TLC59582: Questions regarding Registers and overall behaviour

Maximilian Hillenberg

Part Number: TLC59582
Other Parts Discussed in Thread: TLC59581, TLC5957, TLC5958

Hello, I'm currently working on a board bring up with a single TLC59582, controlled by a MCU.

I have multiple issues, some of them really strange. I have based my code on the example provided here:

https://e2e.ti.com/support/power-management/f/196/t/687405

The first thing is, that the parts I had ordered from digikey were labeled "test spin from TLC59581", which is weird because I had thought the TLC59582 was a finalised product. Also, this was not mentioned on DigiKey or TI's website.

I managed to get it running, but the PWM output seem to be the inversion of the GS data I am sending.

Furthermore, I have no access to the FC1 and FC2 registers. Because of that, I can only use x1 Multiplexing mode. (Which is ok in my application).

I also cannot adjust the current gain, so the whole panel is way too dim.

I'd guess my lack of access to these registers is timing related, but I have checked it multiple times. I have just ordered a MSP432 launchpad to be able to inspect the precise timing of your example code.

I am now wondering why the TLC59582 even exists, if there is no pricing advantage to the TLC59581. I had chosen it because I only needed x16 multiplexing.

I would love to hear your opinion on these issues, I hope its not something obvious that I am missing.

Best Regards

Maximilian

over 4 years ago

0 Maximilian Hillenberg over 4 years ago

Prodigy 50 points

I would like to add, that the Voltage over a Rref of 6.2K is around 0.5V, even though it should be 1.209V? I checked on two boards.
Vcc is 3.3V.

0 Shawn Ding over 4 years ago

TI__Expert 3190 points

Hi Maximilian,

1. TLC59582 is our production device, it is the same family with TLC59581, the only difference is that TLC59581 can support up to 32 scan lines multiplexing, however, TLC59582 can support up to 16 scan lines multiplexing.

2. this technical reference manual may help you for programming http://www.ti.com/lit/an/slva744/slva744.pdf

0 Shawn Ding over 4 years ago in reply to Maximilian Hillenberg

TI__Expert 3190 points

It should be around 1.2V, pls check if you connect the GND to the thermal pad.

0 Maximilian Hillenberg over 4 years ago in reply to Shawn Ding

Prodigy 50 points

Hello Shawn,

Thanks for replying.

Irefgnd is connected to Rref and to the ground plane with a via, as well as the thermal pad which has a via array to ground similar to the example.
Trace length is kept very short, so I don't really see an issue here.

I know slva744, and I have read it multiple times now. I still cannot adjust the gain settings.

Is it possible that the voltage reference is defective?

I am still wondering about the "test spin from TLC59581" marking.

I will test this setup with TLC5958 and TLC5957, they should be pin- and almost software compatible as far as I can see.

0 Hardy Wu over 4 years ago

TI__Genius 12915 points

Hi Maximilian,

Could you help to capture a plot of GCLK / LAT / SCLK / SIN. Maybe the FC1/2 register value was not been changed.

0 Maximilian Hillenberg over 4 years ago in reply to Hardy Wu

Prodigy 50 points

Hi,

I have captured a logic sequence with Digilent WaveForms (https://store.digilentinc.com/waveforms-previously-waveforms-2015/),
you can view it for free by selecting Analog Discovery ( /2) DEMO mode at startup.

I could also provide a jpeg, but with the software it should be a lot better.

GCLK is ran continuously because I have noticed no difference to outputting sequences of 257 cycles per line and Im not using the x16 multiplexing mode.

Timing is very loose because of the mcu

DigilentWaveforms_09Mar2020_Maximilian.7z

0 Shawn Ding over 4 years ago in reply to Maximilian Hillenberg

TI__Expert 3190 points

Hi Maximilian,

The first step to debug is to make sure the REF voltage is right, it should be 1.2V.

Please help to do the below steps

1) Disconnect all the interface wires from MCU.

2) Check the VCC power, and the current when power on(~11mA)

3) Check the V IREF voltage to see if it is 1.2V,

4) if not, please check the schematic/ layout/ soldering, let us know the result

5) if so, connect the digital interface, and check the command format with the timing diagram in TRM file.

Regrads,

Shawn

0 Maximilian Hillenberg over 4 years ago in reply to Shawn Ding

Prodigy 50 points

Hello Shawn,

I have found it! It was an embarassing mistake to make, but I had swapped Pin56 and Pin1, and therefore had a short of Pin1/Iref to GND.

V IREF is 1.2V and i have access to the registers now.

I imagine I was getting dim and inverted, but correctly modulated output from the precharge feature before that.

Thank you very much for your help.

If anyone else has issues with this driver, you can PM/ or mention me in your thread, as I may have made every mistake there is :)

Power management

Power management forum

TLC59582: Questions regarding Registers and overall behaviour