author

splud

Inbox View Profile
14Comments

Tell us about yourself!

  • splud commented on Patchr_io's instructable Explorer Uno PCB Template
    Explorer Uno PCB Template

    A few suggestions:Use descriptive info on the silkscreen. "header" for instance isn't nearly as useful in the completed project as actual pin signal designations. Same for labeling the power connection, and identifying pin 1 for the µC (the square pad at pin 1 is not going to be very visible once parts are in place - and you're suggesting users possibly remove the µC and put it in an Uno to program it, so there's an increased chance of error here) and the ISP header (a silkscreen outline of the 2x3 IDC connector with the keyway on it would work nicely). The Vreg could use a pin 1 identifier as well (or a thicker line on the side where the tab is) - there is no square pad identifying pin 1.Many clone Uno PCBs have soldered-on SMD ICs instead of DIP - users cannot de-socket t...

    see more »

    A few suggestions:Use descriptive info on the silkscreen. "header" for instance isn't nearly as useful in the completed project as actual pin signal designations. Same for labeling the power connection, and identifying pin 1 for the µC (the square pad at pin 1 is not going to be very visible once parts are in place - and you're suggesting users possibly remove the µC and put it in an Uno to program it, so there's an increased chance of error here) and the ISP header (a silkscreen outline of the 2x3 IDC connector with the keyway on it would work nicely). The Vreg could use a pin 1 identifier as well (or a thicker line on the side where the tab is) - there is no square pad identifying pin 1.Many clone Uno PCBs have soldered-on SMD ICs instead of DIP - users cannot de-socket the µC in that case. Also, repeatedly de-socketing the µC will invariably lead to a damaged pin, rendering it no longer a useful Uno. Instead, might I suggest that you consider flashing the Uno with the "Arduino-as-ISP" program and just make the appropriate connections from the ISP Uno to the ISP header on your project board. Cleaner, and less prone to fault. Better yet, just purchase an "USBASP" dongle, which is like US$2.50 from China, and leave your regular Uno as a regular Uno.A nice upgrade might be to use an efficient switchmode regulator instead of a linear, which will be beneficial when running off of batteries (esp if using say a 9V battery).I haven't checked the PCB schematics, but even a casual glance at the images you have here indicates that you do NOT have a current limiting resistor for the red power indicator LED - if you're running 5V here, the LED junction is going to fail pretty quick. Probably should add a series resistor (as you have for the yellow LED), and consider a higher value resistance than a regular indicator LED, so that the LED isn't drawing a lot of current. 510 or 680 ohms instead of 220 for instance.Note your components list does not include the voltage regulator or the PCB itself.

    View Instructable »
  • splud commented on medhatko's instructable Zener Tester
    Zener Tester

    Note that in the instructable schematic based on a LiIon cell, the typical 3-wire voltmeter display should be capable of being powered directly from the LiIon - you can delete the 12V linreg, and the two filter caps, running the OUT from the boost to the resistor, and the ground to the zener. Red lead for the voltmeter should connect at the IN+ to the boost (so after the on/off switch). The regulator and capacitors appear to be leftover components from the higher voltage transformer design, but are unnecessary when you have a convenient 3-4V battery supply. Those voltmeter displays have an onboard LDO.I am unclear on the purpose of the LM317 in this circuit - out of curiosity, I even watched the video, but outside of the names of some components, it appears to be entirely in an Arabi...

    see more »

