Dragon's Lair Fans - Arcade Lifestyle
General Chat => Technical Area => Topic started by: nilfisk on October 07, 2013, 08:45:31 AM
-
Dear Technical Experts,
I have started to fix my Asteroids Cocktail that I got for cheap a few months ago. Turns out it is very clean on the inside and almost unhacked, except for the 5v and ground lines being directly soldered to the PCB. The game itself had a few dead RAMs. I've replaced these with spare 2114s I had at home, and now can play a game blind (coins up, start LEDs flash, game sounds can be heard).
Only problem: I do not get any video, neither in test mode nor during normal operation. The monitor seems to be original and unhacked (Electrohome G05 labeled as "Atari Quadra Scan"), the tube neck is glowing, but the spot killer LED is lit.
I've started measuring X,Y and Z out on the PCB, and it seems Z Out is dead (no voltage measured), but I want to go sure reg. the expected voltage ranges: the manual says that anything between -8/+8 VDC is correct for X/Y, and -5/+5V for Z; the vector monitor FAQ and some posts on KLOV suggest that AC is provided to the monitor instead.
Can someone provide me with some guidance (@Speleo? I've found a lot of info on your website - thanks btw.!).
Cheers,
Martin
-
Here is the Z Axis output.
As you can see the Z level is made by the resistor array, so check the signal at the base of Q7.
If is ok check the 74LS164 (K9) and the 74LS04 (H10) otherwise you must see the scale signals but I don't think will be necessary if game play blind.
If one of them keep Q8 on, the output of Q9 will be allways zero volts.
Remeber that as you can see the Z output is only 0..5V never negative values!
Obviously if the two 74xx are good check the transistors.
-
Technically speaking X, Y are AC voltages because they move below 0 and above 0 and that in an "alternating" fashion. However, this is no regular AC like a sine wave that is coming out of your mains wall outlets.
I would guess that the Z level is a lot lower than the X and Y signals because X and Y needs to move the beam, which requires quite a bit of energy, and Z only represents the brightness and my guess is that this would require less energy to "steer".
I could be wrong though, there is of course also amplification going on on the monitor PCBs of those signals.
The manual should tell you a lot, as always, Arcade Archive is your source for all those manuals and I think Speleo's site has them too.
Of course, an oscilloscope is ideal to check vector outputs :)
-
Hmmh, Z-Axis rarely has an issue... when the spot killer LED is lit, at least one of X or Y signals is missing (or the spotkiller circuit is dead).
X/Y is sort of an AC. They are usually stuck at a high DC level, when defective.
If both X and Y are 'dead', check the Voltages in the D/A section (+5V, +/-15V, +/-8.2V).
In the D/A circuitry mostly the 4016's go defective, next the TL08x, next D/A Converters (rarely).
Of course there can be other parts defective before the D/A, that happens very rarely.
-
Hello,
Thanks for your inputs - highly appreciated, as always!
or the spotkiller circuit is dead
Yep, I am planning to take the monitor out this evening to do a proper check up. Is there a "usual suspect" in terms of the spotkiller circuit failing? Can you guide me to some suitable testpoints here as well?
Sorry for flooding you with questions, but besides repairing my Vectrex a few years ago, this is my first "vector experience" :D
Cheers,
Martin
-
Do you hear any vector shatter going on ?
-
No, nothing, total silence.
If I turn the monitors brightness all the way up, I can see a bright center spot on the screen, but obiously, I want to keep my phosphor coating, so I'm not planning on doing that again.
-
Hello,
Thanks for your inputs - highly appreciated, as always!
or the spotkiller circuit is dead
Yep, I am planning to take the monitor out this evening to do a proper check up. Is there a "usual suspect" in terms of the spotkiller circuit failing? Can you guide me to some suitable testpoints here as well?
Sorry for flooding you with questions, but besides repairing my Vectrex a few years ago, this is my first "vector experience" :D
Cheers,
Martin
I have given you checks for the game pcb - start from there...
-
I completely agree. Better to test the board with an oscilloscope in XY mode. Don't care about Z at moment.
I destroyed a tube using my vector monitor to repair my Hyperspace board >:(
-
Thanks guys, will do as told - no need to doubt the experts :P
-
UPDATE: Done. Checked out all the voltages as shown in the Power Input section of the manual. +/-15 VDC, +/- 8.2 VDC and +5 VDAC checked out good, so I grabbed my scope (a cheap, portable Velemann, but enough for my range of skills) and poked at the outputs. With -4/+4 VDIF, I was able to pick up a sane looking signal at X-Out, but just low voltage noise on Y-OUT.
