|
|
Post by valhalla on Sept 21, 2023 17:30:20 GMT 1
It's been a while since wightdiag did a run on these sorts of things, but I have been itching to get a few projects under my belt for a while. Worried that I was spending too much on tools that may not see a lot of use, I decided to buy a lot of expensive components instead, all of which may not see much use.....and even less if I didn't get around to building the projects.... 
One of the big shortfalls in my armory has been a lack of sensing for pulley/shaft speeds and timing, but also I was interested in getting a cheap array of 3-axis accelerometers put-together, ones where it would matter less if they were driven around the Isle on a wet day, strapped under some knocking Landrover of one denomination or another.
The starting-point needed to be a power-supply, specifically something that could give me a reliable 5-volts for any circuits I needed to power on the sensors (most off-the-shelf solutions will cope with 5v, even those intended for 3.3v Arduino work);
I was going to build-up an LM337 regulator into a circuit, but I found a cheap (and Chinese?) variable voltage supply for model-railways still sitting in my carousel from some long-departed wish/project. I think it was around £2.30 on a certain auction-site, and I only bought it around 4years ago, tops.
Building a power-supply from scratch would have been better, as there really isn't too much to solder around an LM337, but this board needed using. It looks a bit under-engineered (1/4w resistor, for a start) but it ought to be OK for 0.5Amp supply at 5volts. I stressed it out with a load-resistor, and found that the regulator was getting a little toasty, so made the decision to glue it to the side of the enclosure, and use that to dissipate a bit of temperature. The sensors that are going to be driven by this regulator are unlikely to draw more than 100mA between them, so loads of margin for safety/reliability.
The board I used looks like it was designed for an LM337, but they must have bottled-out at the last minute in Chinatown, as the legs into the board have been crossed-over, and this regulator is a similar one from another catalogue - I cannot remember now what they have used.....
|
|
|
|
Post by valhalla on Sept 21, 2023 17:35:01 GMT 1
The variable voltage bit of the board is not really needed - what I wanted was a reliable 5v, and no more...
I set the potentiometer on the board to give me 5v output (almost exactly), measured the resistance across the pot, then de-soldered it off the board and substituted a 680Ohm 1/2W resistor in its place.
Re-measuring the output, I swept the input voltage from 7.5v all the way through to 14.5v, and got a nice steady 5.01v output, so I was happy this was not going to fry the next bit of my project
|
|
|
|
Post by valhalla on Sept 21, 2023 17:55:03 GMT 1
Optoelectronics are very cheap these days, not least the ITR20001/T package that you can buy from all good resellers, and some not-so-good ones as well. I got all my bits through the PiHut (effectively Adafruit bits for the most part) who are quick, cheap, and efficient. Every time.
The circuit above is not quite right, so don't copy this one - it's there to show the concept.....
What I found, after making the first prototype, was that it didn't work. 5v in from the power-supply? Check. Anything out from the pickup? Negative.
These packages do not come any simpler than this - I scratched my head, checked the diagram I had from a YouTube video, re-checked it all - perfect.....
Then I checked the datasheet for this exact device. Something was fishy....the emission device side of things (an IR diode) looked OK, but the receiver (a photo-transistor) didn't look right. Then it occurred to me, someone had swapped the emitter and collector pins around (think power-supply above) and conflated an NPN device with a PNP device. I must post a comment to that effect on the relevant site, which is a good one all the same ("siuautomotive" with Sean).
As the package above was too well heat-shrunk together, I made a second prototype, after having checked my theory on a protoboard (which I should have done first of all, doh!!) and all was well. The second prototype is not as pretty as the first, but it works on its 3metre flylead just fine. I haven't tested it yet for speed-response, but I guess it ought to be OK for what I need - propshaft balance data and engine speeds. 'Scuse fingers on this one:
It only needs three wires to run; a ground (which is shared as a common sensor and emitter ground), a 5v power supply, and a sensor feedback wire. Signal is normally 5volts, but pulls-down towards ground when anything reflective passes the sensor head. Depending on the distance and reflectivity of the surface passing-by, the signal level will rise/fall accordingly. So you could use this as a distance-measurement device as well, but that's not needed here. It is fine with a finger passing near the head, and will pick-up a Tippex line whizzing past on a drill-chuck. I haven't scoped it yet, but the Fluke 87 says it works.
|
|
|
|
Post by valhalla on Sept 21, 2023 19:53:54 GMT 1
The 3-axis accelerometer is even easier to build; just buy the following circuitboard from the PiHut (or anywhere that sells it), solder a few leads onto the board, and you're done !