    Note that in the instructable schematic based on a LiIon cell, the typical 3-wire voltmeter display should be capable of being powered directly from the LiIon - you can delete the 12V linreg, and the two filter caps, running the OUT from the boost to the resistor, and the ground to the zener. Red lead for the voltmeter should connect at the IN+ to the boost (so after the on/off switch). The regulator and capacitors appear to be leftover components from the higher voltage transformer design, but are unnecessary when you have a convenient 3-4V battery supply. Those voltmeter displays have an onboard LDO.I am unclear on the purpose of the LM317 in this circuit - out of curiosity, I even watched the video, but outside of the names of some components, it appears to be entirely in an Arabic language, so I couldn't parse an explanation. Thing is, the 4.7K resistor should do a fine job of limiting current. With a 30V supply for instance, Ohms law dictates I = V/R, so 30/4700 = 0.006382A (6.4mA).The 50V cap used in the schematic would be fine for the 30V transformer supply version, but 1000uF is entirely too much capacitance for this purpose.For charging a capacitor bank from a low voltage source through a boost circuit, I use a zener in series with an LED as both a regulator and as an indicator that the zener voltage has actually been reached. With a lower capacitance capacitor of sufficient voltage instead of the large bank, the same circuit can be used to characterize a zener by measuring the voltage across the zener itself (or being sure to account for the Vf of the LED, which leads to an output voltage (at the capacitor) about 3.0-3.4 V higher than the zener voltage, depending on LED type).Without an indicator like this, you might measure a voltage that is close to your max supply voltage, which could still be lower than your unknown Zener voltage - which doesn't mean that the Zener voltage has actually been reached. If the LED illuminates, it will do so only because the Zener is conducting - if it isn't illuminated, then the Zener is not conducting, and either the Zener is failed open, the Zener rating is higher than your voltage, OR, you're dealing with a non-zener diode with a high reverse breakdown voltage.Attached image shows a compact 16x21mm PCB I made for cap bank charging - the indicator LED gets press fit into a snug hole in the faceplate of the cap bank (so no need for mounting screws) and the board incorporates the boost circuit capable of producing voltages of about 60V for testing. The zener needn't be soldered in place - that could have wires with clips or whatever soldered for clipping to a component needing measurement, or a PCB with pads to seat a MELF SMD resistor on.

    View Instructable »
  • splud commented on giladlotan's instructable DIY Touch Sensor
    DIY Touch Sensor

    I can understand your perspective with respect to a full-on Arduino board perhaps, there are a variety of bare AVR microcontroller ics that are actually MUCH less expensive than the original IC used for this project (when it was still available). To top it off, many of those natively support capacitive touch. I use ATTiny13A uCs in a number of projects, and those can be sourced for less than US$0.50 apiece in single unit quantities, and can drive your logic. ATTiny85 uCs are documented to support 3 QTouch channels:http://www.microchip.com/wwwproducts/en/ATTINY85The ATTiny13A is listed as one of many supported devices on a QTouch informational page, although its own datasheet makes no mention for such support. I don't personally use it for capacitive touch, so I can't say one way or ...

    see more »

    I can understand your perspective with respect to a full-on Arduino board perhaps, there are a variety of bare AVR microcontroller ics that are actually MUCH less expensive than the original IC used for this project (when it was still available). To top it off, many of those natively support capacitive touch. I use ATTiny13A uCs in a number of projects, and those can be sourced for less than US$0.50 apiece in single unit quantities, and can drive your logic. ATTiny85 uCs are documented to support 3 QTouch channels:http://www.microchip.com/wwwproducts/en/ATTINY85The ATTiny13A is listed as one of many supported devices on a QTouch informational page, although its own datasheet makes no mention for such support. I don't personally use it for capacitive touch, so I can't say one way or another for QTouch library support, but it would certainly support a more direct approach to capacitive sensing.The package size for the 8 pin ATTiny devices solidly qualifies as smaller than your thumbnail. They're commonly programmable using the Arduino IDE.Note that capacitive touch can be used for more than simple on/off buttons - sliders and dials can be implemented using it - mostly via a carefully constructed PCB pattern.While I haven't tried to make this particular project (esp since the IC is NLA), one can readily find ICs with similar functionality, such as the AT42QT1012-MAHR (there are a host of devices in the ATxxQTxxxx line - AT for Atmel and QT for QTouch. which is capacitive sensing), which is about 1/5 of the price mentioned for the original IC employed here.

    View Instructable »
  • Tune Up Your Table Saw With a 2x4

    Bummer that left tilt, right fenced saws have the arbor opening on the fence side -- if not for that, then the fixed face of the arbor would be consistently at zero.You can get shims for blades which address this. Precision dado sets would include some, but there are other sources.

