Let's try looking at this way:
Can a controller be made that uses USB as its input?
The answer would be Yes
The top reasons why the DIYC community hasn't jumped on it:
1 - USB restricted in distance. Most people want the flexibility that RS-485 gives you for distance. Now those who are still using Grinch/Olsen boards probably wouldn't mind the USB limits but the majority would not be happy.
2 - Cost. To add the USB module to the controller would increase any design cost. As an example, the USB modules that RPM and RJ used in their USB->DMX dongles each cost in excess of $20. Nobody wants to build multiple boards with a single part that costs so much. But you can get a USB->RS232/RS485 converter for much less, and it will communicate with all the boards.
3 - USB powered controller not likely. Even if you can get the USB port to deliver 500mA then the most channels you could probably have would be 24.
4 - No improvement over existing capabilities. Once you have used up all of your USB ports at 24 channels per port then you would need a method for the USB controller board to communicate with other boards. This would most likely result in RS485, which brings us back to what already exists so all you really did was to create a more expensive first board that has to stay close to the computer.
Can it be done? - Yes
Does it save money? - No
Does it improve capabilities? - No, with the caveat being that the controller operation would be the same as current controllers.
Just some unsolicited input.... 😉
Yeah, that's kind of what I thought. If it was better and cheaper, it would already have been done. The main reason for asking is that when a total neophyte like myself is reading about all the dangles and dongles and 232 to 485 converters and 14 miles of cat5 cable and stuff it just sounds overwhelming.
Of course now that I've read a 1000 posts and looked at pictures, it is starting to actually make some sense! WOW!
I still say one of you guys needs to move out to the Portland area so I can beg, borrow, or steal all your ideas. For a fair price to be paid in the beer of you choice of course... :rolleyes:
Thanks for all your insight guys...
ripbme wrote:Yeah, that's kind of what I thought. If it was better and cheaper, it would already have been done. The main reason for asking is that when a total neophyte like myself is reading about all the dangles and dongles and 232 to 485 converters and 14 miles of cat5 cable and stuff it just sounds overwhelming.
"Overwhelming" is one reason why we do this! This hobby will take over your life if you're not careful. But as it turns out, that's not so bad either!!!
I still say one of you guys needs to move out to the Portland area so I can beg, borrow, or steal all your ideas. For a fair price to be paid in the beer of you choice of course... :rolleyes:
Beer? Free beer you say? Hmmmmmmm
RavingLunatic wrote:
3 - USB powered controller not likely. Even if you can get the USB port to deliver 500mA then the most channels you could probably have would be 24.
How did you calculate that? I am running 2 Renard controllers off USB, 1 with 112 channels (14 PICs), and another with 224 channels (4 PICs). I am not using the PWM firmware, so the power usage is low, though.
don
djulien wrote:How did you calculate that? I am running 2 Renard controllers off USB, 1 with 112 channels (14 PICs), and another with 224 channels (4 PICs). I am not using the PWM firmware, so the power usage is low, though.don
Are you actually powering the boards via USB (the PICS don't get power from the wall?), or just using it for signaling?
djulien wrote:How did you calculate that? I am running 2 Renard controllers off USB, 1 with 112 channels (14 PICs), and another with 224 channels (4 PICs). I am not using the PWM firmware, so the power usage is low, though.don
My rough number was based on the normal Renard design (and PWM as the most probable mode), which the Ren SS24 uses a transformer at 400mA and there isn't much head room left on it (as I remember) and I wouldn't expect anyone to want to push their USB ports to the point of overloading them and damaging their computers.
The number may be low and the channel count could/can be more with careful design. But for the purpose of the post was a rough number given the current controller hardware/design methods.
However, if you have a functioning high channel count USB design then maybe then you provide the OP with the info and he can build one also.
There is a world of difference in power consumption between the PWM and non-PWM modes.
ripbme wrote: I still say one of you guys needs to move out to the Portland area so I can beg, borrow, or steal all your ideas. For a fair price to be paid in the beer of you choice of course... :rolleyes:
Paul,
Nice to have some company up here. I'm in Canby (obviously) and am making big plans for next year. I'm only running about 30 channels this year but hope to be back up to speed next year.
Greg
cbell wrote:Are you actually powering the boards via USB (the PICS don't get power from the wall?), or just using it for signaling?
Yes, the boards are powered from the USB port, so the PICs get power from the USB port. Actually, there is one controller on each USB port, so I'm using 2 USB ports (due to a protocol issue, not necessarily a power issue).
The Ren 112 (14 PICs) is using the +5 and ground leads directly from the USB cable. The Ren 224 (4 PICs) is actually powered by the +5V coming off the PicKit1 (my ZC and clock circuits were wired as a PicKit daughter board and I ran out of time, so I just ran those signals and the +5 and ground from the PicKit to the Renard).
P. Short wrote:There is a world of difference in power consumption between the PWM and non-PWM modes.
As I understand it, it is basically 256x as much (worst case), or 128x (average case), since there are 256 time slots and with non-PWM firmware the optos are only turned on during one of those timeslots, but with PWM firmware once they are turned on they remain on until the end of the AC half-cycle.
RavingLunatic wrote:My rough number was based on the normal Renard design (and PWM as the most probable mode)
That makes sense now; thanks for the clarification. I assumed that non-PWM was the normal case, and that PWM was only used with LEDs, DC devices, or relays, and that non-PWM was used for basically anything incandescent.
I guess it also depends on the number of channels that are turned on at any given time, and that would vary widely from one sequence to another. I have a little "sequence analyzer" tool that tells min and max number of channels turned on during a sequence, if anyone wants to use that to estimate their controller power requirements.
RavingLunatic wrote:However, if you have a functioning high channel count USB design then maybe then you provide the OP with the info and he can build one also.
For the Renard 112, I just built a board and hooked it up to the USB port and a ZC circuit and it worked, but it sounds like I might run into problems if I use too many channels all at the same time. I should probably switch it over to use the P/S for safety.
don