EDIT : Forgot to mention, ignore the 3.3v out and Test pins on this board, we are only interested in the Vin (5v) , GND (0v) , and the three axes outputs
The board needs a power-supply (3.3v to 5v), a ground, and three sensor output wires for X , Y , Z axes. The direction each axis measures is marked on the board, although you end-up covering that up with heatshrink if you want to make this robust. In my case, I knew that I needed to make a means-of-attachment that would allow this to be used in a number of situations. Whatever method is likely to be used, the circuitboard needs to be insulated to prevent a dead-short across the tracks. I used a 3-stage heatshrink method to not only insulate the board, but support the wiring soldered through it. Once this was cooled-down, it was a matter of finding a small magnet.
As I had blown the tweeters on my Disco2 A-post speakers (the perma-damp combined with Blondie had conspired to distort everything), the logical thing was to find the old tweeters in the drawer, gut the cones and wiring, smooth the faces off with a flap-wheel, then cover the working-part of the magnet (the bit that attaches to bodywork) with a black nitrile-rubber glove and tie-wrap, before gluing the sensor assembly to the back of it.
The magnet is not as strong as I would like, but the mass of the assembly is quite low, so it ought to cope with 3g of acceleration, especially if I support the cabling properly on each job;
The outputs in each axis can vary around a static voltage, depending on which way up the sensor sits; there is a background DC voltage that depends on gravitational pull on one or more of the axes. As the signals are likely to be AC-coupled to look at vibrations, not pitch/yaw/roll in DC-mode, then this is fine. Typically, you see around 1.8v when "level" on any one axis. The sensor-board is only good for 3g in any direction, but I'm guessing at this stage that it will suffice for guidance towards offending components. I have the proper NVH kit from Picotech, and if it really matters, I can use that for homing-in on precise data, but 90% of the time you just need an indication of problems, and a "cheap array of inexpensive sensors" will save more money/time than a single expensive unit.
Clearly, if I have more than one accelerometer sensor, or multiple position/speed sensors, I'm going to run out of channels (4-ch Picoscope 4425), so the next stage is to sort-out the power-supply box to allow a "telephone exchange" of leads/sockets to select what needs measuring each time, and on-the-fly. As of now, I'm awaiting a big pack of 4mm banana sockets with capability for shrouded-plugs (high-voltage Pico leads) to allow this selection to occur.
|
|
remmington
Apprentice
Owns Spark Eroder
Posts: 4,966
Member is Online
|
Post by remmington on Sept 21, 2023 20:30:27 GMT 1
|
|
|
|
Post by rhyds on Sept 21, 2023 20:51:12 GMT 1
As I had blown the tweeters on my Disco2 A-post speakers (the perma-damp combined with Blondie had conspired to distort everything),
Tide was high was it?? Great work here chief. I understand some of the concepts and ideas, but have never had the time and patience for deep electronics work. |
|
|
|
Post by valhalla on Sept 21, 2023 21:38:02 GMT 1
As I had blown the tweeters on my Disco2 A-post speakers (the perma-damp combined with Blondie had conspired to distort everything),
Great work here chief. I understand some of the concepts and ideas, but have never had the time and patience for deep electronics work. No - the amplifier went Atomic, which led to their Demise (doo-be-doo)
No skill required here with the electronics. I'll try and post the circuit diagrams. A fine-tip soldering-iron and some multi-core data-cable are all that are required to make this stuff. The wonders of modern electrical assemblies pretty much sort the rest out.
|
|
huffo
Tea Maker
Posts: 237
|
Post by huffo on Sept 22, 2023 7:47:43 GMT 1
Sounds like the Heart of those speakers was made of Glass.
Excellent work Valhalla.
I’ve abandoned far too many electronics projects that simply failed to work, or that died soon after they were born. I find electronics frustrating. It’s not my soldering, but probably poor quality components/boards and me trying to use them beyond their intended scope. I can’t diagnose electronics at all - the reverse engineering that Big Clive does it voodoo magic as far as I’m concerned.
|
|
|
|
Post by valhalla on Sept 22, 2023 11:20:20 GMT 1
I’ve abandoned far too many electronics projects that simply failed to work, or that died soon after they were born. I find electronics frustrating. It’s not my soldering, but probably poor quality components/boards and me trying to use them beyond their intended scope. I can’t diagnose electronics at all - the reverse engineering that Big Clive does it voodoo magic as far as I’m concerned. The thing with modern electronics is that so many basic devices can be bought pre-assembled or integrated onto a single chip, far cheaper than you can source the individual items. As long as you observe the most fundamental precautions (limit the voltage, get the polarity right, watch the temperatures) then it's fairly easy to make things that last well.
