|
A Standard Molex .062" connector would work well IMO.
|
#
?
Jan 3, 2013 02:16
|
|
|
|
| # ? May 31, 2024 09:32 |
|
|
Beat me to it. 2x3 Molex would be an even smaller footprint, have higher current carrying capacity, and if you spend the five dollars for the pin extractor Molexes are not hard to deal with at all.
|
|
#
?
Jan 3, 2013 02:17
|
|
|
Question. Why is Microchip so loving worthless with the quality of software/tools they put out? MPLAB has been pretty lovely for years, and their reboot, MPLAB X really isn't any better. Also less supported. Even though it's also been out for years. Now I'm Googling issues I've been having with the PICkit 3, and it looks like there's a pretty good chance it's bricked my PIC. And from what I've been able to tell, Microchip has been completely silent on this issue. Anyway, if anyone knows what's going on when I get either of these errors, it'd be much appreciated. code:
|
|
#
?
Jan 3, 2013 03:15
|
|
|
Has anyone messed with winding their own transformers? I've had this idea for a while of making a sort of "binary" variac. Instead of a wiper, there would be multiple output windings(maybe multiple input as well) with turns in powers of two, and some relays or something could dynamically reconfigure to provide various outputs levels. The low turn windings could be done in extra heavy gauge, and higher turns could be lighter gauge to give optimal power output over all voltage ranges. Do things like this already exist, is there a name for them? Any tips on winding something like this myself? Any chances of getting just a nice plain core by itself, or should I try to tear apart some existing transformer?
|
|
#
?
Jan 3, 2013 09:07
|
|
|
ante posted:Question. It's been a while since I've used the PIC stuff. One of my first ports of call was to replace the PIC, to see if the chip had gone or if it was a bad connection, etc. The first PICKit 3 I had didn't work properly and I had many errors similar to above, the replacement was a bit better. The PICKIT 3 is very pedantic about the USB controller. So I never managed to get it working directly on my PC usb ports, but had to buy a USB hub to get it working.
|
|
#
?
Jan 3, 2013 15:15
|
|
|
peepsalot posted:Has anyone messed with winding their own transformers? I've had this idea for a while of making a sort of "binary" variac. Instead of a wiper, there would be multiple output windings(maybe multiple input as well) with turns in powers of two, and some relays or something could dynamically reconfigure to provide various outputs levels. The low turn windings could be done in extra heavy gauge, and higher turns could be lighter gauge to give optimal power output over all voltage ranges. Do things like this already exist, is there a name for them? Yeah, this is how substations adjust line voltage sag due to differing temperatures. They have fuckoff big transformers with several closely spaced taps at one end. If the voltage sags, say because warm weather heated up lines and their resistance increased, it'll change a tap to bump the output voltage up a little bit, at the cost of a bit more current on the input. Winding transformers is hoodoo, I have a couple hedgehog transformers that I've been building, but dudes that can hand-wind laminated core transformers deserve props.
|
|
#
?
Jan 3, 2013 15:57
|
|
|
Jonny 290 posted:Yeah, this is how substations adjust line voltage sag due to differing temperatures. They have fuckoff big transformers with several closely spaced taps at one end. If the voltage sags, say because warm weather heated up lines and their resistance increased, it'll change a tap to bump the output voltage up a little bit, at the cost of a bit more current on the input. http://ludens.cl/paradise/turbine/turbine.html I'm sorry it's all in one page, but this guy. He built a hydroelectric plant. Small, sure. But he built it all, by hand, in the sticks. Stuff pulled in by donkeys. He wound his power transformers by hand. quote:With the turbine running smoothly, it was time to make the transformers, to be able to transport the full power to the house. The 600 meter long line was designed to have low loss when operating at 2000 volts. At 230V, the losses are extremely high at anything over a few hundred watts. So, here I go... Yeah. 230V generators, hand-wound xformer to step up to 2kV and a matched one to step back down at the other end. Mad. Props.
|
|
#
?
