[back!] Tempest - Plays blind but tube has glow
- Thread starter Jodo
- Start date
So if the tube is getting the correct voltages and the PCB has tested good with Retroclinic / plays blind and the AR2 has correct +5v now, it looks like a monitor issue after all...
Btw, are the power bricks supposed to be this noisy? This one is positively buzzing...
Btw, are the power bricks supposed to be this noisy? This one is positively buzzing...
One source of buzzing on a wire-wound TX is magnetostriction http://en.wikipedia.org/wiki/Transformer amongst the metal layers. Might be worth having a look for four bolts that run through each corner of the metal laminates and see if they can be tightened as I suspect that increased compression over the laminates will diminish the intensity of the buzz.
There's no two ways about this, a faulty monitor cannot cause a PCB to start resetting like a good-un. It just can't. Even with a very, very slim possibility of a tiny, bit of feedback on the X / Y / Z outputs of the PCB, that's still only the analogue output section of the PCB (DACs etc.). It could impact the PSU and in turn the AR2 though.
My money would still be on power, or PCB. Somethings failing when it's getting hot and overloaded. When Mark tested it, was he running it on your Atari power brick and AR2? I guess not, or even on ANY Atari power brick and AR2. Again, knowing Mark (and I'm not at all saying this in a bad way!), I doubt it - I suspect he's built his own supply voltages for his rig.
I don't know where to tell you to go from here, but it would be interesting to rig a meter up to the 5v from the AR2 and see what it's at to start with, keeping monitoring it and seeing what it's up to at the point the PCB starts resetting.
I know you don't have one of these but at this point I'd be looking to stick a couple of the PCB voltages on my new wazzy scope that can give me max VPP and min VPP and see if the voltage(s) drop too low at any point. But I appreciate that you can't really do that without the kit to do so
My money would still be on power, or PCB. Somethings failing when it's getting hot and overloaded. When Mark tested it, was he running it on your Atari power brick and AR2? I guess not, or even on ANY Atari power brick and AR2. Again, knowing Mark (and I'm not at all saying this in a bad way!), I doubt it - I suspect he's built his own supply voltages for his rig.
I don't know where to tell you to go from here, but it would be interesting to rig a meter up to the 5v from the AR2 and see what it's at to start with, keeping monitoring it and seeing what it's up to at the point the PCB starts resetting.
I know you don't have one of these but at this point I'd be looking to stick a couple of the PCB voltages on my new wazzy scope that can give me max VPP and min VPP and see if the voltage(s) drop too low at any point. But I appreciate that you can't really do that without the kit to do so
guddler said:
Hey fella
Continuous monitoring on the +5v on the (replaced) AR2 gives a constant 5.02v now, with no resetting at all. This has been on for well over an hour now. The previous AR2 started increasing on the +5v after 20 minutes of load so I pulled it; before I pulled it out however I made 1 attempt to adjust the +5v on it and with a 1 mm fractional turn it jumped to +7v!!! (at this point I pulled the plug) The +5 pot on the AR2 was not good; maybe it was this that was expanding under heat after 20 minutes of use and chucking out 7v+?
(@ Tim) - Thanks buddy - I'll take a look at tightening those if present on the TX..
Ahhhhh, well...
Now it sounds like you're at the crux of the problem
What's going to happen on a Tempest if the 5v starts creeping up, at the very least is that the bit slice processors will overheat and become very unhappy. They cause the game to spew vectors all over the place but I think with extreme errors they can crash the aux PCB and in turn the main.
That is from memory though and may not be entirely correct.
Sounds very much like you're getting to the root of it though
Now it sounds like you're at the crux of the problem
That is from memory though and may not be entirely correct.
Sounds very much like you're getting to the root of it though
Jodo said:
aha... well I had one do that once but I swapped out the AR2 so I never actually checked what the fault was but probably bad contact on the conenctors; the game pcb survived though I don't suppose it did it that much good. Anyhow, that is one reason for testing PSUs with dummy loads rather than your game board. I don't know what other people use but I have used things ranging from big power resistors or incandescent light bulbs to a street fighter bootleg.
Unless you have a massive stockpile of monitors you should still consider buying an old crappy scope at some point as long as it does X/Y it should display something that gives you an idea of what is on the display even if it looks a bit mashed. I'm thinking you should be able to get something for around £40 that would do the job, maybe less and weighing that up against the cost of replacing an xy monitor it seems a good deal.
Of course whatever it was appears to have left me with a duff monitor; I'll see what the rebuild kit does for that.
Slightly off topic then but I'd really appreciate a couple of scope make / models to look for and I'm interested in how I'd actually get the signal from the game PCB to show on the scope itself? What sort of connections would I need?
*EDIT* There is actually a shed load of 99p start 'scopes on Feebay right now if you think any of them are worth a punt?
I understand they use BNC connections so I'd need to get a couple of BNC adapters but am I capturing from the X / Y loops on the game PCB or is it not that simple
Jodo2012-09-21 11:34:31
Slightly off topic then but I'd really appreciate a couple of scope make / models to look for and I'm interested in how I'd actually get the signal from the game PCB to show on the scope itself? What sort of connections would I need?
*EDIT* There is actually a shed load of 99p start 'scopes on Feebay right now if you think any of them are worth a punt?
I understand they use BNC connections so I'd need to get a couple of BNC adapters but am I capturing from the X / Y loops on the game PCB or is it not that simple
Jodo2012-09-21 11:34:31
Jodo said:Of course whatever it was appears to have left me with a duff monitor; I'll see what the rebuild kit does for that.
