Beta-Portable Wired/Wireless DMX Tester (See first post for specs and features)

[dlhoppe]

Thanks a ton for the information guys! I will do my best to implement the debounce in software.

In the meantime, here's a look at the schematic using the M14490. It's not much more complex than feeding the switches and encoder into the PIC with pull-up/pull-down resistors (it has built-in pull-up resistors by the way). It also eliminates the need for any debounce logic in the PIC. Although clearly the more attractive option is software debounce resulting in lower cost for this device. I will take a crack at a debounce routine and keep this in my back pocket as a fall-back solution.

I was also concerned about the current required for driving the LEDs directly. At 20ma for each segment, worst case is 140ma to represent the number "8". This is based on the fact that I'm multiplexing one digit at a time. The PIC can handle up to 200ma across all ports, so I think I have enough headroom to go with this simpler approach. Let me know if I'm missing something here though.


Thanks again for your input,
Dave

[dlhoppe]

Yes, I missed the voltage requirements for the ST485. I must have misread the data sheet.

Ok, I added an additional regulator to power the ST485. One thing I'm not clear on is the need for capacitors on the input side of the regulators. Is this necessary when using battery power? The two capacitors (for input side) that are specified in the regulator datasheets are of different properties. If I have to use these, will they interfere with each other being on the same power input lines?

I attached both versions of the circuit. One for hardware debounce and the other using software debounce.


Thanks,
Dave

[dlhoppe]

google "Ganssle debounce" for some algorithms and sample code.

/mike

Mike,

Thanks for this valuable tip. The examples were easy to understand for a newbie like me. I found the info I needed to add really simple debounce logic to my existing code. I'm using the "counter" approach to wait for stable encoder input before acting on it. I added an easily adjustable threshold so I can tune the sampling duration.


Dave

[n1ist]

At quick glance, it looks like the VIH of the ST485 (2V) is less than the VOH of the of the PIC (2.4V) so it is OK to drive the 5V part from the 3.3V.  

Looking at the regulators, they both want a 10uF cap on the output for stability.  The applications schematics also show a 10uF one on the input; I would add one or at least provide the footprint for it on the board; they are cheap.  You will run into a problem with the 5V regulator, however, as 6V doesn't give enough headroom.   There are some RS485 drivers that run on 3.3V, and they are not much more expensive than the ST485 plus the extra regulator.

Looking at the LED, there are two things to take into account - sourcing the segments, and sinking the common cathodes.  You have one pin for each cathode; that pin would have to sink the full current from 7 anodes.  That's why I mentioned needing a transistor there as the PIC won't be able to handle it.  BTW, 20mA per segment seems a bit high these days (though the average current will be about a third of that due to multiplexing, so you may need to hit them that hard...)

/mike

[dlhoppe]

Mike,

Yeah, I looked a bit for a 3.3v 485 chip, but I went ahead and added the other regulator instead. I'll look a little more, or even consider driving the existing one with 3.3v.

Oh yeah!  A big "duh" on my part. I was so caught up with each anode being sourced by a separate pin that I completely overlooked the fact that I had a single pin sinking the common cathode. I assume the current would be "halved" between source and sink for each segment, but even then it's too much for a single pin to sink all the segments. Although I'm not all that sure about the actual current requirements for the LED. I find it odd that the data sheet for the LED unit does not explicitly state the context for the current requirements. I have to assume that it's 20ma per segment. Here's the data sheet. If you could take a minute and look it over, I would appreciate it: You are not allowed to view links. Register or Login  I wouldn't be suprised at all if I'm completely misinterpreting it.


Attached latest schematic.


Thanks again,
Dave

[n1ist]

A quick search on Mouser turned up SP3490CN-L in SOIC or SN65HVD1780P in DIP (but that's very expensive...)  Digikey has ADM3493ARZ or ST3485ECDR in SOIC or ISL3294EIHZ-T in SOT23-6 if you are short on space.  They are all 3.3V parts (the SOT23 one is just a transmitter, but that's all you need.)

You will need a resistor in series with the base of each of those transistors.  As for current, it does look like the recommended current for those displays is 20mA; with 10mA they will be half as bright.  Since you are multiplexing them, they will be a bit dimmer than that.

/mike

 

[dlhoppe]

Cool, thanks again Mike. I'm learning a lot!

Transistor base resistors calculated (correctly I think) and the 485 chip has been substituted for a 3.3v unit.



Dave

[dlhoppe]

Ok,  thanks to Mike and RJ for helping out a total newbie with the hardware aspect of this project. I think I'm at a point now where the circuit design is stabilized enough that I won't let out any smoke. I'll post some updates once I get the hardware put together and start debugging the software.

Latest schematic attached.



Dave

[onesmoothhead]

I would be more than willing to fund and Beta test my portion of this for you.  I know I have been mute on this particular subject, but i am really am a tester, more than a design type of guy/personality.  I will be happy to run this back and forth and I even have a test bed across the street (+25yard) to do all the great data collection. 


You are really on to something, and if this is the only portion I can offer any assistance. I would consider it a privilege. I have been a member for over a year with zero contribution.  I am a communication/securtiy/fiber guy by day. I test constantly and hope you consider my offer.

Again, good work on this tester.


Kevin Lederer

[dlhoppe]

Kevin,

At this point there's no real funding to be concerned about. The hardware cost is minimal and the software is only costing me time. Not sure about PCB printing though. I'll worry about that when the time comes. I'm hopeful others with that particular experience will guide me or step up and participate themselves.

If you have a PIC programmer and want to breadboard the hardware I'd be happy to send you the hex file for testing once I get it to a reasonably runnable state. I believe all the part #'s are listed on the schematic. Although I can post a more detailed BOM later tonight with the specific Mouser part #'s that I'm using.

Thanks for offering. I like the fact that folks are willing to step up and help around here. Makes for a fun project when there's different people with different skills willing to share them. This has been a fun project for me. I always wanted to tinker around with circuit design and microprocessor control. I'm hopeful this will also be useful once it's done.


Dave

[onesmoothhead]

Dave, I do have a PIC programmer and will check into all this weekend.  When you are ready I will be here. You are doing a great job.

Kevin

[dlhoppe]

Here's the current parts list for the prototype. Obviously a work in progress. Once this circuit design proves to be workable I'll spend more time finding cheaper alternatives to some of these items. Notably the 3.3v 485 chip. The current price for this BOM is $26.38 (PCB and enclosure not included) with the 485 chip being the most expensive item at $3.91.


Dave

[Made2Rock]

This looks interesting

Someone pass me some popcorn

Joe

[PJNMCT]

 

[zman]