Jan 4, 2013 02:44
|
|
|
peepsalot posted:Has anyone messed with winding their own transformers? I've had this idea for a while of making a sort of "binary" variac. Instead of a wiper, there would be multiple output windings(maybe multiple input as well) with turns in powers of two, and some relays or something could dynamically reconfigure to provide various outputs levels. The low turn windings could be done in extra heavy gauge, and higher turns could be lighter gauge to give optimal power output over all voltage ranges. Do things like this already exist, is there a name for them? These type of transformers exist, but are expensive. They are used in benchtop linear power supplies. Different taps can be selected to provide different input voltages for the regulator depending on the desired output voltage, reducing the amount of power dissipated as heat in the pass devices. On some of the nicer Agilent power supplies you can hear the relays clicking as you turn the output voltage all the way up.
|
|
#
?
Jan 4, 2013 03:17
|
|
|
FSMC posted:It's been a while since I've used the PIC stuff. One of my first ports of call was to replace the PIC, to see if the chip had gone or if it was a bad connection, etc. DeviceID 00000000 when also detecting a target is a ICSP problem usually. Are PGEC/PGED also being used in application? Can you successfully [i]program[\i] vs. debug it? Have a Real ICE handy to double-check? I have a buddy with a PICKIt3 who always gets 00000000 but he never pays attention to all the parasitics he adds by using the pins for other things.
|
|
#
?
Jan 4, 2013 03:45
|
|
|
babyeatingpsychopath posted:http://ludens.cl/paradise/turbine/turbine.html That was a great read, thanks. the wizards beard posted:These type of transformers exist, but are expensive. They are used in benchtop linear power supplies. Different taps can be selected to provide different input voltages for the regulator depending on the desired output voltage, reducing the amount of power dissipated as heat in the pass devices. On some of the nicer Agilent power supplies you can hear the relays clicking as you turn the output voltage all the way up.
|
|
#
?
Jan 4, 2013 04:14
|
|
|
babyeatingpsychopath posted:http://ludens.cl/paradise/turbine/turbine.html Holy loving poo poo.
|
|
#
?
Jan 4, 2013 05:21
|
|
|
peepsalot posted:Is there a particular name for that style of transformer? Have you looked at the WE-FLEX Flexible Transformer. This doesn't fit your description(power of two turn ratios), but it's the closest thing I can think of. They are designed so people can configure the transformer for their own application without getting a custom one made, but with the right circuits you can switch the input:output turn ratio through quite a few different configurations.
|
|
#
?
Jan 4, 2013 11:17
|
|
|
Slanderer posted:Holy loving poo poo. Agreed, this guy has basically made every DIY project I've ever done seem like LEGO in comparison.
|
|
#
?
Jan 4, 2013 12:41
|
|
|
The same guy did a massive redesign job on some POS inverter he got on eBay: http://ludens.cl/Electron/chinverter/chinverter.html
|
|
#
?
Jan 4, 2013 12:54
|
|
|
I found that page a year ago and am still marveling at the work he did. We're planning our retirement/grey years to be on a mountaintop or somewhere with elevation change, and I was extremely impressed with his work. And yeah, the DIY 2,000 volt power line is ballsy. But he knows what he's doing.
|
|
#
?
Jan 4, 2013 20:54
|
|
|
ante posted:Question. I had the same problem trying to program certain PIC 16f88 chips within MPLAB X. I downloaded the standalone PicKit 3 programmer app thing here: http://www.microchip.com/Microchip....ocName=en538340 and had much better success with it. You can set it to watch a directory and automatically flash new firmware on your chips when you recompile it. I've personally found it easier to use than the ridiculous MPLAB X hidden sub-sub menu buried inside another dialogue that you have to use to get at PicKit settings, even though it is crash prone.
|
|
#
?
Jan 5, 2013 05:21
|
|
|
nobody- posted:I had the same problem trying to program certain PIC 16f88 chips within MPLAB X. The problem turned out to be that I didn't have MCLR connected, despite having it configured as an internal pullup. None of my configuration changes seem to be working properly, actually, and I haven't had time to try and figure out why. Developing in PIC is a process of beating my head against a wall at every step of the way.
|
|
#
?
