DiyLightAnimation
Hardware => Lynx Express => Topic started by: dmorey on July 09, 2012, 08:05:14 pm
-
The short story is that we have four Lynx Express controllers that aren't working. I know at least part of it is related to not so great soldering. (We have the desoldering thing but haven't started.) We put them together a couple of years ago after mass technical problems with Animated Lighting, but got frustrated and just shelved the whole thing. We have a dongle for it that I believe is operational.
We also have 3 Animated Lighting boxes (and RU485 converter and Monster Brain). They all are half working with various non-working channels. They've been like this for years even after getting shipped back to AL a couple of times, hence our attempt to switch to DIY. We'd love to somehow get all these controllers to work together. I was running Maestro on Animated Lighting, and was in the process of converting files to run Vixen instead.
Anyone local in Austin who'd like to have a peak? I know this is a tall order so thanks for even considering ... I'm in northeast Austin. Willing to drive to Round Rock or San Marcos. Houston might even work since I'm there every couple of months to see family. Was hoping to meet in person as we still feel fairly overwhelmed...
-
Here's a discussion link to where we were when we gave up:
http://diylightanimation.com/index.php?topic=4022.0;topicseen
Has some controller pics too. Never got beyond what we discussed on that thread.
-
Sorry, not geographically near you
Hope you get them working, that's a lot invested ... I can understand the frustration
Good luck
-
Lone Star Holidays has a pretty active group in Austin. You might join there and ask the same question.
http://www.lone-star-holidays.com/forums/index.php
I'm three hours away and don't get to Austin much since my son graduated from UT :)
-
Lone Star Holidays has a pretty active group in Austin. You might join there and ask the same question.
http://www.lone-star-holidays.com/forums/index.php
I'm three hours away and don't get to Austin much since my son graduated from UT :)
I offered in a PM -- along with the suggestion of our local LSH meet, but wanted to see if others might have some time. I'm in the middle of a move to a new house, so my time is limited for the next month or so.
-
Thanks for the suggestions all. I will check out the Capitol City group. We moved from Florida last year so hadn't really gotten involved here.
-
If you dont get any offers there I'll look at them but you would have to ship them to me as i'm in Odessa Texas.
-
If you can make this weekend, we might have critical mass to put on a Sunday meeting. Bring them with you and I'll bring a working setup that we can use to compare and swap out pieces to localize the source of your problems. I'll bring a PICkit to do programming, because that has been identified as an important step.
Follow along in the LSH forum and I'll attempt to nail this down.
-
We will see you on Sunday. I'm looking forward to the challenge of those boards. If not I have a sledge hammer you can use. Lol
-
We will see you on Sunday. I'm looking forward to the challenge of those boards. If not I have a sledge hammer you can use. Lol
will has a desoldering station that might be a great tool to help with this work. :)
-
We have more info on Deanna's LEs after a few of us worked together to figure it out -- unsuccessfully.
Any additional ideas will be helpful.
We examined 3 of the 4 units. There were solder points that we touched up, with no improvement. There isn't excessive solder, and no bridging (based on close inspection with a visor).
We had 2 working units from other members that we used to compare the 3 non-working units... after looking at part orientation, soldering, and swapping out all socketed parts -- we came up with nothing. The parts work fine in other units, and working parts still do nothing to help the bad units. We reflashed all of the PICs with current firmware. The PCBs are each v5.
The oddest observation is that the behavior is IDENTICAL across the 3 units in question. board led activity is the same, self-test behavior is the same, voltages (that we checked) were the same.
What is the problem? We have no DMX control of the channels. The led marked "data link" does not light when we have working DMX connected. We can verify the DMX though the other 2 working kits.
Channels do work through the self-test mechanism. We can view the board leds for each channel light in progression, and we can watch strings light up in turn. DMX control is the problem.
Again, we'd swapped out chips to verify them... and swapped them back into the working board to prove that the bad kits are not damaging the replacement chips.
In hindsight, there is one thing we did not check... the regulators. We did not verify each was in the right spot and not swapped. The heat sink had been soldered in place, and rather then remove them to do a closer inspection we left them as-is. Since the self-test worked, and DMX control was the problem we didn't know if regulator swaps might have a role.
We think we can rule out soldering, parts, directionality of chips and capacitors.
Because the problem is IDENTICAL between the 3 units, it seems more likely that there was a bad installation step that was repeated with each unit.
Can anyone offer some ideas?
Thanks!!
-
From what you say I am thinking the data leds are in backwards shorting out the dmx.
Sent from my Charge by Tapatalk
RJ
-
From what you say I am thinking the data leds are in backwards shorting out the dmx.
Sent from my Charge by Tapatalk
RJ
Are you suggesting that the "DATA LINK" led is backwards?