    The tooth is contacting the screw at a different point on the face of the screw. The tolerances for screw manufacture simply isn't as precise as it's being used for here, nor is it's insertion into the piece of wood -- if the screw isn't absolutely square to the block, and parallel to the table, the head of the screw will be on a different plane than the saw blade. A proper dial indicator is the tool to use here.In the spirit of this instructable seeking to achieve the tuneup using a 2x4 and minimal additional tooling, ripping a thinner scrap of 2x4, say to 3/4 or 1" thick, driving a screw through that so that the point penetrates out the back and the head seats fully, then screwing that block into the remaining stock that is to be clamped to the mitre, but with the screw point...

    see more »

    The tooth is contacting the screw at a different point on the face of the screw. The tolerances for screw manufacture simply isn't as precise as it's being used for here, nor is it's insertion into the piece of wood -- if the screw isn't absolutely square to the block, and parallel to the table, the head of the screw will be on a different plane than the saw blade. A proper dial indicator is the tool to use here.In the spirit of this instructable seeking to achieve the tuneup using a 2x4 and minimal additional tooling, ripping a thinner scrap of 2x4, say to 3/4 or 1" thick, driving a screw through that so that the point penetrates out the back and the head seats fully, then screwing that block into the remaining stock that is to be clamped to the mitre, but with the screw point facing outwards will give you a fine point to contact the blade with - a point rather than the not-necessarily-planar screw head. Put that modified block on the mitre (loosely clamped) and adjust it to just contact the leading edge of the saw tooth, tighten the clamp and do your tap adjustments if still necessary. Now, when you slide it to the rear of the blade and rotate the tooth into place, you have just one point to contact and have eliminated planar error.Note that checking and correcting for slop in the mitre gauge rail should be a step that is taken before using the mitre gauge to square anything.The belt should be removed from the arbor and the blade/arbor checked for slop. With the belt in place and tensioning the arbor, you might not readily notice the propensity for movement, but when the motor is running and the arbor is under varying loads as you feed material, that could cause it to wobble.

    View Instructable »
  • Custom PCB Prototyping Using a Laser Cutter

    I can get a really nice looking PCB by doing the vector outlines only, and it is WAY faster to prep too. However, if you don't have suitable clearance between pads for SMD components, you're going to have a greatly increased chance of bridging (ESPECIALLY when there are heatsink pads underneath some devices). This wouldn't be as much of a problem if you were using solder mask, but this method of prototyping doesn't involve solder mask.Printing a raster image on the laser can end up resulting in jagged edges on the verticals. That can be mitigated by following up the raster etch with an outline cut, which then ensures everything has a crisp transition.Avoid jacking up the power too much - it will delaminate the copper (won't etch the copper itself, just cause the adhesive bond underne...

    see more »

    I can get a really nice looking PCB by doing the vector outlines only, and it is WAY faster to prep too. However, if you don't have suitable clearance between pads for SMD components, you're going to have a greatly increased chance of bridging (ESPECIALLY when there are heatsink pads underneath some devices). This wouldn't be as much of a problem if you were using solder mask, but this method of prototyping doesn't involve solder mask.Printing a raster image on the laser can end up resulting in jagged edges on the verticals. That can be mitigated by following up the raster etch with an outline cut, which then ensures everything has a crisp transition.Avoid jacking up the power too much - it will delaminate the copper (won't etch the copper itself, just cause the adhesive bond underneath it to gas out and separate the copper). Might not seem like a big issue if the spots being heated are areas you want to remove anyway, but this delaminating impacts adjacent material - so that thin trace you DO want may lift away from the PCB (and in fact, may etch away because the etchant gets underneath it).Experiment with different paints. Primer versus top coat enamel, versus lacquer, etc. You can have different results between Ferric Chloride (the brown stuff) versus Cupric Chloride (the bright green stuff made from priming Muriatic Acid and Hydrogen Peroxide with Copper). I have some Nalgene bottles with good seals that I use for "shaking it up" with small prototype PCBs - complete etch in < 5 minutes.If you botch a laser etch (say because the PCB design wasn't registered proper to your copper and ran off the edge), you can clean the PCB and repaint it to use again later. If you just spray over the botched area, you'll have an uneven thickness of paint and an area which may be more prone to partial etch.Note carburetor cleaner also works great for removing paint from the PCB.For the best results, make sure your cutting deck is level and the top of the PCB material is at the focal point of the laser.

    View Instructable »