Jan 6, 2013 05:42
|
|
|
Is there anything I should keep in mind when designing a system that has to accurately transmit signals up to around 100kHz? I'm trying to implement a system that's meant to measure high frequency noise signals in a cryogenic environment. I'm pretty confident about most of the details, but I'm trying to see ahead of time if there's going to be any problems related to the high-frequency part of it. Everything past the breakout box part has to fit in a tube approximately the diameter of a soda can, so space is a premium. 1) Breakout Box - Box that I plug my instruments into that leads down into the belly of the cryostat. They attach through BNC connectors, and here's where I'll have low-pass pi filters installed to help stop outside electronic noise from getting to my samples. 2) Transmission lines - I'm going to be using Thermocoax cabling as my transmission lines. Thermocoax cables are designed to be used as heating elements, but they also act as high frequency filters due to their small size which causes losses from the skin effect. The pi filters get rid of noise up to around 1GHz, but due to parasitic inductances you get spikes in the filtering above that. The thermocoax cables don't start attenuating until around 1GHz, but they're pretty effective after that. There's going to be 6-8 of these lines but they're pretty small so space shouldn't be an issue. 3) Connection to sample holder - I have to solder the wires to some kind of connector that attaches to my sample holder. The connection has to be removable, as the sample holder can be switched, so the connections have to be able to plug in and out. This will be attached to a PCB (that I'll have to design) that I plug my sample into. I'm slightly unsure about how I should attach these wires to the sample holder - our current set-up has 20 wires soldered into a plug, but it's just plain copper wiring and doesn't carry high frequency signals. I've heard of other labs using SMA connectors, but I'm wondering what is necessary in order to keep the integrity of the signal intact. How important is the connector when maintaining the integrity of the signal? What type of connector should I be looking for that does not affect high frequency signals? 4) Sample holder - PCB attached to a machined piece of copper which has a small holder in which we plug in our samples. This is small, on the scale of a credit card or two. The signal lines go to the place where we plug our sample into, which then goes through connectors we made with a wirebonder to our actual devices under study. Anything I should keep in mind when laying out the PCB for this?
|
|
#
?
Jan 7, 2013 21:16
|
|
|
ante posted:The problem turned out to be that I didn't have MCLR connected, despite having it configured as an internal pullup. None of my configuration changes seem to be working properly, actually, and I haven't had time to try and figure out why. MCLR is "special", and pretty much every PIC datasheet in the first chapter or two gives you the sample connection for MCLR, doesn't it? Mandatory pull-up to VCC, and you can add a cap/series resistor if needed afterward as well. I don't see how the internal pull-up on it would have ever worked, as there's probably a pseudo-race condition in which it notices MCLR floating before it can even get to the point of activating the internal pull-up. Also,  at the homemade hydro-plant, that is insane.
|
|
#
?
Jan 7, 2013 22:42
|
|
|
Insurrectum posted:Is there anything I should keep in mind when designing a system that has to accurately transmit signals up to around 100kHz? I'm trying to implement a system that's meant to measure high frequency noise signals in a cryogenic environment. I'm pretty confident about most of the details, but I'm trying to see ahead of time if there's going to be any problems related to the high-frequency part of it. Everything past the breakout box part has to fit in a tube approximately the diameter of a soda can, so space is a premium. I'm not a cryogenics expert but I know a thing or two about coaxial components. If you are only going to 100 kHz, you have basically nothing to worry about electrically. Focus on the mechanical challenges. These guys (Thermocoax) provide little to no information, but what I gather tells me they are lossy as hell. Even so, 100 kHz should be no problem. They are apparently nearly impossible to solder to. Stick to crimp connectors if possible, preferably both the center contact and the shield ("dual crimp"). "Wrench crimp" or similar connectors will be an improvement, but still somewhat unfortunate. Glue-encapsulated heat shrink will be your friend if you need an environmental seal. SMA is just fine for this application. Avoid non-threaded connectors at all cost unless you absolutely need them. Do not use right-angle connectors. Solder joints and swept contacts don't react well to thermal contraction. I expect PTFE is a superior option at absolute zero to Noryl/Delrin/PE, so stay away from the absolute bottom of the barrel connectors (China). Don't worry about electrical performance of your connectors, this frequency range is a cakewalk for anything smaller than a football.
|
|
#
?
