Tell us about yourself!

Complete Your Profile
  • Home Made Bezel / Window for LCD, LED, TFT Displays.

    The types of flathead countersunk hex head bolts in the attached picture are great for such projects, especially the black bolts. I've used them a lot for such panels and also for mounting black items like IEC mains input sockets. They provide a fantastic finish, you just have to be careful and precise with your countersinking. Available on eBay from various sources. I use the M3 type most often but other sizes are available. I'd avoid pan-heads of a different colour on the outside as they don't look nearly as professional as the countersunk flatheads. If you don't want to countersink then use the button head ones, they look much nicer than the pan heads.I've used the milled edge rebate technique in the past to produce a flush finish and it works great ... IF you can get someone to do t...

    see more »

    The types of flathead countersunk hex head bolts in the attached picture are great for such projects, especially the black bolts. I've used them a lot for such panels and also for mounting black items like IEC mains input sockets. They provide a fantastic finish, you just have to be careful and precise with your countersinking. Available on eBay from various sources. I use the M3 type most often but other sizes are available. I'd avoid pan-heads of a different colour on the outside as they don't look nearly as professional as the countersunk flatheads. If you don't want to countersink then use the button head ones, they look much nicer than the pan heads.I've used the milled edge rebate technique in the past to produce a flush finish and it works great ... IF you can get someone to do the milling for you. I was lucky, I did. You don't NEED a CNC machine but a plain old milling machine IS required for a really good accurate finish. If you have a decent pillar drill, you can use a simple cross feed milling table to do the milling with your drill. eBay has a great one for under £50 that I'm planning on getting. http://i.ebayimg.com/images/g/8EQAAOSw32lYvTNS/s-l...However, you CAN achieve the effect by taking advantage of the ability of true perspex to be acetone welded invisibly and you build the total from the top piece that fits exactly into the cutout hole (very accurate cutting, filing and trimming is needed. Then before black painting the reverse, acetone weld a complete piece over the top piece like a layer cake. the second layer is larger than the first so provides the rebate step. It's tricky but can be very successful. You can then black mask the rear as before.Another trick is to use stand-offs behind the front panel, bolted to the front panel via countersunk holes at the front with the relevant bolts, then mount the operstional display etc to the stand-offs. The front bezel can then be applied OVER the countersunk bolts using double-sided adhesive or using a silicone bead as it should never need to come off again. I am using this process currently for a frequency meter display.

    View Instructable »
  • ATtiny85 Two-Channel Lead Acid Battery Charger

    This is a second attempt at posting a comprehensive and relevant comment as after I'd written a detailed entry I was than asked to login (!!) and the system "lost" all my input!!! Arrrgggh! Anyway, here goes for a second attempt (I have the text backed up first this time!) ...There are some obvious major improvements that can be made to this overall clever design but the most obvious is probably the one with the biggest cost ratio benefit and also a big space saving one. I have drawn an alternative design for the final output stage with the Darlington (for which I can't see a specification in your notes, but I may just have missed it). To anyone that knows classic current source design your three LARGE, hefty and somewhat expensive 25W resistors can be reduced massively in siz...

    see more »

