Newhaven Products > Character LCDs

I2C ACK issue

(1/2) > >>

jjbrewer:
It was reported on the other forum that the I2C controller bounded to the NHD-C0216CiZ-FSW-FBW-3V3 COG character LCD could only sink 4ma of current.
In order to get a decent waveform with the cable length I am using I have 2.5K pulllups on my PCBA. The COG cannot pull SDA down hard enough for a valid ACK.
Would a NXP PCA9605 2 wire buffer, placed  between the COG and the other devices and cable, help?

Michael_L:
The NXP PCA9605 could help, but I would like to ask if you could try isolating the display on the i2c bus, and see if the ACK gets pulled low enough when there are no other devices present on the bus.

jjbrewer:
The issue is I need it to work with these other devices at the end of a 36" ribbon cable at 100kHz.
I have attached a VISIO drawing and pictures from a O-scope of the wave forms. The 1.5K pictures are with the 1.5K pullups and the 20K are with 20K pullups.
The EPROM and PsOC have no issue pulling down the SDA with the 1.5K pullups.
I picked this COG because of its size. It was the only one I could find that will fit in the space I have.
My next experiment will be placing the NXP PCA9605 between the I2C bus and the COG.

The attachments button only allows 4 and won't send BMPs, so I zipped it up and changed the extension to pdf, rename to .zip

Michael_L:
Thank you for the images.  The recommended pull-up value is 10kΩ for each i2c line.  From the look of your pictures, 20kΩ seems to do the trick for you correct?
I am not familiar with the NXP PCA9605, but it seems to me that if you use larger pull-ups the display will work great. :)

jjbrewer:
I was able to insert the NXP PCA9605 between the COG LCD and the other devices and the ribbon cable.
I was able to use 10kOHM for the side the COG LCD is on and iw as abke to ACK quite well.
On the other side, the buffer was able pull it low even with the 1.5kOHM pullups I started that the PsOC and EPROM were okay with.

Navigation

[0] Message Index

[#] Next page

Go to full version