DLPC910: differences between Left and Right side of DMD

Roy Netanel

Prodigy 40 points

Part Number: DLPC910
Other Parts Discussed in Thread: DLP9000X,

Hi,

We are working new board controlling on DLPC910 and DLP9000X.

We are receiving DDC Version of "2" (OK) and DMD Type of "F" (also OK).

1. Left side mirrors of DMD looks different from Right side

2. we send training pattern and INIT_ACTIVE goes low.

3. we send first frame (1600 rows) followed by MCP --> RST_ACTIVE goes high and stays high --> In the right half of the DMD mirrors changed as expected, In the left side mirrors changed but to something unknown (seems different in the top and bottom side)

4. we tried to send PWR_FLOAT and still Right side of the mirrors and the left side looks different.

we tested DLP900X voltages, DCLKOUT_A - DCLKOUT_D (400MHZ), SCTRL_A - SCTRL_D and they seems OK

We are trying to understand if we are dealing with logic design issues or HW issues, from your experience what could be the problem?

over 3 years ago

0 Roy Netanel over 3 years ago

Prodigy 40 points

IMPORTANT UPDATE:

3. RST_ACTIVE goes "high" and then "low" (in all frames writes) but the mirrors are changed only on the right side of the DMD

0 Fizix over 3 years ago in reply to Roy Netanel

TI__Guru 54705 points

Roy,

Please check all of the rail supply inputs to the DMD - VBIAS, VOFFSET, and VRESET as a starting point.. Make sure that they are present on all specified input pins for each voltage.

As a second step if the first does not show anything, it may be necessary to inspect the interposer for possible missing (or rotated) "C" springs and remount.

Fizix.

0 Roy Netanel over 3 years ago

Prodigy 40 points

this is what we see if we write all ones

0 Fizix over 3 years ago in reply to Roy Netanel

TI__Guru 54705 points

Hello Roy,

Thank you for the image. I need a little more time to study this.

Fizix

+1 Fizix over 3 years ago in reply to Fizix

TI__Guru 54705 points

Roy,

Offline we determined that the P/N of clock C was reversed so that SCTL_C was being read incorrectly and data was not being properly written to the right side of the DMD (left side in the picture).

I am marking this as complete.

Fizix

0 Roy Netanel over 3 years ago in reply to Fizix

Prodigy 40 points

As you wrote, after double check we saw that clock C P/N was reversed.

After we fixed clock C P/N left side of DMD started to work.

0 Fizix over 3 years ago in reply to Roy Netanel

TI__Guru 54705 points

Roy,

Fantastic! Thank you for letting us know.

Fizix

DLP®︎ products

DLP products forum

DLPC910: differences between Left and Right side of DMD