Re: Asteroids proper grounding and monitor DC reg bug.

From: <js_at_cimmeri.com>
Date: Mon Sep 15 2008 - 12:52:11 EDT

William Boucher wrote:

> It sounds like you really did your homework there and I think it makes
> great sense with respect to the monitor and the video lines. Granted,
> I've not taken the time yet to absorb all of that yet like I plan to,
> but my first thought is with regards to the AVR and the Vsense leads
> and how your mods might effect the functionality of those. You said
> that you did not add another ground to the AVR. You could if you
> wanted to as long as it remains on the input side and not on the
> output side. There are 2 leads each for +5V and ground return running
> from the AVR to the game board. They are NOT shorted together but are
> separated by a 10 ohm resistor on each side. The idea with that is to
> provide the bulk power to the game board through the fat leads so that
> the AVR can read the board voltage via the thinner leads and then
> compensate for Vdrop over the fat leads. When the connection between
> the fat leads and the game board is degraded or burnt out, all the
> game board current tries to go through the little 10-ohm sense
> resistors and burns them up. Replacing the gameboard power leads with
> fatter wire wouldn't help. Improving their connection to the
> gameboard itself would. If you were to short the circuit ground leads
> together at the AVR or bypass them with a separate ground to the game
> board, the Vsense function on the negative side of the supply would be
> defeated and the AVR would not be able to regulate, or even properly
> measure, that side.
>
> You said that prior to the mods that you could read tenths of a volt
> between the gameboard ground and the power brick ground... You're
> supposed to. That voltage drop occurs over the wires running from the
> brick to the AVR and then more again over the fat leads from the AVR
> to the gameboard. The Vsense leads between the AVR and the gameboard
> are then used to compensate the supply so that the gameboard chips see
> an accurate +5V supply relative to themselves locally. So long as the
> ground mods that you made do not interfere with the AVR's ability to
> sense voltage drop over the leads connecting its output to the
> gameboard, then it should continue to regulate properly.
>
> William Boucher

<< my first thought is with regards to the AVR and the Vsense leads and
how your mods might effect the functionality of those. >>
*
Normally, the PCB's floating ground is tied to earth ground through the
3 video gnd wires, then through the deflection board, then to the
monitor chassis (which itself is normally tied to the power brick's
earth through the DC regulator).

By removing the video wiring-earth ground link, and providing a more
direct one directly from the power brick, I'm not changing the operation
of Vsense. Vsense is concerned with the relative difference between 5v
and PCB ground... not the PCB gnd and earth ground. Accordingly, the
mod didn't even necessitate an AR boardd adjustment.
*

<<You said that you did not add another ground to the AVR. You could if
you wanted to as long as it remains on the input side and not on the
output side. There are 2 leads each for +5V and ground return running
from the AVR to the game board. They are NOT shorted together but are
separated by a 10 ohm resistor on each side. >>
*I might have a different AR board than you... as my board has a big
ground plane running in a complete circle along its backside
circumference (a ground loop I need to open, I now see). The 2 floating
ground wires from the brick tie into this circle, as do the two main
grounds going to the PCB. In this case, then, the AR board is just a
splice for those 4 floating ground wires; and essentially a big,
adaptive resistor for the positive volts wires between the brick and
PCB. No need for or benefit to be had from an earth ground reference
or shield.

*<< The idea with that is to provide the bulk power to the game board
through the fat leads so that the AVR can read the board voltage via the
thinner leads and then compensate for Vdrop over the fat leads. When
the connection between the fat leads and the game board is degraded or
burnt out, all the game board current tries to go through the little
10-ohm sense resistors and burns them up. Replacing the gameboard power
leads with fatter wire wouldn't help. Improving their connection to the
gameboard itself would. >>
*Well, I'd never just replace the wire without making sure their
connection to whatever was optimal. ;-)

*
 
<<You said that prior to the mods that you could read tenths of a volt
between the gameboard ground and the power brick ground... You're
supposed to. That voltage drop occurs over the wires running from the
brick to the AVR and then more again over the fat leads from the AVR to
the gameboard.>>
*I was seeing both VDC and VAC. I didn't want to see any VAC at all,
and yes, there's always VDC drop over wiring, but the question is of how
much is tolerable. Since the PCB is normally tied to power brick earth
ground via its video cable-def board-monitor frame link, ideally, the
PCB and the power brick should have as close to zero VAC and VDC
differential as possible. The real aim here is that both the PCB
ground near the video outs _and_ the deflection board ground near the
video input connector be as close to each other as possible for the most
accurate video signaling. By improving that ground reference, I did
perceive a very slight improvement in video quality.

 JS
*

---------------------------------------------------------------------------
** Unsubscribe, subscribe, or view the archives at http://www.vectorlist.org
** Please direct other questions, comments, or problems to chris@westnet.com
Received on Mon Sep 15 12:52:13 2008

This archive was generated by hypermail 2.1.8 : Mon Sep 15 2008 - 13:50:01 EDT