Slightly off topic then but I'd really appreciate a couple of scope make / models to look for and I'm interested in how I'd actually get the signal from the game PCB to show on the scope itself? What sort of connections would I need?
*EDIT* There is actually a shed load of 99p start 'scopes on Feebay right now if you think any of them are worth a punt?
I understand they use BNC connections so I'd need to get a couple of BNC adapters but am I capturing from the X / Y loops on the game PCB or is it not that simple
Watch out dude, sounds like you're turning into a nerd
That would do the job, yes. (BNC cables that is!)
For just testing the presence of an XY picture then you want BNC on one end, either crocodile clips or test hooks on the other and connect them to the X and Y test points. CPC sold the cables years ago when I got mine but I'm sure many places sell them cheap enough. For any other use you need a x1 scope probe. Most instructions say youse a x10, x1 switchable one set to x10 but I've always found that n this old sh*t were working on that just amplifies the amunt of noise you see soi stick to x1. Works well enough for me.
I've actually got a scope I was going to be selling now I've upgraded to a modern Rigol digital storage scope but I was going to ask £50 for it as its compact compared to a lot of them out there (a relative term when it comes to CRT scopes, NONE of them are remotely small). It actually can be powered from a battery and back in its day was considered portable field equipment. Damned if I'd have wanted to carry it about though and who knows what the battery must weigh!!
Need to offer it to Bazfish first though as he was after one.
For just testing the presence of an XY picture then you want BNC on one end, either crocodile clips or test hooks on the other and connect them to the X and Y test points. CPC sold the cables years ago when I got mine but I'm sure many places sell them cheap enough. For any other use you need a x1 scope probe. Most instructions say youse a x10, x1 switchable one set to x10 but I've always found that n this old sh*t were working on that just amplifies the amunt of noise you see soi stick to x1. Works well enough for me.
I've actually got a scope I was going to be selling now I've upgraded to a modern Rigol digital storage scope but I was going to ask £50 for it as its compact compared to a lot of them out there (a relative term when it comes to CRT scopes, NONE of them are remotely small). It actually can be powered from a battery and back in its day was considered portable field equipment. Damned if I'd have wanted to carry it about though and who knows what the battery must weigh!!
Need to offer it to Bazfish first though as he was after one.
guddler said:
If Baz doesn't want it and it and you're OK to offer a little guidance on getting it displaying Tempest with me, then I'm interested please
Alpha1 said:Watch out dude, sounds like you're turning into a nerd
I love this sh*t
Jodo2012-09-21 12:19:44
I'd say the cheapest option would be an oldish analogue 'scope -must be capable of x/y mode and you'll see all the retrace lines without a Z channel but I wouldn't worry about that. I don't know that you actually need probes but they make life easier, I suspect a piece of wire would suffice possibly adding a resistor in series to be on the safe side. Not sure of minimum scope speed (yes I'm sure you can work it out from the monitor specs but I haven't) I would say mine is 20meg tops. So in essence I can't say exactly what scope would work best 
, I'd just go for something cheap and nasty and dump/sell/give away if it not suitable. Check this out:
Apart from the incorrect suggestion that retrace lines were left in for test purposes it's bang on for a simple demonstration. My scope doesn't get as good a picture as that, eg I can't read the text properly on the self test screens but it's good enough.
, I'd just go for something cheap and nasty and dump/sell/give away if it not suitable. Check this out:Apart from the incorrect suggestion that retrace lines were left in for test purposes it's bang on for a simple demonstration. My scope doesn't get as good a picture as that, eg I can't read the text properly on the self test screens but it's good enough.
Just in addition, if you can get a scope with a Z input it's preferable. If you're only interested in checking for vector output on the X and Y axis then you don't need it. As I'm finding however, if you have an issue with the Z blanking circuit you can only see it on the real monitor. I've got a Hameg 203 and it works pretty well for just the X and Y stuff.
Dan
Dan
Hey all,
I think we were looking, and possibly still are, at multiple faults. To recap, and I apologise if I didn't update the thread at the time with this info:
- The original fault of no picture was solved by moving the key plastic in the edge connector that was blocking pin 4
- The cab then played perfectly for 20 minutes before the game started resetting again. It appeared the AR2 was having a fit when warmed up so that was replaced and fixed the resetting
- Unfortunately although the game played fine, the screen now displays faint coloured clouds that change colour in time to the attract mode. After 30 seconds the tube starts to stink so power must be killed
- (Current status) The entire tube assembly is with Ordy for a rebuild and LV kit. The AR2 is with Mr P-Man for the sense mod
I think we were looking, and possibly still are, at multiple faults. To recap, and I apologise if I didn't update the thread at the time with this info:
- The original fault of no picture was solved by moving the key plastic in the edge connector that was blocking pin 4
- The cab then played perfectly for 20 minutes before the game started resetting again. It appeared the AR2 was having a fit when warmed up so that was replaced and fixed the resetting
- Unfortunately although the game played fine, the screen now displays faint coloured clouds that change colour in time to the attract mode. After 30 seconds the tube starts to stink so power must be killed
- (Current status) The entire tube assembly is with Ordy for a rebuild and LV kit. The AR2 is with Mr P-Man for the sense mod
so, it now sounds like the game is over deflecting on one axis, probably the deflection pcb this time (rare), but there's a chance its the game pcb itself.
if the monitor has been removed and sent away, why dont you measure from gnd to the X and the Y posts on your game pcb while its running, just to make sure its not the game pcb.. let us know what the voltages are when the game is in test mode...
if the monitor has been removed and sent away, why dont you measure from gnd to the X and the Y posts on your game pcb while its running, just to make sure its not the game pcb.. let us know what the voltages are when the game is in test mode...