Going back and comparing/tracing back, I suddenly noticed X being dead as well. Rechecked all the voltages as initially, and sure enough +8.2 VDC is now down to around 1.5 VDC; all other voltages still check out ok, incl. the -8.2.
So I suppose I have to fix the DAC power section now before moving on.
Again, thanks for your prev. inputs, guys.
Martin
-
Got the 8.2v back. I originally suspected the zener diode at CR14, but since replacing it didn't change a thing, I went further down the path and found the 14016 at D12 to be bad - removing it brought back things to normal. I'll get a replacement on monday, but as I have to move abroad for a few days (work related), I won't find any time to continue my investigations before sunday next week.
Martin
-
Hmmh, Z-Axis rarely has an issue... when the spot killer LED is lit, at least one of X or Y signals is missing (or the spotkiller circuit is dead).
X/Y is sort of an AC. They are usually stuck at a high DC level, when defective.
If both X and Y are 'dead', check the Voltages in the D/A section (+5V, +/-15V, +/-8.2V).
In the D/A circuitry mostly the 4016's go defective, next the TL08x, next D/A Converters (rarely).
Of course there can be other parts defective before the D/A, that happens very rarely.
:-)
-
Hello Guys,
Finally found the time to follow up on this. I do now have all voltages back to normal (indeed, a bad 4016 and a fried zener diode). Also, I can see what I consider a normal vector signal on X-out with my scope (8v/20 ms Div, looks like a proper pattern for the test screen). Also, on Y-out I believe to see something that makes sense, albeit at a much weaker signal level (0,4 to 1 VDiv). I can confirm that the opamps for Y-out are good (socketed and replaced with a known good TL081 for B/C12 and a TL082 for A12).
It seems that the output coming from the DAC is much lower for Y already, at least when comparing the two DACs. I can confirm that proper voltages are supplied to both of them again (5/-15) according to the manual. Any suggestions for next steps? As the DACs are somewhat expensive, i don't want to shotgun this one by just replacing it.
As always, thanks for any hints!
Martin
PS.: also followed up on Z-out being dead. I am quite sure it's a fried 164 (all three transistors and the 7404 are confirmed to be ok, and the 164's output seems to be stuck), but I need to get one tomorrow in order to confirm this.
-
Do you have capabilities to desolder the DACs and swap x vs y?
-
Hi,
I have thought of that, but have as of yet refrained from doing so, out of fear of damaging one of them. I've tried taking a look at the input side of the DACs (pin 1-10, iirc) beforehand, but haven't really been able to tell right from wrong here. I guess I'll just buy in on a replacement if there's no other option.
Cheers,
Martin
-
After some weeks, another (rather disappointing) update:
- Finally got my hands on a spare DAC and socketed/replaced the one for Y-Out: no changes.
- Also replaced the LS164 for Z-Out, also without success. Still dead.
So, I now do have:
- Proper voltages everywhere on the board (checked and re-checked)
- X-Out: working as expected.
- Y-Out: known good TL081, TL082, DAC - seems to be somewhat active, but far too weak (already coming from the DAC, pre-amplified)
- Z-Out: known good transistors (all three of them), LS04, LS164 - dead Z-Out (I do see activity at the base of Q7, but beyond that point, the signal seems to get lost).
- Game still does play blind.
Learnings so far:
- I need to get a proper scope. The old Velleman portable is suffering from software issues and frequently locks up during normal operation.
- My lack of knowledge hits me hard here, so I'd appreciate any further hints.
Cheers,
Martin
-
Finally, with a lot of help, and a second, working boardset as comparison, Asteroids is up and running again. Turns out not only the PCB was severly shot, but also the monitor needed some work (reflowing the Molex header joints, bridging one of the two ceramic resistors, so fortunately a rather easy fix).
(http://img10.imageshack.us/img10/6977/n1rm.jpg) (http://imageshack.us/photo/my-images/10/n1rm.jpg/)
Still a lot of stuff to cover (replace "big blue", rebuild AR, ...), but at last some visible progress :)
-
Replacing Big Blue is almost never neccesary.
If you plan on getting more cabs, buy an ESR meter and just check those caps instead of pointlessly replacing them with possibly inferior modern types...
Nice work by the way, would be interesting to read the PCB fixes you did.... ;D
-
Woodgrain looks totally cool on this table, well done Martin! ;)