    This is a second attempt at posting a comprehensive and relevant comment as after I'd written a detailed entry I was than asked to login (!!) and the system "lost" all my input!!! Arrrgggh! Anyway, here goes for a second attempt (I have the text backed up first this time!) ...There are some obvious major improvements that can be made to this overall clever design but the most obvious is probably the one with the biggest cost ratio benefit and also a big space saving one. I have drawn an alternative design for the final output stage with the Darlington (for which I can't see a specification in your notes, but I may just have missed it). To anyone that knows classic current source design your three LARGE, hefty and somewhat expensive 25W resistors can be reduced massively in size and cost by utilising the power and gain capability of the Darlington to do the shrinking for you.In the alternative design shown, the classic current source config is shown where a series resistor in the emitter circuit has a fixed voltage applied across it by virtue of the base current drive having 3 x 0.7v diode drops and in the emitter(s) current path we have 2 x 0.7v diode drops plus the voltage drop across R4 which MUST therefore be 0.7v (by virtue of Kirchoff's law). So to define the output current all we have to do is make R4 the correct value defined by R4 = 0.7v/Iout (Iout = charging current - max). For 5A charging current this means R4 should be 0.7/5 = 0.14 ohms. Not an easy value to obtain cheaply in the required wattage of 5 x 5 x 0.14 = 3.5 watts. It's also not a standard value in E24 or E48 series and even though the 1% E96 series contains the value 3.5 watt rating is not a common one - 3 watt and 5 watt can usually be found easily but not always.Something else to note about the original design, 15 volts is only 0.6v ABOVE the 14.4 needed for max charging and you will likely find that 14.4 cannot be reached due to the fact that unlike a normal PNP single transistor, a Darlington can never saturate due to its inherent design features. The 15 volt supply, even in the original, should ideally be increased to at least 15.5 to 15.8 to be certain and another volt on top for the alternative design. The battery voltage under charge will ONLY ever increase above 14.4 if it is being OVER charged as the current being supplied will effectively determine that voltage.Hmmm! ... This may not seem such an attractive design after all, well not until you apply an old engineer's trick - using the parallel resistors rule i.e. use 4 x 0.56 ohm 1 watt, easily (and cheaply) available from ALL of the E12 (10%) E24 (5%) AND E48 (2%) common series, albeit at maybe a higher tolerance in the lower series. Hardly anyone uses the old E12 series any more but if, like I do, you keep lots of old and recovered components, then you may well find that you have useful values of those. Obtaining 1 watt versions of 0.56 ohm resistors is FAR more easily, and MUCH less expensively achieved than 25 watt ceramic coated power resistors anyhow. In the diagram these four are represented as R11 through R14 and being only 1 Watt can be mounted on the circuit board much more securely than a bird's nest arrangement - just remember to mount them on longer leads between 3mm and 5mm ABOVE the board to avoid scorching as they probably will get quite warm if not hot.VR1 and R15 can be used (second old engineers' trick) to "trim" the value of the other four DOWN to a value that will INCREASE the current output. Should you find that your output current WITHOUT R15 and VR1 is already too HIGH however, then change just ONE of the 0.56 ohm resistors for the next value UP in the series i.e. 0.68 ohm. Alternatively if you have the luxury of being able to measure accurately the individual low resistance values and have multiple units from which to chose you may find you can trim the current value through picking and testing and even using non-identical values for the others. VR1 and R15 should be an order of magnitude LARGER than R11 through R14 combined so VR1 should be a max of around 1-2 ohms! Arrrgh! Usual engineer's dilemma number 2 - solve one issue to have another - just try obtaining a low cost wirewound pot of 1-2 ohms - if one exists at all! This being the case you will need instead to resort to a bit of "Set On Test" with two series resistors of the right kind of values - in the range of 0.47 to 1.2 ohms. BUT, because these resistors are roughly an order of magnitude higher (both connected in series remember) than the other four in parallel they will only need to dissipate a total of around 1/10th the power i.e. 350mW and that divided between the two should easily be handled by standard 0.25 watt E48 series resistors.True, all of that means a higher component count but ALL the extras are low (or much lower) power devices and can ALL be board mounted thereby dramatically increasing mechanical stability, reducing internal heat (ideally the Darlington should be mounted on an external heatsink) and greatly improving long-term reliability. Before I forget, the diodes D2-D4 only need to be large enough to handle the Darlington output current divided by the Darlington gain plus say 5% margin. Rather than worry about exact calculations just use any old standard 0.5 amp to 1A rectifier diodes, min PIR rating of 2 x VCC = 30 volts means you have a MASSIVE range from which to choose. You MAY get away with the ubiquitous 1N4148 (only 200mA max though) or the just as ubiquitous, and more appropriate 1N400x 1A series diodes (1N4001 - the entry level - is 50v PIR so way more than adequate). The latter series of diodes can usually be found by the dozen or more in old, non SMD electronic gear so ALWAYS salvage these when you can.I hope that even if this alternative output design isn't what you want to use for this, that the concepts prove useful for any other designs you might conceive.The circuit was drawn using the GREAT, free, on-line schematic tool from Digikey: https://goo.gl/BmBRx7

    View Instructable »