Big Clive makes it look easy. You have to remember that even before he was Big Clive, he was well-regarded in the electronics field amongst professionals and amateurs alike. Being time-served in any profession helps a lot in this respect, so "muscle-memory" takes over whenever you are doing a diagnosis on components and assemblies. The main thing is to start "simple" and work your way up, don't try to fix your Arcam A12FMJ Amplifier and get all frustrated when you get nowhere.......(I know a thing or two about this, right now.......  ) |
|
|
|
Post by Karl on Sept 26, 2023 19:47:48 GMT 1
do you fancy making me up a pressure pulse sensor please then 😄😄😄😄
|
|
|
|
Post by valhalla on Sept 26, 2023 21:27:40 GMT 1
do you fancy making me up a pressure pulse sensor please then 😄😄😄😄 I'll look into it !! I haven't done a pressure sensor yet, but I have some ideas about the sorts of things that might be useful. One of those ideas would be to ensure that a delta-pressure can always be piped across a transducer - so you can easily measure a pressure-drop dynamically over a DPF or inlet tract, for example. More often than not, you have to use two pressure sensors, two oscilloscope channels in parallel, then use a "Maths Function" to do the delta-P. Far easier and more efficient if you could just throw another pipe onto the back of the transducer. |
|
|
|
Post by valhalla on Sept 28, 2023 19:43:32 GMT 1
Well, it isn't as pretty as I would have liked - the story of my life - but the sensors and the 5v power-supply box have come together this afternoon. In my defense, I haven't been 100% well this last few days, and this job had to be done indoors and away from my electrical workstation;
It has become obvious that I'm going to have to "stack" enclosures if I want to get multiple sensors of each type into this assembly; I have used too small a base enclosure to get all the banana sockets in. That's what comes of not planning ahead!
What I have done is to bring the 5v supply out to external sockets, so stackable test-leads can be used to not only reference 5v and ground for the sensor signals, but also they can be used to bring that 5v up to adjacent modules. By keeping it all flexible, it might just be that the whole project is more useful. As it stands right now, the main module here has the power-supply, an optical pick-up, and a 3-axis accelerometer.
The finishing-off of strain-reliefs on all the cables has yet to happen, especially on the 12v input cables (with crocodile clips for the battery terminals), but that is relatively trivial now.
The next post or two will explain how I was able to solder indoors.....
|
|
|
|
Post by valhalla on Sept 28, 2023 22:09:36 GMT 1
I had almost forgotten one of the tools I built a few weeks ago. I'm quite often required to make solder repairs to equipment at remote locations, and until recently this entailed using a gas soldering-iron. Although this particular tool had served me well, the locations of remote work were often in barns and animal-sheds, meaning that the risk of conflagration was pretty high with a naked flame present (albeit behind a catalyst).
This particular project turned-up by chance on YouTube one evening, a North American post, and as soon as I saw it, I knew it was the answer to most of the problem above. Broadly-speaking, this low-voltage DC device can drive a small soldering-iron, control its temperature up to high 300's celcius without too much self-destruction, and heats-up (also cools-down) very quickly.
Sadly, the unit is designed for 24VDC, meaning that powering it by car battery is not going to work. Happily, it will work down to 18VDC, and that means that with a bit of purchasing of adapter;
the unit can be modified together with it (an inline fuse, for a start!!!!) to produce a device that mounts on top of a Milwaukee M18 Fuel battery. What I did was to fully strip the unit apart, do the mechanical modifications and some proper DC wiring/fuse within that enclosure, then seal it all back up together;
The unit has the footprint basically similar to an M18 Fuel pack, meaning it is incredibly compact. The iron that comes with it is a chisel-tip 4-pin generic unit - not a bad bit of kit, but not my preferred tool. Sadly, the spare iron I bought has come with a 4-pin plug of the wrong diameter, so I'm going to have to modify that with a replacement (probably CPC Farnell).
|
|
|
|
Post by valhalla on Sept 29, 2023 21:40:03 GMT 1
The iron warms-up pretty quickly - under 15seconds from room temperature to circa 330celcius. It struggles a little bit towards the end, but the temperature-control characteristics are adequate, if not benchmark, to keep the iron around the right sort of temperature.
The voltage indication is quite useful as a guide, but I've noticed that you can use this iron on and off for over 30mins and not even make a difference on the battery-guage (bar-graph) on the M18 Fuel cell itself, so I reckon that this could be good for 180mins at-least.
The only problem I need to deal-with is the "sleep" function on it. I thought that it would be a good idea to leave this programmed "on", given the remote-work aspect of it, but I have found that the iron is too insensitive to my grip, and it quite often cuts the current through the heater just at the wrong moment, hence the less-than-impressive soldering on the +ve post inside the sensor/supply enclosure above - it all went cold at a critical point.
|
|
|
|
Post by Karl on Sept 30, 2023 20:12:29 GMT 1
blo*dy brilliant little home made stuff
Love it
You do know you can buy cordless soldering irons
😄😄😄
|
|
)