinking maybe I could build an active filter using the crystals in the feedback loop? In simulations I found that a normal non-inverting opamp with the crystal between the inverting input and ground (+ series resistor to limit gain) worked at lower frequencies but then seemed to actually have the opposite effect above 2 MHz, maybe the series capacitance of the crystal becomes a limiting factor? I didn't really get anywhere with putting it into an inverting configuration (good since it would allow negative gain for other frequencies) since it seemed to actually have a notch-filter effect. Or maybe I should just use the crystals as purely passive filters (crystal+loading cap) and implement a normal active filter with gain to stick after it?
|
|
#
?
Apr 27, 2013 19:05
|
|
|
|
| # ? Jun 3, 2024 21:06 |
|
|
Trip report: toaster oven for solder reflow. Product links are all at the bottom. I could not find the Oster '6293' oven that so many folks have used and instead picked up a newer model, 'TSSTTVMATT', from Amazon. It is slightly less powerful (1300 watts versus 1500) and still has analog controls but otherwise appears to be identical to the 6293. Verdict is still out on accuracy of the temperature dial as I did the first couple boards by sight, my cheap multimeter and thermocouple are not trustworthy. The solder paste I used was Kester EP256 63/37 no-clean in an 8cc syringe with a 19 gauge tip, also from Amazon. This seemed like the better choice as I won't be getting in to stencils just yet, everything is just one-offs or prototypes, and it worked out very well. I kept the stuff in our garage fridge so it was a little difficult to get it flowing at the beginning but once it had warmed up a bit it started moving better. I just tapped once or twice on the plunger with my thumb and a 1mm long 'drop' was pushed out, that was then pressed onto a pad. The circuit boards are multi-channel LED drivers for my aquarium and were printed by Laen at OSH Park. The toaster oven:  PCB with solder paste applied:  PCB straight out of the oven:  A PCB I hand soldered a couple weeks ago for cleanliness comparison:  I am extremely happy with the results! The only things I really need to iron out are placing the proper amount of solder paste and measuring the accuracy of the temperature dial, although I think eyeballing it went drat well anyways. So, for anyone thinking of moving to SMT/SMD components, don't let the soldering put you off. It is really quick and easy! Toaster oven: http://www.amazon.com/Oster-TSSTTVM...ords=TSSTTVMATT Solder paste: http://www.amazon.com/Kester-EP256-...ds=Kester+EP256 Printed circuit boards: http://oshpark.com/
|
|
#
?
Apr 30, 2013 02:48
|
|
