Well, I may have been quiet the last few weeks. I was VERY busy.
PIXAD8 PCBs arrived April 18th. Quickly assembled one and proceeded on testing. Worked on hardware and software all week and finally on Saturday afternoon got rolling test patterns working on a sting of WS2801 lights.
Now working on LPD6803 and then proceeding on to TM190X. Still a lot of software work to do, but that's what I do best. Have to finish E1.31 code, add ARTNET II code, add configuration screens etc.
But the hardware is fully tested at the logic level.
Today I received a 50 amp 12VDC power supply. I used MR16 50 watt mini-spots to test full loading on one of the power busses. Nothing melted (the wires to the lights got real hot but halogens will do that) but the +V and Ground traces got a little warmer than I would like under FULL load (28 amps). The whole room was getting hot with 350 watts of halogens trying to set my work table on fire. A note on the power supply: at first I had trouble with onrush current. I can't connect more than 12 amps of lights and turn power on and off at the power strip. For testing I had to use the individual string fuses as a switch and add a string (2x 4 amps) at a time. This is a problem with my supply and the use of halogens for testing.
Bottom line:
1) The logic is solid.
2) The software is functional but needs a lot of refinement to be ready for prime time but I have no worries on that issue.
3) Three mounting holes between the connectors across the front of the board do not have enough clearance around them for the screw head. Easy PCB design fix. Adds .5in to board length.
4) +V and Ground traces will be beefed up. Add 50% to +V and 100% to Ground. It was half sized but double layered. The power/heat didn't dissipate well enough. Adds .75in to board depth.
5) While redesigning I will switch from the standard blade fuse to the mini-blade fuse. The docs and listing for fuses and fuse holders from Keystone were confusing and I finally found the holder style I wanted in the mini-blade size.
So a good week of work with a lot of progress. I will refine the PCB design tomorrow and submit it soon. I will order a small first run of 55 boards, have 5 boards hand assembled, do a quick test and then, hopefully, release the others for production. If the 5 boards have problems I will junk the 50 PCBs and make any final tweaks necessary.
Attached are some photos of one board with just fuse holders and connectors being used for load testing and some photos of another board fully assembled with an ECG-M32MX sitting on top hooked up for software testing and development.
I also have the ability to remotely download firmware updates just about tested and will probably release this combination as the first boards that won't require a PICKit for updates. It still might be a good idea to have one available somewhere to borrow in case of an emergency since this is a new technique for me and could have some little quirks at first.
Everybody asks price. Still have to figure all costs but probably looking at $120 per set. People have asked about an enclosure. I looked into having an enclosure with pigtails or connectors and all of them added a lot of cost or reduced the amperage per string for a reasonable cost solution. I am going to acquire a quantity of CG-1000 and CG-500 enclosures (if it fits into the CG-500). I think a weatherproof enclosure with your cables coming into the box (any size or type) is more flexible. Of course you can make your own enclosure. I expect the CG-1000 to come in at about $15 including a mounting plate that adapts our hole pattern to the CG-1000 layout.
-Ed
oops - photos were too big. I will add a reply in a few minutes with links to the photos on our site.
PIXAD8 PCBs arrived April 18th. Quickly assembled one and proceeded on testing. Worked on hardware and software all week and finally on Saturday afternoon got rolling test patterns working on a sting of WS2801 lights.
Now working on LPD6803 and then proceeding on to TM190X. Still a lot of software work to do, but that's what I do best. Have to finish E1.31 code, add ARTNET II code, add configuration screens etc.
But the hardware is fully tested at the logic level.
Today I received a 50 amp 12VDC power supply. I used MR16 50 watt mini-spots to test full loading on one of the power busses. Nothing melted (the wires to the lights got real hot but halogens will do that) but the +V and Ground traces got a little warmer than I would like under FULL load (28 amps). The whole room was getting hot with 350 watts of halogens trying to set my work table on fire. A note on the power supply: at first I had trouble with onrush current. I can't connect more than 12 amps of lights and turn power on and off at the power strip. For testing I had to use the individual string fuses as a switch and add a string (2x 4 amps) at a time. This is a problem with my supply and the use of halogens for testing.
Bottom line:
1) The logic is solid.
2) The software is functional but needs a lot of refinement to be ready for prime time but I have no worries on that issue.
3) Three mounting holes between the connectors across the front of the board do not have enough clearance around them for the screw head. Easy PCB design fix. Adds .5in to board length.
4) +V and Ground traces will be beefed up. Add 50% to +V and 100% to Ground. It was half sized but double layered. The power/heat didn't dissipate well enough. Adds .75in to board depth.
5) While redesigning I will switch from the standard blade fuse to the mini-blade fuse. The docs and listing for fuses and fuse holders from Keystone were confusing and I finally found the holder style I wanted in the mini-blade size.
So a good week of work with a lot of progress. I will refine the PCB design tomorrow and submit it soon. I will order a small first run of 55 boards, have 5 boards hand assembled, do a quick test and then, hopefully, release the others for production. If the 5 boards have problems I will junk the 50 PCBs and make any final tweaks necessary.
Attached are some photos of one board with just fuse holders and connectors being used for load testing and some photos of another board fully assembled with an ECG-M32MX sitting on top hooked up for software testing and development.
I also have the ability to remotely download firmware updates just about tested and will probably release this combination as the first boards that won't require a PICKit for updates. It still might be a good idea to have one available somewhere to borrow in case of an emergency since this is a new technique for me and could have some little quirks at first.
Everybody asks price. Still have to figure all costs but probably looking at $120 per set. People have asked about an enclosure. I looked into having an enclosure with pigtails or connectors and all of them added a lot of cost or reduced the amperage per string for a reasonable cost solution. I am going to acquire a quantity of CG-1000 and CG-500 enclosures (if it fits into the CG-500). I think a weatherproof enclosure with your cables coming into the box (any size or type) is more flexible. Of course you can make your own enclosure. I expect the CG-1000 to come in at about $15 including a mounting plate that adapts our hole pattern to the CG-1000 layout.
-Ed
oops - photos were too big. I will add a reply in a few minutes with links to the photos on our site.
