Re: Tempest vector distortion

I noticed someone had beaten the high score I had quickly put on the machine
a few days ago, so I played 4 games and got 208K or so points. Not great, but
I can't spend too much time playing it, because I have work to do. :-)

I also used the opportunity to notice when the display was acting up. It seems
to have problems with the levels that are not closed loops that are centered on
the screen. The staircase, bowling alley, double U shape, and other similar
levels seem to have a similar foldover issue like the TEMPEST logo, and it
only starts once monsters are put on the play field. For the brief moment when
the playfield is drawn and before it puts monsters on it, the display is fine.

I'm thinking I should just change all 4 chassis deflection transistors. I have
spares at home.

I'll adjust the brightness too. Hopefully that will fix all the problems I'm
seeing.

I'll let you know in a few days once I've worked on it.

___
Ken

On 08/09/11 16:21, Andrew Welburn wrote:
> If you turn up the brightness in the attract mode on the early revision you can see the beam is sent way off the edges of the screen. Yes, they shipped it like this. On a brand new monitor this was of little concern back in the day.
>
> Fast forward 30 years, and This either fritzes out the monitor completely or causes foldover and other weirdness at several points including inter level zooms and the firework display at the endgame. The deflection transistors can cope with normal gameplay, but this over deflection causes partial saturation of the transistor junction, and you see the weirdness with folding or curling. Once the game returns to normal deflection size, the transistors behaves normally.
>
> Either you change the game code and cross your fingers the monitor holds out a bit longer, or you find the chassis mounted deflection transistor that is breaking down under load and put in another.
>
> I have seen this folding/curling on 3or4 tempests in the past. Some people don't notice it, think it's normal, or just live with it. Others like me want to find out why and fix it :) it is common, and it's a problem that only shows up on marginal monitors, it's easily fixed, but you need to swap out transistors one at a time to locate the duffer. Because gameplay isn't affected, some people leave it be, or can't understand why it happens. If you study tempest on a scope display, you can see where the beam is allowed to wander off to a maximum deflection position rather than be brought back to rest at zero. The updated rom code addresses this lazy coding by adding return to zero command every now and then on the fireworks, title screen and inter level zooms... At least I think it does, I rarely find the updated roms on tempest pcbs, it's super common to find the over deflecting romset all the time. I don't remember which revision is which, suffice to say early revs over defl
ect, laters don't... There is paperwork from atari that describes what the new revision roms do, but I think the rom rev was only issued as an optional, not a compulsory upgrade, I think they mention something about 'extending the life of your monitor' with it.
>
> Andrew Welburn (mobile)
> www.andys-arcade.com
>
> On 9 Aug 2011, at 23:53, Ken Sumrall<k_lists@scrapheap.net> wrote:
>
>> Really? They shipped the game with that error in the attract sequence? Or
>> does it only happen on monitors that are marginal?
>>
>> I did notice what I thought was an odd EPROM configuration. There are only
>> 3 EPROMs on the board. 2 are over by themselves, away from the CPU. Then
>> there is an EPROM in a socket near the CPU with some wires running from it
>> to various points on the board, and written on the label it says something
>> like "Tempest 256K ROM". Is this a known hack to put most of the game in
>> a single ROM? Should I just burn a new set of ROMs of the final version of
>> the code?
>>
>> I'll get video of the curved vector lines and also of the EPROMs on
>> the board and post them. Probably sometime tomorrow.
>>
>> ___
>> Ken
>>
>>
>> On 08/09/11 15:30, Malcolm Mackenzie wrote:
>>> The foldover at the end of the attract sequence is a code error in early
>>> revisions of the game.
>>>
>>> The faint lines you're seeing are indeed retrace lines. Adjusting the brightness
>>> should fix that. If not, you may have a problem in the z amp section of the
>>> monitor or board.
>>>
>>> The curving lines are a bit of a mystery...
>>>
>>> --- On *Tue, 8/9/11, Ken Sumrall /<k_lists@scrapheap.net>/* wrote:
>>>
>>>
>>> From: Ken Sumrall<k_lists@scrapheap.net>
>>> Subject: VECTOR: Tempest vector distortion
>>> To: "vectorlist@vectorlist.org"<vectorlist@vectorlist.org>
>>> Date: Tuesday, August 9, 2011, 3:14 PM
>>>
>>> Hello,
>>>
>>> We have a small arcade at work, and I recently got the cabaret Tempest machine
>>> there to work again. Basically I just pulled the logic board and reseated all
>>> the socketed chips and the interboard connector cable.
>>>
>>> It now comes up, stays up, and shaking the cabinet doesn't cause the game to
>>> crash. However, I'm seeing some vector distortion. When it displays the
>>> TEMPEST logo in the attract mode mode, near the end of the animation, the
>>> vectors foldover, and then (mostly) return to normal. You can see a video
>>> of it here:
>>>
>>> http://www.scrapheap.net/tempest/
>>>
>>> Also, on some levels (in particular, the one I call the staircase level)
>>> straight lines get drawn with a definite curve in them, and the screen tends
>>> to pulsate a bit with the size of the image growing larger and smaller. My
>>> first reaction is I need to work on the monitor, but then I thought maybe I
>>> need to adjust the X and Y size and position pots on the logic board.
>>>
>>> I also see some faint lines which I assume are the beam location being moved
>>> with the beam turned off. I assume I fix this by adjusting the screen
>>> brightness.
>>>
>>> Suggestions anyone?
>>>
>>> Thanks in advance for your help!
>>>
>>> ___
>>> Ken
>>> ---------------------------------------------------------------------------
>>> ** Unsubscribe, subscribe, or view the archives at http://www.vectorlist.org
>>> ** Please direct other questions, comments, or problems to chris@westnet.com
>>> </mc/compose?to=chris@westnet.com>
>>>
>> ---------------------------------------------------------------------------
>> ** Unsubscribe, subscribe, or view the archives at http://www.vectorlist.org
>> ** Please direct other questions, comments, or problems to chris@westnet.com
> ---------------------------------------------------------------------------
> ** Unsubscribe, subscribe, or view the archives at http://www.vectorlist.org
> ** Please direct other questions, comments, or problems to chris@westnet.com
>
>
---------------------------------------------------------------------------
** Unsubscribe, subscribe, or view the archives at http://www.vectorlist.org
** Please direct other questions, comments, or problems to chris@westnet.com
Received on Tue Aug 9 20:46:38 2011