(http://img72.imageshack.us/img72/1108/ledp.png)
To test that theory, Deanna can run a 5V power source over each of the 3 leds sitting next to each other to ensure that each light with that same 5V+/ground order. right?
And ALSO, Deanna should check to ensure there is no bridge (in case we missed it).
A side question... has anyone ever seen the case where a bridge occurred on the front side of the board, and under the part where visibility is obscured?
-
If you applied 5V directly to the LED you'll burn them out. You need a resistor in line with the led. Another way to check it would be to lift one side of the 1 K resistor (one closest to the Leads) with the lead closeest to the DMX jack. apply voltage to the board (carefull 110V shock warning) if the LED is in right you will have 5V on both sides of the resistor to ground if in backwards then there will be 0V on both sides of the Resistor.
Yes I have on large part with lots and lots of solder but I really haven't like under a socket.
-
Deanna are you going to try checking the LED? We will meet again in August if you want to bring them again.
-
From what you say I am thinking the data leds are in backwards shorting out the dmx.
Sent from my Charge by Tapatalk
RJ
Are you suggesting that the "DATA LINK" led is backwards?
I read that as Data LEDS (plural as in 16) not Data Link Led (singular).
-
All 16 of the channel leds will light up without problems.
I did not think that part of the board to be the problem as they were working through the self-test mechanism. I assumed that control of the 16 channels would happen upstream -- where the self-test or DMX would drive their operation.
"...thinking the data leds are..."
I took led(s) to mean... (3 kits) X (1 data link led) = (3 data link leds)
:)
-
The one thing can't get my head around. If the Data link LED is in backwards I would think it would pass a DMX signal and the Express would work normal with out a Data link light.
hmm got to go to work
-
Can you post some pictures?
Check 485 chip near terminate jumper
If last unit in DMX string, make sure terminate jumper is present
Make sure remove for wireless jumper is not removed
Make sure terminate resistor is the 120 ohm resistor
Examine pic socket solder joints to make sure there is a lead coming through all holes (pin not bent under on socket)
Check voltages to make sure regulators are not swapped. The '485 chip should have 5 volts on one of the pins and the pic should have 3.3. I don't have a board handy, can someone tell us which pins those are?
Good luck!
RM
-
Can you post some pictures?
Check 485 chip near terminate jumper
If last unit in DMX string, make sure terminate jumper is present
Make sure remove for wireless jumper is not removed
Make sure terminate resistor is the 120 ohm resistor
Examine pic socket solder joints to make sure there is a lead coming through all holes (pin not bent under on socket)
Check voltages to make sure regulators are not swapped. The '485 chip should have 5 volts on one of the pins and the pic should have 3.3. I don't have a board handy, can someone tell us which pins those are?
Good luck!
RM
Deanna has the units back in her possession, and she'll need to provide pictures.
We did check both 485 chips, swapped them, swapped them back. we know the chip is good, and we've verified that the legs are not bent.
We did NOT enable DMX termination. this was a test bench setup, and we did not have multiple items on the DMX bus -- only a USB DMX dongle, and then the unit under test. All 3 failed in this configuration, while our 2 working units continued to work in this same configuration.
checked the colors of the resistors, and they matched the working units.
didn't check the sockets/board connections for bent pins... but this is unlikely the cause for 3 units having identical problems.
We did not check the regulators... unfortunately. didn't know the pins and expected voltages... but in hindsight we could have compared to the working units. :(
It sounds like the first thing to test is the data link led... though I don't understand why a LED in reverse would prevent channel control -- i'd expect only the LED to be off.
The second thing is to test voltages on the regulators. Having more info on specific pins and voltages will help Deanna to make these checks.
Thanks.
-
Normally when the board gets no dmx data its one of three things (assuming the soldering is done correctly and the parts are right)
1 - RS-485 chip is bad
2 - Soldering and or RJ-45 jack is bad (DMX IN)
3 - Data LED is backwards
One thing that could also be tried. Move the cat5 from the DMX IN to the DMX Out
Note: NOT DMX Out 2
Rick R.
-
Here are pics of two of the LEs. The third one is at Bill's house as he swapped one of his good ones for one of my screwy ones.
I can try testing other stuff, but am new to this and not really clear how to perform the suggested tests.
-
One thing that I had done on one of my PBC's was then I inserted the sockets was bent the pins. You should be able to pull out the chips and check the connection from the top of the socket and the back of the PCB's where the pin comes thru the PCB. while you have the chip out also check the top of the sockets with a meter and make sure there are no shorts between the pins.
-
I pulled the Data Link LED flipped around backwards. Took Dave M DMX tester (yes that was plug for it) and applied a DMX signal into the express. Channels all work with Data Link in backwards.
