Hackaday Links: September 21, 2014


Obviously the best way to show you’ve pwned a piece of hardware is to run Doom on it. This time around it’s not just a hardware hack but a security hack. [Michael Jordan] demonstrates a network vulnerability on a Canon printer by making it run Doom. [Thanks to all who sent this tip to us]

Fans of the photo-resist method of PCB etching will appreciate this Bromograph built from an old all-in-one-printer. When gutted, the body of the printer makes a nice enclosure for four UV Lamps. The treated copper clad goes face-down on the scanner glass with the printed transparency between the two. [Michele’s] early testing shows really great results.

[Tom] sent in a link to this video biography of the rLab hackerspace. It shows off the space and its members but also tells the story of a tight-knit community. We enjoyed hearing that almost everything in the space is salvaged and repaired; a great way to acquire equipment on a tight budget while also building the skills of the members.

The bulbs in projectors can be quite expensive to replace. This hack adds an RC filter to the bulb and claims to greatly extend its life. Does this really work and why isn’t it built into the projector?

[Steve Maher] built a GRBL board that is the same form-factor as a DB25 connector housing. It’s basically an Arduino derivative that includes a USB connection, a separate jack for STOP and CYCLE START switches. If you’re not familiar, GRBL is an open source project that lets your drive parallel-port-based hardware with a machine that doesn’t have a parallel port.

Finally, have you heard of the ZofzPCB program before? It’s a way to visualize your gerber files in order to perform a final sanity check before sending your deign off for fabrication. [Thanks Boldport]

20 thoughts on “Hackaday Links: September 21, 2014

  1. Re bromograph: nice up cycling job. Got me thinking about actually keeping the scan track assembly and scanning one lamp across the board.
    Finding a lamp small enough might be hard. Perhaps a row or two of UV LEDs?

      1. I built my current UV box (apparently, “bromograph”, who knew) in a multi function, well, sort of, I took the scanner head and controls off an old brother multifunction and built a larger box underneath for the light source and diffuser. (Aside, mine reuses the original character LCD and control panel from the printer, driven by an arduino using the BroLCD library).

        It uses 5mm UV leds, lots of them, about 100, and exposes dry film in about 4 minutes 30 seconds. However the LEDs are about 25cm distant to the artwork, and there is a diffuser between them (necessary because even 25cm wasn’t enough to eliminate hotspots from the narrow beam of the LEDs, it was either use a diffuser, or probably quadruple the number of leds).

        However in a “scan” type of setup, especially if the LEDs in the “bar” were very closely spaced, you could have them much closer to the artwork, this would greatly reduce the exposure time for the area above the leds. It is possible that it might be feasible.

        I have a stack of old multi functions, maybe I’ll try it.

    1. If the bulb really does experience lots of square-wave driving, a big capacitor in parallel would help somewhat to increase its lifespan. That capacitor, though, is far too small. The “resistor wire” is unnecessary, and shows the lack of understanding on the part of the designer.

      It’s basically the same idea as putting a capacitor across your voltage lines on an MCU: It absorbs voltage spikes, preventing internal damage.

      1. I disagree with the similarity to a bypass cap across an MCU’s powerlines:

        The bypass cap in the MCU example is there to temporarily absorb voltage spikes, to prevent a rapid-rate-of-change in the voltage supplied to the MCU. Specifically, ic = C*dV/dT for a linear current source / drain.

        This cap in parallel doesn’t do that. It’s effectively a resonant LC tank circuit.

        The lamp’s initial inrush current is extremely high, adding a capacitor in parallel does not decrease the lamp’s inrush current, infact it will increase the inrush current presented to the lamp’s driver, presumably a MOSFET or transistor.

        Here is a short read that has some valid points:

        The parts proposed by allegro are constant current lamp drivers with a peak inrush limit. Some of the parts limit to 1 Amp. This means the lamp will warm up more slowly but the lamp’s bond wires will not experience the usual 10Amp current rating because the inrush would be limited to 1 Amp peak.

        Rarely do lights burn out when they’re already on. It’s typically when you turn them on (due to inrush current destroying the lamp bond wires) or off (due to inductive voltage spike.)

        If the lamp is considered an inductor: vL = L* dI / dT. By opening (turning off) the circuit you are effectively making dI / dT an extremely large number. The inductor’s stored energy will be presented as a voltage that will attempt to go to infinity. This is why arcing happens. Eventually a path is found to dissipate the energy stored in the inductor.

        1. A search on the bulb NSH200R or the lamp assembly PL9777 would show that at $200+ this is unlikely to be a tungsten filament. That said, the schematic as pictured in the related article shows a fixed filament in the bulb, which is certainly not correct.

          It’s not clear what this hack really does; it takes active circuit management to operate an HID bulb. This looks like it is something that circuit tolerates, not something that makes the bulb last longer. It’s a 4000hour bulb operated in ECO mode. How long does it usually last?



        2. While the bypass cap may absorb a spike, its principle function is to supply more current when the IC next to it needs more current during state changes. The capacitors at the power supply are “too far away”. The current surge without the nearby capacitor will momentarily pull down the power supply rail and possibly force other IC’s on the rail to change state.

  2. Re: RC filter for bulb. If the bulb is using AC, this circuit is hemorrhaging power, and I suspect the cap will will burn-out long before the bulb. If AC, a series inductor can be used to the same effect and will be more robust, although it will still hemorrhage power. If DC, the inductor is definitely the better idea.

    1. Older CNC Machines that use Mach3, etc. use a parallel port to send step data directly to the motor drivers. You can’t do this with USB-parallel adapter because it is too slow.

      GRBL takes G-code (through USB or from SD) and processes it into step commands. This “adapter” has an arduino-compatible chip loaded with GRBL, takes the G-code sends those step commands through the DB-25 connector, driving your older hardware as though it was through a parallel port.

      1. Yes. I was referring to the post here “If you’re not familiar, GRBL is an open source project that lets your drive parallel-port-based hardware with a machine that doesn’t have a parallel port”. That description of GRBL could actually be the description of the project.

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.