Jan 8, 2013 01:05
|
|
|
Does anyone remember the name of that relatively inexpensive sattelite modem for sending short messages was? I remember that the company had units for a few hundred dollars that were either bare, or in IP housings... It's new, and I think I saw it in this thread sometime late last year, but maybe it was somewhere else... Alternatively, does anyone have any ideas for cheap methods of relaying tiny amounts of data from a remote, mobile sensor? I had a prepaid cell with an accessible AT modem that could be convinced to send/receive txts over a serial port (in the headphone jack!), but the plan I was able to get was brutal--I think my minutes vanished after a month or 3 months or something. On an unrelated note, it turns out there is a Raspberry Pi thread in SHSC, complementing the Embedded Programming thread
|
|
#
?
Jan 8, 2013 17:32
|
|
|
There's Spot which you can hack, but I think you're right, I remember seeing something akin to a dedicated module you could fling into any project and get instant satellite connectivity. Might have even been in this very thread.
|
|
#
?
Jan 8, 2013 18:05
|
|
|
Slanderer posted:Does anyone remember the name of that relatively inexpensive sattelite modem for sending short messages was? I remember that the company had units for a few hundred dollars that were either bare, or in IP housings... It's new, and I think I saw it in this thread sometime late last year, but maybe it was somewhere else... Oooh, I got this. I designed the communications controller used on a transatlantic balloon attempt. It caused lots and lots of heartburn. There are two cheap satellite modules out there: The Digi M10 (on the Orbcomm Network), and the Iridium 9602 (on the Iridium network). Of the two, the Iridium modules is easier to use and has better coverage. It is, however, more expensive. The Spot tracker can be hacked to get some location information out, but nothing more than that. The Delorme InReach contains an Iridium 9602 modem, and the Bluetooth protocol for it has been partially decoded. Unfortunately, it's still hacky and inflexible, and the modem inside is locked to Delorme as a service provider. Unfortunately, buying the modem is only half the battle. You also have to find someone willing to provide you with service. In the case of Orbcomm, you can buy the Digi M10 starter kit, which comes with 90 days of free service (they didn't cut us off after the 90 days were up, though). There was only one company willing to work with us when we bought our iridium modem: http://www.nalresearch.com/ Since then, a company in the UK has started selling devices: http://rockblock.rock7mobile.com/ The 9602 uses either an AT protocol, or a pretty simple binary protocol. The Digi M10 can only use a complicated binary protocol. Furthermore, Orbcomm's constellation is not large enough to provide constant coverage, antennas are hard to find and even harder to construct, and their satellites don't communicate with each other, so the satellite must be in communication with your device AND a ground station in order to transmit. I strongly recommend going with the 9602.
|
|
#
?
Jan 8, 2013 18:26
|
|
|
Slanderer posted:Does anyone remember the name of that relatively inexpensive sattelite modem for sending short messages was? I remember that the company had units for a few hundred dollars that were either bare, or in IP housings... It's new, and I think I saw it in this thread sometime late last year, but maybe it was somewhere else... http://rockblock.rock7mobile.com/
|
|
#
?
Jan 8, 2013 19:09
|
|
|
sixide posted:I'm not a cryogenics expert but I know a thing or two about coaxial components. If you are only going to 100 kHz, you have basically nothing to worry about electrically. Focus on the mechanical challenges. Ok, I have a much better idea on what I'm doing now.  Here's a general overview of what I'm doing. There's a problem, though: Thermocoax: Here's a pdf with another group who did something similar to what I want to do. The main difference is they solder the sample end of their wire directly into the chip holder, while I have mine go into SMA connectors (because my sample holder has to be detachable). My problem is on the other end, where they have to somehow go from vacuum to atmosphere, and preferably through a connector (so I could take off the breakout box if something in it needed to be adjusted). Our current set-up uses Fischer connectors mounted through a vacuum sealed port (picture here). Is there a way to have something similar to this with coaxial cables? Or should I do something like this, where the coaxial cables go straight through? And maybe I have SMA connectors on the other end outside of the epoxy, and screw them into my breakout box? I'd basically be building my own connector, at that point. Still thinking this out. Any ideas?
|
|
#
?
