[GSoC Weekly Overview]: Good Momentum!


So, it was quite an interesting week on project (actually only 5 days since my exams ended on 28th). Every other day, I was working on almost different area. First day on VHDL, then on PCB soldering, then PCB designing, then some embedded (I2C) and so on, you get the drift!

First, I got to assemble a semi-assembled PCB sent by Jahanzeb. During assembly, I found out one more problem. The R7 (ie Red Color's 8th Bit) signal was unconnected to the P1 resistor network. It was not a major issue hence can be ignored for a while, since only 1 bit information would be missing, resulting only in lesser Red color’s intensity. Finished the assembly.

Then moved on to testing the I3C2 controller. An I2C controller required for initialization and configuration of AD9984. It failed to work. Then, I tried the I2C controller from here. Fortunately, it worked and successfully sent data to 2 PCF8574A's (I2C I/O Expander & LED Driver IC)  connected to the I2C bus. But, I'm inclined to using I3C2 due to its excellent implementation and easy programmability. Also, I've a backup option now, just in case!

Incorrect signal wiring!

Okay, now it was time to test the PCB for life. On powering up and running I2C scan, the AD9984A failed to show up, whereas the other 2 PCF8574A's were showing up! Something was wrong! I checked the voltages, and whoa! PD and VD voltages were *very* low (0.58V and 0.39V respectively) instead of both 1.8V. Weird! I checked the resistors on the feedback circuit to see if they were of correct value. They were. Some how, my eyes got attention to the feedback circuit, and I realised they were incorrect!

Screenshot - 31_05_2014 , 2_40_21 AM.pngerror1.png
On the bottom is the wrong circuit, and on the top is the correct circuit. One can see, It was slight but major error! Fixed the schematic, but the PCB already fabricated had to be fixed with some effort.
In the meantime, I ordered samples for different other fixed voltage (1.8V) LDOs both pin compatible and less pin count(SOT23) ones from Analog Devices and TI.

PCB Correction/Fixing:

Now, the PCB fixing part, It required some trace cutting and rewiring and also resoldering of some resistors. My blog post tells exact things fixed on the PCB. Here are the pictures of fixed PCB:

VD Power Rail Fix
PD Power Rail Fix

Now, I after correcting the PCB, I powered it up and measured the voltages They were fine! Hurray!

I2C Scan Test:

Now, moving on to I2C Test. Let me first tell me about my I2C test setup. I’m using BeagleBone Black for I2C scanning. One could easily use other hardware tools such as BusPirate, RasPi or simply a programmed microcontroller. But, since I had BeagleBone Black available with me, I decided to use it.  I’ll be writing a separate blog post on I2C section!

So, I ran the scan, and voila!, the AD9984A shows up in the scan along with 2 other PCF8574A ICs connected on the same bus, highlighted red in the picture below:

Screenshot - 01_06_2014 , 2_55_16 AM_edited.png

This I2C test setup is only for testing, once I get initialization commands right, I’ll use the I2C core on FPGA for this. For testing, the current setup is far more convenient than compiling VHDL firmware everytime your I2C program changes!

A major task done, I then proceeded on the PCB design. I corrected the VHDCI connector footprint, since the earlier footprint was not letting the plastic retainers go into the hole due to smaller diameter. I also updated the AD9984A IC’s footprint with larger pads. They will help me in soldering. Current pads are too small. Maybe I can manage with them?

After I2C running, major task is to configure the AD9984A hence I read the configuration information from its datasheet. It will more time to complete. Also, I did the footprint assignment and PCB components placement for the next design iteration of the PCB. Some improvements in the next design compared to current design would be updated VHDCI footprint, comfortable pads, test points and maybe smaller power supply section.  

Upcoming Week:


Priority would be to complete the PCB beta design and also to learn to configure
the AD9984A and also test its configuration. First, of all, configuring the AD9984A for getting correct pixel clock from the IC would be the main priority. It could take a week to get this correct. Once, I get the pixel clock, hopefully by next week (10th June), then the actual VGA capture would become a trivial task. Hopefully within 3-4 days of getting the correct pixel clock (~14th-15th June).

Thats all folks! :)

Posted in , . Bookmark the permalink. RSS feed for this post.

Leave a Reply

Search

Swedish Greys - a WordPress theme from Nordic Themepark. Converted by LiteThemes.com.