DLPC910: DLPR910 replacement. SPI Flash configuration file?

Hello Finlay,

If you are planning to switch to SPI configuration there are hardware changes that also will need to take place.  Are you aware of these?  We can provide the file.  Please accept the friend invite.

Fizix

Hello Fizix,

We are aware of the hardware changes of dlpc910 part, we acturely design our board depending on the project 00570, the schmatic of dlpc910 and dlpr910 part is exact the same as 00570. The reason we need to design our own board is we need to add our newly developed edge-smoothing algorithm to the board and also add a huge memory sram to it.

Thanks for reply!

Finlay

Hello again Finlay,

Thank you for the confirmation.  You should receive an email several hours from now.

Fizix

Hello Fizix,

Thank you for your help. I haven't receive your email yet, yet all the other emails from TI have been received in my email account. Please check the email address I sent to you by private message. 

Thanks,

Finlay

I received the email, thank you Fizix!

Finlay

You are welcome.

Hello Fizix,

Now we download the firmware to the spi chip xcf16pfsg48c, then the dlpc910 chip can load the configuration and run(can see this breathing LED working). But we monitor the signal SCPENZ, we have a pulse with only 15ns pulse width, and 6.25k freq. The dlp9000x book says it needs at least 1us pulse width to enable the chip. My question is, to have a good SCPENZ signal, is there another version of this firmware can be tried, or we should focus on the hardware trouble shooting?  BTW, we build the board with DLP9000X+DLPC910+XCF16PFSG48C+app FPGA. Thanks!

Finlay

Fizix，

We are trying to verify the SCPENZ signal again, we download the DLPC910A.mcs firmware to the prom chip XCF16PFSG48C and watch the SCPENZ signal, or we download the DLPC910A.bit to DLPC910 chip directly, in both way we couldn’t have a good SCPENZ signal, this is what we have on the scope:

But if we watch a SCPENZ from a working DLPC910 control board, this is what we have:

This two was totally different.

So we doubt it could be a PCB failure or something with the DLPC910 chip, but looks like it’s not. This is what we have if we write a test program to the DLPC910 chip to make it output a pulse to the same pin SCPENZ:

The yellow channel is the test signal on pin SCPENZ, pulse shapes are well formed, so the reason we don’t have a good SCPENZ signal doesn’t look like a pcb failure, but a firmware failure.

So now my question is:

1. Is the firmware DLPC910A.mcs fit for prom chip XCF16PFSG48C? Could it be downloaded to another SPI chip?
2. Could we have the older version of the firmware to try?
3. are there else hardware changes that we need to notice to make the DLPC910 to read ID from DLP9000X?

Thanks!

 Finlay

Hello Finlay,

TI does not make the older FW available as a file.  It was available only in preprogrammed XCF16 DLPR910 chips.

Please try downloading via JTAG the bit file to the DLPC910 directly on the working board and see if it fails SCPENZ or if it works as expected.  Downloading the bit file via JTAG will not cause an issue with what is programmed into the DLPR910 on the board.

Fizix