I'm wondering if it might be a good idea to come up with a new entry-level controller board. The Grinch currently fills that role, it's cheap and easy to build. The difficulty, of course, is that parallel ports are becoming increasingly rare.
One possibility is to resurrect something like the Renard8, either in its original dimming configuration or in a non-dimming configuration. The benefit of the non-dimming configuration is that the zero-crossing circuit would no longer be required, making it that much simpler and easier for people to build, test and use. One drawback to this approach is that it still requires the PC to have a serial port (which is becoming rare at the same rate as the parallel port), or a USB-serial converter (which is added expense and complication).
Another possibility that occurred to me would be to design a board that connects to the USB port with just an MM232R, a PIC and a RS485 transmitter on it. This would be fairly easy to build, having only three active parts on it and a few connectors (one of them a programmable part, unfortunately). This could be designed with or without dimming (same consideration as above), and the downstream data format might be either Renard or DMX (with the right choice of a PIC).
The goal, in any case, would be to design a single-board solution that is as simple to build as possible, with as few complications for the newby as possible. There may be other solutions that people might want to consider.
Thoughts?
I think this is an excellent idea.
I also like the idea of a choice between dimming and non-dimming as the whole zero crossing concept may be difficult for beginners to pick up.
Personally I would prefer to stick to the serial port, and give people the option to use USB -> RS232 adapters where necessary. At least this way people have both options to choose from.
Would this board be daisy-chainable? I'm thinking it would be good for my web control stuff.
I think its as good idea also - one suggestion to take it to the "next level" would be have a way to program it by just having it connected to the PC, no programmer needed.
fkostyun wrote:I think its as good idea also - one suggestion to take it to the "next level" would be have a way to program it by just having it connected to the PC, no programmer needed.
Someone could pre-program the processor chips with a USB and/or serial bootloader... Thereafter they could be updated or reflashed without a special programmer.
JEC wrote:Someone could pre-program the processor chips with a USB and/or serial bootloader... Thereafter they could be updated or reflashed without a special programmer.
It's a good idea. There are even mid-range PICs (like the PIC16F689) that are reasonably cheap and can update the Flash program code under programmatic control. It is a bit more work than I really wanted to do, involving both firmware (not too difficult) and PC code. I wonder if KC could be convinced to put something like this in Vixen.
The PIC16F689 (and cousins) are 20-pin devices, which would allow 12 PWM outputs along with a few spare pins for status LEDs or the like. So far I haven't found any 14 pin mid-range parts that software-updatable program space.
Oops ... I mis-read the datasheet - the Flash memory is apparently read-only for the PIC16F689-series parts. So I need to keep looking.
Further edit...I would now be much more inclined to go with the PIC18F1220 (or one of it's cousins). The PIC24 series is certainly tempting, but IFIRC they are power hogs, so that it might not be possible to keep the power budget within the 100 mA available with the minimal USB configuration.
There was once a product called Dasher, It was serial came strate off the serial port just on/off just like you meation very easy not a lot of components on the board 8ch with pic16F628. I still have some controlers, self etched boards all the parts exept the PICs. I would like to run them off of Vixen someday mabey. I do have clone firmware for the chip that came out after christmas cave got bought, never used it. For that matter someone could write new firmware and still use the old boards. I still have all pcb files for this also but would probably want to redew them. I don't know if there is any interest in reserecting this old system or not.
Russ
I have been toying with this idea as well
i personally am a big fan of the 18f2550 chips. little bit more exspensive than the low end chips but it would allow 12-16 outputs and i dont think anyone would care of the project cost anotehr $2-$3.
this would allow USB or serial connection without any other chips. there is USB and Serial bootloader code available allready and also some simple diy development board designs allready around. search bitwacker for 1 example
i also would be interested in the 24 series pics but maybe they shouldnt be in a beginners board.
Also with the whole ZC thing i would love to get away from triacs and the ZC problem.
can we do some other type of design and do away with ZC alltogether? maybe dual triacs or something. (not my area).
anyway, just my two cents
Peter
The 18F chips - even the small ones - are cheap and powerful. Lots of DIP packages for those interested. 24F looks cool but they require a lot more babysitting re: power supplies, etc.
$0.02
How about if we expanded the grinch.
Now that I finally understand the concept of a serial driver it makes the most sense.
Develop 16 & 32 channel grinch based boards with on board SSRs.
The RJ45 connector would be for serial connection of multiple grinches.
In addition it would have a pin header that would accept a piggy back control module.
That module would be like a Ren-C or the new DMX-grinch or future wireless designs.
The concept being that if the control module output a serial string it could be plugged into the grinch-ssr.
To switch from paralell to serial-grinch to DMX-grinch would only mean swapping the control module.
Just another off the wall idea.
Joel
A lot of this depends on who is going to step forward to do it.
I don't have a particularly strong itch to use such a board myself, since I have so many other designs available to me. So if I end up doing it, the design will have as little new development as possible. This means that it would use an MM232R module, since I don't have enough itch to do a native USB design (especially since RJ ended up taking a module approach as well). It also means that it would probably use one of the existing designs, possibly an 8-channel Renard. Since it is intended for newbies it would keep 110 VAC off the board (i.e. no integrated SSRs) in the interests of reducing possible trouble spots. And I'm debating whether I should omit the zero-crossing all together (to avoid confusion on the part of the newbies), or leave it as an option. And unless I can get more than 100 mA out of the MM232R it would be limited to 12 channels so that it could use the available USB power.
On the other hand, if someone else wants to do it themselves in some other manner, more power to them. As long as I thought that it was sufficiently simple for the newbies to build I would keep my mouth shut (except when asked).