-
RS485 pin 5 gnd
Pin 8 5V
Pic pin 28 3.3V
pin 27 gnd
I'm very sorry Deanna my eyes are bad can you take another picture of the back side around the DMX jacks/RS485 chips
-
If you look at picture #3, there seems to be a dark ring below the outputs below LED 10 and 12. Possible arcing?
-
If you look at picture #3, there seems to be a dark ring below the outputs below LED 10 and 12. Possible arcing?
on one unit only, there was a short when water entered the case. The channels test out ok using the test mode.
-
Another photo of the back and also a close up of the back near the 485 chip. I can unscrew the other and take photos of the back but pretty sure it is the same thing on both. Hopefully these pics are better. It's hard to get the lighting right since the solder reflects.
Yes, that dark ring is from an ill-fated attempt to have it run a show in the rain. Only on that one box and didn't seem to affect it. (And, no it didn't run that show. We already had a show up and figured it was worth a shot.)
Pretty sure it isn't bent pins. We already checked that.
-
PS, Hello again Bill! These were the ones that you PIC programmed and looked at a couple of years ago when we used to live in Orlando. The fun never ends, lol.
If you look at picture #3, there seems to be a dark ring below the outputs below LED 10 and 12. Possible arcing?
-
Remove for wireless needs a jumper on it.
If you don't have a jumper, wrap a piece of bare wire around the two pins.
It Will act exactly as you say it is without this jumper.
If this is the only board that is hooked up, you also need a jumper on the terminate pins. It may work without this jumper, but you may see unpredictable results.
Rm
-
The only time you should remove the "remove for wireless" jumper is if you have a ex/RX module installed.
-
The only time you should remove the "remove for wireless" jumper is if you have a ex/RX module installed.
the LEs were definitely missing this wireless(er... wired) jumper!
the terminate jumper was also off... it had wire insulation in place of a jumper -- in other words, jumperless.
Bill, Deanna -- can you give this a try? ...this sounds like the best explanation so far (and the most embarrassing as well :) )
...i just looked back in the earlier thread. i see that one board has the terminate jumper, the other does not -- BUT THEY ARE BOTH MISSING THE WIRED JUMPER.
-
I'll ask my hubby if he knows how to do this, but honestly it's like that whole statement is in Greek to me. I wouldn't know where it's located or what to put there. Also, I'm guessing it's a part I'd need to order? If it's not on there we don't have it.
The only time you should remove the "remove for wireless" jumper is if you have a ex/RX module installed.
the LEs were definitely missing this wireless(er... wired) jumper!
the terminate jumper was also off... it had wire insulation in place of a jumper -- in other words, jumperless.
Bill, Deanna -- can you give this a try? ...this sounds like the best explanation so far (and the most embarrassing as well :) )
...i just looked back in the earlier thread. i see that one board has the terminate jumper, the other does not -- BUT THEY ARE BOTH MISSING THE WIRED JUMPER.
-
If you look at the board there is a blue jumper in place where it says TERMINATE, there should be one also there it says REMOVE FOR WIRELESS. If the wireless module is not installed the blue jumper (or any jumper, something thats makes the too pins connect together). You might be able to test just by moving the blue jumper from the TERMINATE to the REMOVE FOR WIRELESS, But looking at the pic's you do need a jumper on both.
-
all you need is a jumper (little blue plastic part) and place if over the two pins coming out of the board. no tools required. if you have other kits, you may already have jumpers you can use in case you've misplaced the original that came with the kit.
otherwise i'll bring some spares when we have or next austin get together.
The arrow is pointing to the jumper to be installed:
(http://img546.imageshack.us/img546/8651/jumper.png)
Note the legend on the board: :)
(http://img221.imageshack.us/img221/1245/jumper2.png)
-
For testing you could just wrap a bare wire around both pins or make a little ball of tin foil and stick on the two pins. Anything that will short the two pins together.
-
Or just wrap a small piece of wire around it.
-
Or just wrap a small piece of wire around it.
paper clip?
jumpers from old computer hardware (hard drivers, etc)
:)
Rick R.
-
PS, Hello again Bill! These were the ones that you PIC programmed and looked at a couple of years ago when we used to live in Orlando. The fun never ends, lol.
Hi Deanna,
Welcome back to the fray, when you guys met up with the Austin group, were the PIC's reprogrammed? Just to make sure I did them right, and they weren't corrupted some how. And the jumpers would make a difference too. Hope you guys get them running,
Bill
-
PS, Hello again Bill! These were the ones that you PIC programmed and looked at a couple of years ago when we used to live in Orlando. The fun never ends, lol.
Hi Deanna,
Welcome back to the fray, when you guys met up with the Austin group, were the PIC's reprogrammed? Just to make sure I did them right, and they weren't corrupted some how. And the jumpers would make a difference too. Hope you guys get them running,
Bill
I did reprogram each of them, as a standard first step. Reprogramming made no difference.