DLPDLCR3310EVM: Power Supply and Start-Up Issues for DLP EVM

Hello Joshua,

Welcome to the E2E forum! Thank you for your business.

You mentioned that you purchased another DMD; were any changes made to the EVM before plugging it into the power supply? Second, there is a small 5A fuse between the power input (J4) and the 19V rail. Is that fuse currently intact? And third, has a short occurred on the DLPC3437 controller between Vin and GND?

The GUI will operate without an EVM plugged into your computer. It won't be able to perform any functions but the program can be opened. Which GUI are you using and which version?

Best,

Maxine

Hello Maxine, 

We did not make any changes to the EVM before plugging it in. We were running an initial test to see if the system powers on before we tried working with it. The 5A fuse appears to be intact. Nothing on the board appears to be broken. I'm not sure how to check if the short is on the DLP3437. Could you help with instructions on how to do this?

The software version for the GUI I am using is the DLP v7.4EVM software downloaded from the TI website. When I open the software, nothing appears to run, not even in the task manager. 

Hello Joshua,

To test for a short on Vin, you can turn your multimeter to the resistance setting. Place one probe on test point #4 (TP4) (or any point on the 19V rail) and the other on a ground point. A reading of 0 ohms would indicate a short. If there is a short on the 19V rail, then some components could also be shorted or damaged, possibly the DLPC3437.

V7.4 is the version of the GUI I have on my end. I would try to re-download the GUI and see if that works.

Best,

Maxine

Hi Maxine, 

On the system board, the TP4 and GND resistance is around 140 Ohms. I can try to use a different multimeter to measure, as I believe there may be problems with the one I'm using. Are there other possible test points that could be the cause of the short? I will redownload the software and try again with the GUI. 

It seems there's a short on the display board from ground to TP54 and TP48. 

TP4 to GND is about 108 Ohm

Hello Joshua,

To narrow down where the issue may be, here are some things you can verify in a step-by-step manner:

1. Power the EVM with 19V from the lab bench supply and measure the voltage from TP4 to ground. Is it 19V?

2. After applying 19V but before turning on the PROJ_ON switch, are LEDs D1 and D7 are illuminated? At which step in the 3310 EVM Guide's 'Quick-Start Procedure' does operation fail?

3. Finally, if the LEDs are illuminated but there is no output after turning on PROJ_ON, then the PMIC should be probed to see where (if any) there is a failure on the start-up procedure. There is a graph on page 18 of the DLPA3000 PMIC datasheet that details how the waveforms appear upon start-up. All waveforms (except for DMD_VRESET) should go from low to high. To check this, probe each net to see if a measurable voltage is there. 

Any luck with the re-download?

Best,

Maxine

I've decided to try a different power supply, and now the system is turning on and appears more stable.

TP4 to ground is 19V. 

D1 and D7 are on now. Everything seems fine until I try to turn on the PROJ_ON switch. 

When I flip the PROJ_ON switch, the output of the power supply drops to 11V and an overload is triggered. The PROJ_ON D5 LED turns on, but the D6 LED begins blinking. I'm not sure how to test each of the waveforms from the DLPA3000 PMIC datasheet. I was trying to find connections between the test points and specific powerup timing points from the waveform graph on the datasheet, but I could not find anything in the documentation. Any help with the process for probing the PMIC would be appreciated. 

Just for some context, for our application, we only need the DMD chip. We don't need any of the LED drivers or anything else on the EVM. We only want to interface with the DMD chip and use our own illumination method to create a structured illumination pattern. I'm not sure if you can tell whether the issue is with the LED driver circuits within the display board or if it's with anything relating to controlling the DMD itself. Is there anything we can disconnect that may be introducing problems into the circuit that we may not need for our application?

Hello Vivek, 

How do I test the timings of the startup sequence for the board? I'm not sure which test pins to probe and when to do them with the oscilloscope in order to perform an exact diagnosis for the issues with our EVM. Thank you. 

Hello Vivek and Maxine,

I've redownloaded the GUI software and it works on my computer. The device is now stable when I turn it on and when I turn the PROJ_ON switch to the on position. The device is not displaying an image. After turning on the PROJ_ON switch, the D1, D6, and D7 LEDs on the display board are on. D1, D5, D7, and D9 are on. Not sure what the issue could be now. Is there anything I can do in the GUI that could help diagnose the issue?

