Jump to content

Dmills

Regular Members
  • Posts

    753
  • Joined

  • Last visited

Previous Fields

  • Member Status
    Working in the industry
  • Current Employment or place of study
    At present I am doing a theatre referb job in Liverpool. Capital building projects are a nightmare!
  • Full Name
    Dan Mills

Contact Methods

  • Website URL
    http://
  • ICQ
    0

Dmills's Achievements

Experienced

Experienced (11/14)

  • First Post
  • Collaborator
  • Posting Machine Rare
  • Conversation Starter
  • Week One Done

Recent Badges

  1. Also a bridge rectifier right on the input so you can power the thing with pretty much any wall wart that comes to hand, it means that polarity becomes irelevant, and that even an AC supply would work.
  2. I suggest a polyswitch in series with each ring power line, and a fairly butch cap across the power input, that way a shorted cable does not take the whole thing down. Regards, Dan.
  3. More subtle, but tune that parallel LC on the power input to 455KHz and do the tally signallying by coupling a pulse coded 455KHz onto the power line. Good ceramic filters are readily available for this frequency (Used in AM radios) and the detector is straightforward, then stuff the resulting bitstream into a small pic as a uart and decode which tally should be on. The tally controller could actually sit in line with a completely unmodified beltpack, line comes in, control carrier is sampled, line goes out to the pack with power via an LC lowpass to not pass the tally control carrier. This actually sounds fairly straightforward to implement, 1200 baud should suffice so nothing very sophisticated is required. Regards, Dan.
  4. What is the function of the L1,C10 tuned circuit? I could see it if C10 ran from the junction with the diodes to ground forming an RFI filter, but a parallel tuned circuit there makes little sense that I can see. Just curious. Regards, Dan.
  5. I will do Mouser if nobody else want to, they are worth the trouble. Regards, Dan.
  6. Top tip when buying more then a very few PCBs, it usually costs exactly the same for a panel full as it does for four or five, and for a small board a panel can contain a lot of boards. Board houses will however never offer a full panel unless you explicitly ask them (So always ask). For something as low density as blueclone we would normally go to Printed Wiring, for something more critical Wurth have a very good prototype shop, but that is more for fine pitch and BGA sort of work. If getting boards that will not be stuffed for a while, ENIG surface remains much more solderable a year after you get the boards then OSP (Spit!), plus it looks really pretty. 73 Dan.
  7. There is a picture of some teardropped connections a few posts down this page http://dangerousprototypes.com/forum/viewtopic.php?t=3622&p=36846 Note how the tracks smoothly merge into the pad. This helps to reduce the risk of drill breakout and also gives a little more copper in the pad area It is only worth it if you cad system makes it easy (altium does, as I suspect do most of the serious cad systems), if not then forget it, too much like hard work. Regards, Dan.
  8. Excellent work. Just a nitpick based on the board photos, it is hard to tell from a photo, but it might be worth bumping the default anualar ring size up a couple of notches. Particularly on the connectors, but also generally it looks very much tighter then I would normally use for a through hole board, and increasing it will make registration of the two layers somewhat less critical as well as making removal of a component less likely to damage the board. Also, adding stitching vias will help tie the two polygon pours forming the ground plane together, never a bad thing. It looks to my eyes like C11 is across L1, I would have expected to to be to ground? It looks on the photo like there may be something funky going on under R20 component side, dead track? Via tenting? Your call, but normally our house style tents both sides of vias. Teardropping connections? if you cad package does this it is useful on a through hole board, the extra copper makes accidentally pulling a pad during rework slightly less common. Sorry, doing PCBs professionally these days and you come to notice these things. Regards, Dan.
  9. How about doing the mic switching with a picogate D type or similar, so that a short interruption to the power from the master station can be used as master 'mic off' command from the prompt desk? I believe some of the Clearcom kit does something like this and it is usful when some muppet leaves a mic on and walks away. Also, some compression and/or noise gating? Possibly just a jfet across the mic line with fairly simple minded feedback? Could we have an option to put a fixed resistor in line with the bottom of the volume control so it is not possible to take the headphone volume all the way to zero by accident (Hate to say it but I have missed a cue due to that one). Final thought, if the signal lamp control voltage were to be defined as on say between 1/2 and 3/4 and off between 1/4 and 1/2 of the supply voltage then it would be possible to implement a red LED to indicate a shorted or pin 2/3 reversed line, that would save a lot of fault finding pain on occasion (A confidence LED on the front would be good). The 1/4 supply would be provided by the power supply via the terminating network so any cable short or open could be detected by the outstation and some faults could be picked up by the master station. I think this would probably be backward compatable. I have access to Altium 10 and would not mind getting involved in the design if that toolchain matches what you guys are using. Regards, Dan.
  10. Silly thought, but what are the 610s using as snubbers? Replacing a triac designed for snubberless operation with one designed for use with a snubber will often cause all sorts of amusing problems. Another thought, were the original parts by any chance sensitive gate types? Replacing these with conventional units will often cause poor behaviour (either failing to trigger reliably or triggering but generating excess heat). What is the PSC measured at the dimmer supply terminals, the presence of large induction motors or close proximity to a sub station can often cause a supply to apparently be very stiff. Regards, Dan.
  11. Indeed, I have 4 Vortex 6 which routinely get run off a single 16A socket, but there is something a little subtle to notice. The over current protection when fed from the grid is normally quite slow for overloads of less then about 5 to 10 times the protective device rating, so as long as the voltage drop does not become excessive on a transient you can pull a LOT of power for a few tens or hundreds of milliseconds from a supply that is protected at a fairly low level (More so if class C breakers are in use). Combine this with the fact that peak/average ratio for music is usually reckoned to be about 8 or so, and you can see how my 24KW of amps might be quite happy on a 16A (RMS) supply, in spite of the odd few milliseconds when they draw several times that. Now the critical point with generators is that (particularly with small ones), the source impedance may be much higher then that of the grid, and large transient current demands may both saturate the magnetics, confuse the voltage regulator and cause frequency changes, none of which are good. While working with RMS currents makes reasonable sense when grid connected, small generators (Like the amplifiers themselves) really need to be rated for nearer to the peak demand. A good trick with small generators is to hang a KW or so of static load (site lighting) on there as well as the PA, it makes the rapid power demand variations smaller in percentage terms which helps the voltage and speed regulation. Regards, Dan.
×
×
  • Create New...

Important Information

We have placed cookies on your device to help make this website better. You can adjust your cookie settings, otherwise we'll assume you're okay to continue.