Jan 9, 2013 21:51
|
|
|
movax posted:MCLR is "special", and pretty much every PIC datasheet in the first chapter or two gives you the sample connection for MCLR, doesn't it? Mandatory pull-up to VCC, and you can add a cap/series resistor if needed afterward as well. I never had any problems with just setting MCLR to internal pull-up and routing that pin to an ICSP header only; my project isn't very intensive though. He might be using a different PIC than me.
|
|
#
?
Jan 9, 2013 23:36
|
|
|
Insurrectum posted:Ok, I have a much better idea on what I'm doing now. Air will rush in down the center of the coax given sufficient pressure differential (more than a few psig) unless the connectors have a strong seal at both ends. Basically, don't run cables straight through your bulkhead. Would something like this do the trick? http://www1.futureelectronics.com/doc/JOHNSON%20COMPONENTS/142-1000-003.pdf Basically, the seal pin has a glass bead around it and a metal ring on the outside. This gets soldered into a bulkhead and a pair of flange mount connectors with female contacts in the back can connect to either side as shown for the test fixture. There are companies making non-coaxial glass seal pins as well (DE9, DB25, various). Alternatively, AMP has some hermetically sealed feedthrough adapters. These cost $150 or so and depend on a robust seal being made around them at the bulkhead with a good O-ring or similar.
|
|
#
?
Jan 10, 2013 00:18
|
|
|
So, the Product Assurance people in my building have several of this dc load, one of which I've had perpetually signed out for most of the past year (mostly for various battery testing). They're pretty good, even though the Windows software is *terrible*---most of the pulsed load functionality is broken, tons of bugs in the battery discharge characterization scripting, weird default behaviors that enable loads when you least expect it, etc....There's also a python library and a labview driver, and they probably work better for actual automation stuff. Regardless, it's a pretty solid load. But remember how I said the Product Assurance people own these? Due to weird budgeting priorities, my engineering group doesn't have any (we have an ancient, enormous HP load, but it's not really very useful). I'm trying to get some new test equipment on the budget for the year, including a couple of electronic loads and a DC supply with a PC interface (either serial or isolated USB or something). But the loads are still pretty expensive, so I'm wondering if there might be any cheaper alternatives. Anyone have any experience with these? On a marginally related note, I watched this teardown of a BK8500 yesterday (I still can't stand listening to this guy). Anyway, in the video we see that there is a bank of MOSFETs, a group of power resistors, and two zigzags of wire sticking out of the board. He claims the wire would be a current shunt--but why would there be two of them? There are spaces for 2 more, which could correspond with high-power versions of this load...As for the other power resistors, I assume they correspond with the bank of MOSFETs (and the op-amps they surround) and form multiple, parallel op amp current sinks, in order to give higher resolution over full range. Does that make sense, or am I just being stupid? I really wish more power electronics people did DIY stuff on the internet, because this stuff is fascinating. There are so many design mistakes I could overlook (if only because they are to protect from non-obvious edge cases and faults) that I don't want to make anything like this on my own.
|
|
#
?
Jan 10, 2013 01:11
|
|
|
I was thinking of building an audio amplifier with digital volume control driven by, say, the Arduino I have lying around. I like the idea of controlling random things around me with my computer and amplifier circuits are fun. I was thinking I would control the volume by building a voltage divider and adjusting the value of one of the resistors, as described here: http://embedded-lab.com/blog/?p=2967 However I was wondering, do I really need the digital potentiometer chip? I was reading elsewhere that for reasonably low frequency circuits you can make a fairly effective variable resistor by using PWM to control a transistor that repeatedly shunts power away from (or into, depending on how you set it up) a normal resistor, the idea being that if you use a 2k resistor at a 50% duty cycle the circuit would only "see" a 1k resistor. I'm just wondering if this would be unacceptably noisy for an amplifier and I should just get some of those digital potentiometer chips, or if I could just smooth it out with a capacitor or otherwise handle it. If I don't need to get those chips, I think I have everything I need to breadboard the circuit right now.
|
|
#
?