Hello Joshua,

It is good to hear that there is progress on your GUI software. 

Do you see any activity on D8 or D9 on the display board? If not, please probe to ensure that the 2.5V, 1.8V rails from the DLPA are coming up.

The GUI can be used to re-flash the firmware, which may be helpful in this case. Please try connecting to the EVM with PROJ_ON in the off position. In the Firmware tab of the GUI use the 'Browse...' button in the Update Firmware section and point it to the EVM default firmware. Then try updating by clicking 'Start Update.' This firmware image is available on the EVM product page. My hope here is that a re-flash will clear any potential corruption of firmware. Other than this, the GUI may not be of particular use to you as the DLPC needs to initialize to gain communications with the GUI.

Regards,

Austin

Hello Austin, 

D8 and D9 on the display board initially are on after flipping the PROJ_ON switch to the on position, but then after a few seconds they turn off. I have updated the firmware to the one that's provided on the TI website. That didn't fix the issue. The LED's seem to be pulling a large amperage. Around 6000mA according to the GUI IntelliBringht section.

Hi Josh,

D8 and D9 turning off after a few seconds is good news. This indicates that the initialization process is completing. I also had not realized that you were able to establish communication with the controller. The EVM appears to be in better condition than expected.

If you are using the EVM firmware (EVM Firmware: 7.1.2 on the Information tab), the 6000mA current for the green and blue LED's is expected. the red LED should be pulling roughly 4534 mA.

Please try navigating to Status menu in the Advanced GUI. Issue the Read System Status command. Do the return parameters differ from what can be seen in my screen grab?

Additionally, were you able to successfully re-flash the board?

Regards,

Austin

Hi Austin,

I was able to re-flash the board. I downloaded the most up to date firmware and flashed it. After looking in the advanced system settings, the system status is all the same for the board as the picture you sent. There must be something wrong with the LEDs or the illumination circuitry. I can visibly see the DMD inside of the light engine optics moving as I update the image being sent to the EVM. Changing from a solid field pattern to a different one changes the DMD, and I can see it if I look through the lens of the EVM. For some reason, I believe theres an issue with the LEDs. In the Illumination settings, the CAIC RGB LED currents are all at 0, but the RGB LED currents are all around 1000. Not sure if this has anything to do with it. 

Joshua,

So long as CAIC Is disabled the CAIC RGB LEG currents should not be an issue. You can confirm that CAIC is disabled within the Image menu. There is a Read CAIC LED Output Control command that will return 0x0 (Manual RGB LED Current Control).

Can you confirm the power rating on your supply? The GST90A19-P1M should be rated to well above the needed current, but it could help to be sure there was no mix up.

If we are investigating the LED circuitry, what is the resistance between the anode and cathode for each LED with the board disconnected from power? This should be in the order of a few hundred kilo ohms.

Also, what is the voltage difference between the LED anode and cathode? Is it possible for you to connect a current probe on the LED lines to observe the current?

Regards,

Austin

Hello Austin, 

I've measured the resistance across the LEDs. The red and green LEDs are measuring around 110-114kOhm, but the blue LED is significantly lower. It's measuring around 150Ohm. The voltage differences are as follows: Red - 9V, Green - 5.4V, Blue - 20.6V. I'm finding it difficult to connect a current probe. 

I was looking at other posts in the forum, and I believe the issue is similar to this post https://e2e.ti.com/support/dlp-products-group/dlp/f/dlp-products-forum/731513/dlpdlcr3310evm-dlpdlcr3310evm . I'm not entirely sure, but I believe the DLP3000 PMIC and/or the LED Driver might be broken. As I mentioned earlier, the DMD itself is still operational, and I could even display video from the HDMI port. I could see the images forming on the surface of the chip through the lens of the EVM, but the illumination LEDs are not turning on. 

Hi Joshua,

I would like to give you a test image to try on the EVM. Please accept my friend request so that we can coordinate.

Regards,

Austin

Just accepted the request. 

Hi Joshua,

Please coordinate with Austin to continue the discussion, and I will close this forum here. 

Thanks,

Lori