Author Topic: Communication Issue  (Read 10654 times)

Offline tbone321

  • Sr. Member
  • ****
  • Posts: 4054
Re: Communication Issue
« Reply #15 on: March 02, 2015, »
What exactly were RJ's instructions?
If at first you don't succeed,
your not cut out for sky diving

Offline LasVegasF250

  • Full Member
  • ***
  • Posts: 91
Re: Communication Issue
« Reply #16 on: March 05, 2015, »
You are not allowed to view links. Register or Login
Follow RJs instructions with the new chips and it worked perfectly.

You are not allowed to view links. Register or Login
What exactly were RJ's instructions?

Bump

Offline cpt_hammer

  • Jr. Member
  • **
  • Posts: 14
    • The Hammer Lights
Re: Communication Issue
« Reply #17 on: March 05, 2015, »
Sorry. RJ did the initial post but someone else had the fix.  Look at this post:   You are not allowed to view links. Register or Login

Offline cpt_hammer

  • Jr. Member
  • **
  • Posts: 14
    • The Hammer Lights
Re: Communication Issue
« Reply #18 on: March 05, 2015, »
Specifically the post from technoweenie on 4 January:

UPDATED: Success!!! See update at bottom of post.
--------------------------------------------------------------

Now that Christmas is over and mostly packed away, I got around to building my Aether II from the latest co-op. Having a problem trying to get it working. When I power it up, I get a bright flash, and then the white driver stays on in operate mode. No DMX control at all.

Here is what I have done / tested / observed;

- With switches all towards LED, the board powers up, flashes all colours and then the white stays on. I measure 0.2 volts on white driver control pin and 3 volts on all other colours.

- When I switch outer switch to program, the board powers up with a flash and then sits dark as expected.

- Loaded the test hex file and board cycles all 4 colours with a brief dark time between each as expected. Runs test sequence if switch is in operate or program.

- Loaded normal hex and set address to 5 at program location 2800. Write to board and same behaviour as above. No control and white stays on.

- Placed Aether first in a DMX chain with a known working pixel string to confirm DMX. With Aether powered off and terminator switch out (towards LED), pixel string responds as expected. When I power Aether on, I loose control of the pixel string.

- Based on above, I swapped out 485 driver ship. Same results.

- Swapped out PIC and reprogrammed. Same results.

- Re-sweated all solder joints on PIC and 485 IC sockets and all around this area.

I don't have a Lynx dmx dongle, but I tried my Entec Pro USB dongle with the Lynx programming utility using instructions in wiki. No response from board. (no flashing red).

I am programming the board with a Pickit 3 and the newer beta version of the utility. I could not get the ver 1 of the utility to work. Kept crashing with a windows exception error, but I did get beta 1 version working. Could this be the problem? I have the proper settings, PIC 24 family, PIC24FJ32GA002 selected, powering PIC from Pickit dongle. It writes, reads and verifies correctly.

I am stumped, especially since the test hex file works fine. Also strange that the dmx chain downstream is interrupted when I power the Aether on. Could a regulator be oscillating? I don't have a scope at home to look at this, but could take it to work.

One question, does the Aether program hold the last known dmx value in memory somewhere? Could that be why white comes on in operate mode only? But why would it not come on with test hex file?? Hmmm.

Edit:

- measured the voltage at pin 2 and 3 of the 485 chip. Expected low (active low for RE and low to disable DE), but I am seeing 2.0 volts there. See the same at pin 15 of the PIC as well. If I pull the SN75176 I then get 0.7v as expected. No shorts or solder bridges I can see anywhere along the trace.

Also downloaded the latest v3 beta of the pic utility and same thing. Loads, reads and verifies no problem. I can switch back and forth between the test hex file and regular aether file. Test file works as I would expect. Hmmm... strange gremlins indeed. Think I will bring the board to work and put a scope on it to poke around.

Edit #2:

I had a look at the PIC data sheet and noticed a couple of things. Pin 8 is not connected to Vss ground and pin 13 is not connected to Vdd 3.3v supply. The data sheet advises they should be along with a 0.1 cap across them.  I also notice the internal regulator is enabled by pin 19 pulled to ground. This should then require a 10uF cap from pin 20 to ground, however it is a 0.1uF cap instead. I might be off on a wild goose chase, but I am stumped on this without a scope. Let me know your thoughts.

Edit #3 Success!!! <does happy dance in shop / knocks box of parts over...>

I came back to the aether board after a bit of a break. Replaced the 0.1 uF cap with a 10uF 20v tantalum that runs off pin 20 of the PIC (cap closest to switches right over middle of pic. I then soldered a jumper from pin 8 to pin 27 to connect second ground rail and pin 13 to pin 28 for second supply rail. Took 0.1uF cap I removed and placed it across pin 8 and 13. I tested the board and it now was not acting erratic, but the DE line was still sitting at 2.0 volts. I soldered a 1K pull down resistor to pin 2 and ground of the 485 chip. Problem solved!!!

I think what is happening is that a small percentage of users are having the PIC chip blow the pin 15 output internal to the chip due to pin 8 not being connected to ground. It may be trying to sink too much current thru the internal ground bus. The data transmit enable line isn't really needed in this application, so I just pulled pin 2, 3 of the 485 chip to ground to release the dmx bus.

As a quick check if anyone else experiences this issue, measure pin 15 of the PIC. It should be 0.7v or lower with the aether powered up with normal hex file. If it is like mine and sitting at 1.9 to 2.0 volts, solder a 1k ohm resistor from pin 2 or 3 of the 485 ship, to pin 5 ground. This will permanently put it in receive data mode.

Also, the beta release of the Pickit programming utility did work for me in the end. Also the Entec DMX pro was able to program the start address after I fixed the board up. Now we are shining!!

Maybe I could decorate for Valentines day???!!!

Offline woodbury

  • Jr. Member
  • **
  • Posts: 2
    • บล็อกเกมสนุกมันส์ๆ
Re: Communication Issue
« Reply #19 on: February 29, 2016, »
What RS-485 is it using?
ติดตามบล็อกเกมสนุกมันส์ๆได้ที่นี่
You are not allowed to view links. Register or Login
You are not allowed to view links. Register or Login
You are not allowed to view links. Register or Login
You are not allowed to view links. Register or Login

Offline LasVegasF250

  • Full Member
  • ***
  • Posts: 91
Re: Communication Issue
« Reply #20 on: March 01, 2016, »
These are the one's that I received with my last build.

SN75176AP