Jan 10, 2013 03:16
|
|
|
So I've wanted to get into electronics for a while, and seeing this made me wonder if making a copy of it would be a decent beginner project? I don't have any electronics experience so I don't expect to throw it together in an afternoon. But other than that, just go buy a soldering iron / some solder, the parts and take a crack at it?
|
|
#
?
Jan 10, 2013 03:28
|
|
|
Slanderer posted:On a marginally related note, I watched this teardown of a BK8500 yesterday (I still can't stand listening to this guy). Anyway, in the video we see that there is a bank of MOSFETs, a group of power resistors, and two zigzags of wire sticking out of the board. He claims the wire would be a current shunt--but why would there be two of them? There are spaces for 2 more, which could correspond with high-power versions of this load...As for the other power resistors, I assume they correspond with the bank of MOSFETs (and the op-amps they surround) and form multiple, parallel op amp current sinks, in order to give higher resolution over full range. Does that make sense, or am I just being stupid?
|
|
#
?
Jan 10, 2013 03:52
|
|
|
Parallel Paraplegic posted:However I was wondering, do I really need the digital potentiometer chip? I was reading elsewhere that for reasonably low frequency circuits you can make a fairly effective variable resistor by using PWM to control a transistor that repeatedly shunts power away from (or into, depending on how you set it up) a normal resistor, the idea being that if you use a 2k resistor at a 50% duty cycle the circuit would only "see" a 1k resistor. I'm just wondering if this would be unacceptably noisy for an amplifier and I should just get some of those digital potentiometer chips, or if I could just smooth it out with a capacitor or otherwise handle it. If I don't need to get those chips, I think I have everything I need to breadboard the circuit right now. Also you should know that digital pots often work horribly as variable attenuators for audio, since when you change the resistor settings, you will often get enormous "glitches" on the pots, which can sound horrible in an audio circuit.
|
|
#
?
Jan 10, 2013 04:00
|
|
|
Insurrectum posted:Still thinking this out. Any ideas? I don't know a lot about this subject myself, but the lab down the hall does high frequency work in cryogenic/vacuum environments and I could try to corner one of their folks and ask what they'd use. I'm interested to know the answer myself. In terms of a 'problem statement' you want to get a 100kHz sine wave passed in through a vacuum seal and down to your sample, correct? Any particular power requirements? Are you wedded to the idea of using SMA or is just any 'ol plug that will do the job fine at this point? What type of cryogen are you using - N2, He, superfluid He? Any other limitations like being non-magnetic or a maximum allowable thermal load?
|
|
#
?
Jan 10, 2013 04:06
|
|
|
ANIME AKBAR posted:It could work, as long as the PWM frequency is very high, and you filter it carefully. Basically the simplest implementation of such a scheme is to make a "variable" gain amplifier that can either have some finite gain, or zero gain, and just modulate between those two states. Following on this, the default pwm frequency on your arduino is approximately 490 Hz, which is absolutely audible. You'll need to bump it up quite a bit. Human hearing goes up to around 20khz, so that's an absolute minimum there. Depending on what pin and what timer is on that pin the pwm frequency will top out at 61500hz or 31250hz, either of which should work for you. Keep in mind there are some functions depend on these settings, timer0 is used for mills() and micros(). I don't think that either timer1 or 2 is used by default (by things other than analogwrite() anyway) A couple of things to get you started. icantfindaname posted:So I've wanted to get into electronics for a while, and seeing this made me wonder if making a copy of it would be a decent beginner project? I don't have any electronics experience so I don't expect to throw it together in an afternoon. But other than that, just go buy a soldering iron / some solder, the parts and take a crack at it? They give you a parts list, and the code. Only other thing you need is the long exposure camera. It's all though hole, no particularly odious parts. Go for it.
|
|
#
?
