TM1804 protocol on the TP3244

[Tabor]

There has been alot of discussion on the forums about running pixel nodes at higher voltages and running longer string without power injection.

Below is a quick video of the TP3244 running the TM1804 protocol.

This particular string is running at 12v and has 100 pixels on it, without any addition power injection.
I am using vixen and I am just running through some test patterns.

This is just one string on one output, all four could be used.

Cost is around 55c to 58c per node from Ray, exc postage

So we have another choice in pixels on the TP3244.

There are some good reason to use these type of pixels and some reason why I wouldnt.

So let the questions and discussion fly.

TP3244 running TM1804 protocol

[TheBanker]

I'm at work and cannot see the video, its blocked, but.....are you saying you will be able to run four strings of 100 pixels each for a total of 400 pixels, at 12v with the TP3244?  Have you tried to push it greater than 100 or did you just stop there because that is the length of the string?

Will

[hbomb341]

That's AWSOME!!!!

Harrison

[dmoore]

Tabor -

Excellent job there on adding the TM1804 protocol.  The TP3244 really is the swiss army knife of RGB pixel boards!

My question is on the speed of update - you had mentioned that the refresh rate is lower than a 2801 - how low is it?  Also, what is the color depth of the TM1804 compared to the 6803 and 2801?

Thanks!

[mrpackethead]

colour depth is 8 bit..( 256 )..

[mrpackethead]

@tabor

I'd love to compare notes on this with you tabor,  the converstation we had last night, made me do some more work on it, and i got it running as well..     I dont' have 100, just 5, as i just got a few sample chips to mess with.

Is there any chance you could measure the Current that is flowing out of your power supply into the entire string and the voltage ( i'm assuming 12V ) that is across it?  ( set all the pixels to white ( FF FF FF )   I know you dont like doing Maths, so i'll multiply the number and figure out what power its pulling..

I smell something getting very hot here.

[ɟɐsʇǝppʎ]

Great work once again Tabor. When will you release this firmware for the tp3244 to support the TM1804 protocol?

[TimW]

Assuming the leds are consuming the same current as usual (in parallel off the supply) I was wondering where the extra power @12V is going to be dissapated as well! Is it twice as hungry?

[XDU]

Excellent work T

[ɟɐsʇǝppʎ]

So these appear to be the same strings using the same TM1804 IC chip running at 12VDC that RJ appears to be using and going to sell in his group buy. Well its great to see there are a lot of choices out there for the community.

It would be interesting to do a wattage comparison between these strings and the 5vdc strings, i would expect these to be fairly inefficient

[gunther]

I was wondering about the power too..  I would think the chip designers would be much smarter than to just drop the full current across the 8-9V margin.  Is there some sort of switching regulator on the TM1804 die?

Also, does anyone have a pointer to the TM1804 data sheet in English?  The only thing I have is in Mandarin.

[TimW]

Is there some sort of switching regulator on the TM1804 die?

hope so!

[ɟɐsʇǝppʎ]

So to keep things simple for all i am just going to use round numbers and an online calculator to show the inefficencies of running a single LED at 12vdc compared to 5vdc.

The calculator is as http://ledcalculator.net/

Power supply = 12vdc
LED voltage drop = 3v
LED current rating = 20ma
Number of LEDs = 1

The calculator reports

You will need 1 x 470 ohm 1/4 watt resistor.
Each 470 ohm resistor consumes 188 milliwatt.
Total power consumed by the resistors is 188 milliwatt.
Total power consumed by the LEDs is 60 milliwatt.
Total power consumed by the circuit is 248 milliwatt.
Total current drawn by the circuit is 20 milliampere.

Power Supply = 5vdc
LED voltage drop = 3v
LED current rating = 20ma
Number of LEDs = 1

The calculator reports

You will need 1 x 100 ohm 1/8 watt resistor.
Each 100 ohm resistor consumes 40 milliwatt.
Total power consumed by the resistors is 40 milliwatt.
Total power consumed by the LEDs is 60 milliwatt.
Total power consumed by the circuit is 100 milliwatt.
Total current drawn by the circuit is 20 milliampere.

So the basic comparison 

12vdc you will use 248mw
5vdc you will use 100mw



Now that is a 248% increase in energy used for the same light which means you need more power supplies to run the same amount of lights and an increase in your electricity bill.
More energy is being dissapated through heat than what is used to drive the LED at 12vdc


 

[Tabor]

Some quick info.

Current draw at 12v is interestingly 3.683A.

44W total

I would have expected the current draw to be greater.

Talking of the heat, whenworking out the protol to this nodes, I found them to get warm.
When using them normally they dont seem to heat at all. When leaving the colour at FFFFFF they do warm, but only after long operation.
Prob not an issue in normal use.

[AussiePhil]

Tabor

Great work as always.

OK so we have 44W of power input, approximately 14W above the theory level of a .3Wx100count 2801 string.

44W is nearly four times the level specified for POE applications on CAT5 and the current of 3.6A is over 10x the current specified of 350mA.

There is no way i would supply power for these over CAT5

must be time to order a 128 count string to have a play.

Phil