1804 on PixAd8

[riri7707]

Ok,
LS located under RJ45 board and ordered with ACT series from Mouser (We have here Mouser from Germany, so it's fine : $.075 per chip (4 ordered for 2 x PixAd8)....
Actually i will have 14 ribbons 5 meters 1803 ( 30 chips per ribbon)...
So for moment all seems fine...
Just need the firmware please
Regards from France...
Henri

 

[David_AVD]

Hmmm... the post from danozfw disappeared.

Anyway, the chips to replace are the 74LS541 at positions U7 and U8. You will need to carefully remove the top board to get to them. Use a thin flat blade screwdriver (between the IC and its socket) one end at a time to wedge them out.

Make sure you get the new 74ACT541 chips the right way around when re-installing. There is a mark or notch at one end of the chip to guide you.

When ordering the new chips, make sure you get the DIP package one, not the surface mount version!

 

[danozfw]

Yeah, realized I was wrong when I looked at the new chip..I did replace them and am now working successfully using a 3 meter extension (no flicker) with the following from Ray.

http://www.aliexpress.com/product-fm/374338740-TM1804-IC-led-pixel-module-DC12V-input-20pcs-a-string-wholesalers.html


http://www.aliexpress.com/product-fm/380807523-5m-waterproof-LED-digital-strip-by-tube-DC12V-input-without-controller-10pcs-TM1809IC-m-30pcs-5050-wholesalers.html

http://www.aliexpress.com/product-fm/486534040-TM1809-IC-led-pixel-module-3pcs-5050-SMD-RGB-LED-DC12V-input-20pcs-a-string-wholesalers.html

 

[danozfw]

It looks like I was a little hasty with my assessment of working "successfully". Running 1804 and 1809 modules and strips at 2480Kbps (I don't recall where I arrived at that number, I believe I read it in a post somewhere) in test mode on the PixAd8 I am able to light all the pixels with no flicker. However, using LSP I can run a sequence only with solid colors using the "On" command. Anything else (blends, ramps, etc) causes multicolor twinkling of the pixels. Is there something I may be missing?

 

[kane]

It looks like I was a little hasty with my assessment of working "successfully". Running 1804 and 1809 modules and strips at 2480Kbps (I don't recall where I arrived at that number, I believe I read it in a post somewhere) in test mode on the PixAd8 I am able to light all the pixels with no flicker. However, using LSP I can run a sequence only with solid colors using the "On" command. Anything else (blends, ramps, etc) causes multicolor twinkling of the pixels. Is there something I may be missing?

Are you using the latest firmware (version H) - if not, probably worth contacting Ed to get a copy.

Also, I believe 2000 is the optimal level, so maybe try that, although Ed did mention that 1809 may be different, and there are slow and fast versions of 1804, so perhaps try a few different options

Kane

 

[j1sys]

the final H release is now on our download page. it is mainly a fix for the TM180x. it is also a unified release with the initial source code for the P12R. Now i can fix problems rapidly for both products and keep them in synch.


Kane - you should download also. you have a pre-release H.


As Kane said, the BPS for the TM180x varies by chip and configuration. The TM180x uses a 3 bit encoded timing signal to transmit each actual data bit. for a 0 we send 100, for a 1 we send 110. So the BPS is actually the encoding bit speed not the data bit speed. according to our reading of the documents in english and chinese for the TM180x chips their should be 4 different settings:


TM1804 - low speed = 1000kbps, high speed = 2000kbps. most of the modules we've tested with work at 2000


TM1803/9 - low speed = 1666kbps, high speed = 3333kbps. this is theoretical based on the docs. we do not have any 1803/9 in house for testing.


So i would try all four values to see which works best with your pixels.


also the two driver chips (74LS541) are best replaced with 74ACT541 units. all orders in the last two weeks have been shipped with the ACT chips along with a tube of the LS chips in case of problems.


we were going to ship some ACT to customer but most that need them for TM180x have found it easier to pick up a couple of them locally. if you need some let us know and we can ship you a couple. but especially for our Aussie friends the shipping time could be troublesome.


The LS chips have worked for months with 2801 and 6803 at 5V. We think the 12V and the signal/receivers of the TM180x work best with the ACT and we have found no problems using them with the 2801 and 6803. but YMMV and we wanted to include the LS chips just in case for this season.


-Ed

 

[danozfw]

I did some quick sequence testing with the "H" firmware and everything seems to be working okay. I plan on doing some more thorough testing later this week. The 2000Kbps timing works best with my TM1809 strips as well as TM1804 and TM1809 modules.

Thanks for the help

Dan

 

[kane]

TM1804 - low speed = 1000kbps, high speed = 2000kbps. most of the modules we've tested with work at 2000


TM1803/9 - low speed = 1666kbps, high speed = 3333kbps. this is theoretical based on the docs. we do not have any 1803/9 in house for testing.

Just to bring up an old thread - I decided to pick up a string of these "ping pong ball" style pixels to see what they're like:
http://www.aliexpress.com/fm-store/701799/209889132-577940096/50mm-diameter-full-color-ball-type-led-pixel-module-pixel-node-DC12V-input-3pcs-5050-RGB.html

These are tm1803 chips, so I had to play around a little bit to get them to work.

At 2000 (which is what I run all my other tm180x stuff at), nothing lights up
At 3333, very random stuff going on.
At 1666, get a heap of flickering, and wrong colours lighting up.
At 1200, all correct colours lighting up, but a tiny bit of flicker.
At 1000, all correct colours lighting up, and doesnt seem to be flickering.
At 800, wrong colours light up.