Jan 10, 2013 06:28
|
|
|
My recommendation is an actual pot with a dc motor, that's how commercial products used to do it. Reading back the volume could be a problem, but if you can find one intended for quadrophonic audio then you could put a DC voltage on the spare track and the value will be proportional to the position of the pot (log scale makes it a little harder). About two years ago I was able to make an Arduino read an IR remote, that controlled two of these, one was a volume, the other acting as an input selector, when you turned the pot it would jerk the motor at the input boundaries. When you changed input using the remote or by turning it manually it would also remember the volume, and it could cancel a remote-initiated input change if you moved the selector manually (stopping the motor if the value went the wrong way). I never built it up properly with actual inputs and I had trouble converting the voltage reading into evenly spaced angles on the pot then (I now know more math, and could probably do it properly now), but if you can get motor controlled pots that's the best solution IMO.
|
|
#
?
Jan 10, 2013 10:26
|
|
|
Aurium posted:Following on this, the default pwm frequency on your arduino is approximately 490 Hz, which is absolutely audible. You'll need to bump it up quite a bit. Wow, I had no idea it defaulted to a frequency that low, I always assumed it was in the mid-tens-of-kHz range for some reason. Thanks! longview posted:My recommendation is an actual pot with a dc motor, that's how commercial products used to do it. I also had no idea that pro models would just use motors to turn pots, I always thought they'd have a more... elegant solution? I guess it makes sense though as it lets you manually adjust the knobs too. I have some research and experimenting to do, thanks guys.
|
|
#
?
Jan 10, 2013 13:57
|
|
|
A more high end solution usually involves replacing the volume control with a really lovely rotary encoder, and doing the volume control in a DSP I think. Personally I think a nicely geared smooth motorized pot is a very elegant solution, doesn't involve digitizing the audio signal, no need to use possibly lovely electronic trimpots. It also looks cool when the knob rotates by itself.  I have an old Technics receiver that does the input selection using linear sliding switches, the knob is connected with a wire and has a little gearing system to push/pull the slider in the switch.
|
|
#
?
Jan 10, 2013 14:15
|
|
|
There are also plenty of high quality voltage controlled VGAs and attenuators which are suitable for audio. Not sure why anyone would use motorized pots anymore.
|
|
#
?
Jan 10, 2013 14:52
|
|
|
ANIME AKBAR posted:There are also plenty of high quality voltage controlled VGAs and attenuators which are suitable for audio. Not sure why anyone would use motorized pots anymore. Because audiophiles are weird as gently caress. Digital audio is such a solved problem that it's not even an issue. Hardcore audio people seem to invent arguments about why it sucks based on invented problems, some stemming from bad experiences with early digital equipment 30 years ago and cheap chinese crap. A variable gain audio amplifier is also a non-issue, since you can find single IC solutions for it which are essentially guaranteed to have amplifiers and filters with basically no distortion within the audible range. Digital potentiometers aren't really a drop in replacement for physical ones, since designs may need to be altered (different current handling capacities, different supply voltage filtering requirements, etc). As for a DSP, you can use one, sure, but it's way overkill (except for cheap low-end ones). It make sense when you're doing a full solution with equalization, filtering, and whatever the gently caress else you want to throw in.
|
|
#
?
Jan 10, 2013 15:30
|
|
|
|
| # ? May 31, 2024 09:32 |
|
|
This is pretty annoying: the IR serial interface for my U1272A uses an outdated Prolific controller chip, meaning it's no longer supported in Windows 8. It works with the old driver from before they started their anti-clone chip driver check, but that bluescreens the computer after about an hour or two of logging. There's no updated version of the hardware or driver except the Bluetooth thing which won't work on iOS. Looks like I'll be opening this thing and hacking in a new USB serial chip then... the PL2303TA is apparently pin compatible. longview fucked around with this message at 18:52 on Jan 10, 2013 |
|
#
?
Jan 10, 2013 18:50
|
|