News:

LATEST RELEASE:  FPP 8.5.1 - Download from here - https://github.com/FalconChristmas/fpp/releases/tag/8.5

+-+-

+-User

Welcome, Guest.
Please login or register.
 
 
 
Forgot your password?

+-Site Stats

Members
Total Members: 16898
Latest: Kap10G
New This Month: 14
New This Week: 2
New Today: 1
Stats
Total Posts: 135492
Total Topics: 16990
Most Online Today: 166
Most Online Ever: 7634
(January 21, 2020, 02:14:03 AM)
Users Online
Members: 2
Guests: 115
Total: 117

troubles getting UCS2904/TM1814 to work properly

Started by ocupmoc, January 09, 2024, 04:40:24 AM

Previous topic - Next topic

0 Members and 1 Guest are viewing this topic.

ocupmoc

Hi all,

I've been looking at some new pixels to decorate the outlines of our house. Browsing Ali I came across a 12VDC string with UCS2904 or TM1814 IC's. Link to Ali
Is there a major difference between the protocol WS281x uses and the protocol UCS2904/TM1814 uses or could it just be a matter of modifying timing?
The only response I get out of that string is that it stops it's automatic changing of colors whenever I try to control them.
I am using 2 BBG with homemade rgb-123.com 24 port capes and also 1 Pi Zero, all of them show the same phenomenon.

Any and all help would be greatly appreciated by this newbie.

Cheers,
Arco

tbone321

I still have quite a few TM1814 pixels in my display and while they will work using the WS821x protocol, you cannot put them in the same string with WS281x nodes.   I run mine using a F48 and have no issues with them.

ocupmoc

Thanks @tbone321 . Do you have any idea what the voltage is on the output for the strings?
Looks like t could be an electrical problem. Both the Pi 'cape' and BBG cape have an output of 3.3V while the TM1814 expects something around 5V.
Just ordered a bunch of level shifters to see if this will help, will update in a week or 2.

CaptainMurdoch

How far is it to the first pixel?  Most (all?) WS281x want 5v on their data line as well, but will also work directly wired to the Pi's 3.3V GPIO output.  The level shifter definitely helps for multiple reasons, but you might try with a shorter cable and see if it works as a test.
-
Chris

MikeKrebs

Quote from: ocupmoc on January 10, 2024, 02:37:12 AMThanks @tbone321 . Do you have any idea what the voltage is on the output for the strings?
Looks like t could be an electrical problem. Both the Pi 'cape' and BBG cape have an output of 3.3V while the TM1814 expects something around 5V.
Just ordered a bunch of level shifters to see if this will help, will update in a week or 2.
All the capes I looked at at rgb-123.com have level shifters. Did you not install them? 

Also, the power needs to be injected as you can't run many pixels through the board.

If the data line at full on is only doing 3.3v, that seems like a board trace problem. Did you correctly align the pins so you are getting the 5v that drives the level shifters? 

If you aren't doing 100% intensity at full on, you can't really measure the data without a scope of some kind.

dkulp

To be clear, the TM1814 pixels would definitely NOT work.   It's a different wire protocol.   (it's actually inverted ws281x.   Replacing the buffer/level shifter with an inverting version may work).

Not sure on the UCS2904.
Daniel Kulp - https://kulplights.com

ocupmoc

Many thanks for all the feedback!

@CaptainMurdoch the first pixel is only approx. 50cm from the output, so that shouldn't be a real issue.
@MikeKrebs I am using the 24 port version which uses transceivers (SN74LV245ARGYR) which have a 3.3V output, they can actually also be used as level shifter from 5V to 3.3V
@dkulp I hope you are wrong about this otherwise it will take another couple of weeks before I will get it working properly. It is only 348 days until Christmas so I need to hurry ;)

CaptainMurdoch

Quote from: ocupmoc on January 11, 2024, 05:05:39 AMI hope you are wrong about this otherwise it will take another couple of weeks before I will get it working properly. It is only 348 days until Christmas so I need to hurry ;)

Listen to what he (Dan) says, he's put a lot of time and effort into it. :)

Here's an arduino thread discussing it: https://forum.arduino.cc/t/rgbw-tm1814-on-arduino-uno/606596
-
Chris

ocupmoc

Ordered some level shifters yesterday that can also be used to invert the signal, just in case  :D
Those should come in within 12 days.

@dkulp Dan, do you think these would suffice?: https://www.aliexpress.com/item/1005006358586542.html

Many thanks for your help lads.

ocupmoc

So, I got the pcb's in and tested both the levelshifters and inverters and both just output garbage. So I took the scope to see what is going on.
It turns out that the original buffers already output 5V despite what the datasheet says, without load that is but with load that shouldn't drop hardly anything.
I'm thinking I just continue to use the ws28xx instead of trying to fix something that might not be broken...

BTW it was really cool to see how the signal is built up with the scope.

Support FPP

+- Recent Topics

Light Elf Bluetooth IP65 Laser by Kap10G
Today at 11:45:04 AM

HA MQTT plugin for single model by s1godfrey
March 20, 2025, 08:21:46 PM

Virtual Display not working? by dkulp
March 19, 2025, 11:45:01 AM

PixelOverlay-Clock - Revisited Again by jnealand
March 18, 2025, 06:00:53 PM

PixelOverlay-Clock.php works but I have questions... by datajack
March 18, 2025, 04:38:32 AM

Backup Feature not functional by efamlights
March 16, 2025, 11:58:28 AM

WS2811, Raspberry Pi 5, GPIO, could not initialize. by Ulysse-srfc24
March 16, 2025, 05:07:48 AM

K16A-B Issue by nolatron
March 15, 2025, 01:52:26 PM

F16V5 and SRx1 v5 by jkp
March 15, 2025, 07:19:44 AM

Differential Board: PSU & Enclosure by hakko808
March 14, 2025, 09:17:01 AM

Powered by EzPortal
Powered by SMFPacks Menu Editor Mod