and the VGA saga continues

GoogleGroups
1

It is mostly as this VGA thing dont want to cooperate, after testing several methods without any effective results (== display converted to VGA output), Im currently testing the VGA adapter made of resistor ladders (and a couple of sync logic inverters using npn transistors) from this page:

http://www.linusakesson.net/hardware/beagleboard/vga.php

Im sure this is very familiar for BB C4 users (from the number of threads and web pages that are related to this vga method), so here is my question, Have anyone tested this on a BB xM?

I have constructed this adapter, and verified again and again that all connections, traces, and components are properly soldered, but the BeagleBoard (xM) simply dont want to boot when this adapter is connected to the LCD expansion port (not even any serial output), after the power is applied, it simply lit some of the leds, but nothing after that

I have purchased the same resitors that are used in that page (after a month and a half of shipping and delivery time) 60.4 omhs and 121 omhs, low tolerance, metal-film type

Is it because the outputs (lcd data bits) are conencted to the ladder that draws 7.5 mA? and the nominal current of the pin is only 0.1mA? but this has not prevented the C4 from working with this adapter

I even have have connected the BB directly to the PC power supply (5v rail) that can provide 8 ~ 10A (more than enough for the BB)

As soon as the adapter is disconnected, the BB boots up

I have reviewed several times, there are no shorts or other issues

Ok, if this method dont work, I suppose we have the active DVI to VGA converter from fit-PC or some Asian DVI-VGA converter (there is a $24 unit that seems to work ok as per the reviews)

Lioric

2

Hi, I would avoid the resistor ladder myself and go active conversion.

I also see these on Amazon. I have not tried them and can not speak for their
ability to perform the task.

http://www.amazon.com/HP-HDMI-Vga-Display-Adapter/dp/B0038KFTZK

or

http://www.amazon.com/gp/product/B003HC85D2/ref=pd_lpo_k2_dp_sr_2?pf_rd_p=486539851&pf_rd_s=lpo-top-stripe-1&pf_rd_t=201&pf_rd_i=B0038KFTZK&pf_rd_m=ATVPDKIKX0DER&pf_rd_r=1BRG76Y2XMDWZRFMMN4C

djlewis

3

http://www.beagleboardtoys.com/shop/article_BB-vgab/Beagle-Vga-board.html?shop_param=cid%3D1%26aid%3DBB-vgab%26

Gerald

4

@Gerald

Im the guy that have tested two of the BeagleBoard VGA adapters with the same “washed out color” display (and a very odd green color issue, almost as if sync is muxed in the green bits or gamma levels are way off), I contemplated the possibility of changing the FSAdj resistor value of the THS3135 silicon used in that adapter, to modify the voltage levels and “change the gamma”, but I tested the adapter and they are alredy using a 37.5KOmh as recommended by the TI datasheet

I think that: either my BB xM have an issue with the LCD expansion connector, or the BBToys VGA board was designed for C4 and not really tested in xM (that is why I asked if there is anybody that have tested with xM)

Probably it would be a good idea to contact them (circutco) and ask them for a test and validation with a xM (the communication with them was really slow so I decided to test other options)

@djlewis

Thanks, yes I have them on my list of things to test for this VGA issue, the problem is that there is no actual real tests or data that will backup that any of them will work, and seriously with what I have spent only getting VGA output (with no real results) I would have purchased a top of the line fit-PC2 with LVDS, VGA, DVI, 1.8GHZ 2GiB ram for my project

What I an postponing is creating an adapter board with the SN75LVDS from TI and connect the panel directly, it might probably work, but will require a custom fabricated board with a pre soldered chip on it and we are talking about 50 to 100 usd, Im doing some test soldering TSSOP 56, but at the rate of the shipping time to this country and the expenses involved, this is a backup plan , where an Asian lcd controller board goes for only 10 usd (the thing is finding the right controller)

And the size of the LCD expansion headers is not helping this custom fabrication either

Lioric

5

The interface on the C4 and the -xM are the same, so it should work on the -xM. There may be some differences that have crept in somewhere, so I need to go back and take a look to see if something got missed.

The VGA board was tested on the -xM but it has been a while back, on the first generation.

Some things that that could an issue include:

-The fact that there are option resistors on the LCD control lines on the -xM to handle the pin swapping required for the difference in the actual output pins used. I don’t think that is the issue but it is one difference between the C4 and the -xM.
-One other possibility is the resolution used. It may have an affect on the colors as things are changing more rapidly on the higher rresolution. One thing you might check is the inductors on the output lines of the VGA pins themselves. You could try shorting them out

-Another possibly issue could be the settings of the timings in SW that affect the porches on the video signal. There may be an issue on the retrace time required by an analog device like a VGA monitor. Those could be adjusted as well.

Gerald

6

Thanks Gerald

-The fact that there are option resistors on the LCD control lines on the -xM to handle the pin swapping required for the difference in the actual output pins used

Yes, that doesn’t seems to be the issue here

-One other possibility is the resolution used

Same result on any resolution, from 648x480 up to 1024x768, exactly same result

One thing you might check is the inductors on the output lines of the VGA pins themselves. You could try shorting them out

Do you mean shorting the R, G and B lines of the VGA connector? Im not sure what that would test (All this is somewhat harder without the actual schematics of the adapter)

-Another possibly issue could be the settings of the timings in SW

Yes, no issue here as well, all the timings are correct and validated in other operating systems (and with other video signal generators)

To me, this seems to be an incorrect FSAdj value for the resitor or a fault in the lcd expansion(signals) header

But being without an oscilloscope at hand is not helping either, so I will put this BB Toys VGA adapter in the storage box for the time being and continue to test other alternatives (or purchase a Panda board or a fit-PC2 and call it a day)

Lioric

7

I have the guys taking a look at it at the moment. I will let you know if we have any results. Right now we are seeing some issues on the -xM. But, they don’t have a root cause yet.

Gerald