I, Robot #158 Restoration

Someone has too much time on their hands. Or is over-thinking things.
I did not realize you run your I,Robot 24/7.
Fascinating. :unsure:

Personally, adding noisy (acoustic/emi) fans in any of my machines is the last thing I'd do.
I'd rather keep the backdoor off the machine if I was that paranoid. YMMV
 
Someone has too much time on their hands. Or is over-thinking things.
I did not realize you run your I,Robot 24/7.
Fascinating. :unsure:

Personally, adding noisy (acoustic/emi) fans in any of my machines is the last thing I'd do.
I'd rather keep the backdoor off the machine if I was that paranoid. YMMV
I played it for several hours last night non stop and the monitor started getting jumpy by the end 😅
 
You could always break down the Disco and mount it more like an Amplifone vector. I might have a spare set of Amplifone tube brackets if you wanted to try it. Of course that would make it less historically correct. You could also add another fan. I do have extra fan mounting brackets. I had a few of them made.
 
You could always break down the Disco and mount it more like an Amplifone vector. I might have a spare set of Amplifone tube brackets if you wanted to try it. Of course that would make it less historically correct. You could also add another fan. I do have extra fan mounting brackets. I had a few of them made.
I thought the same thing, but it then (like you said) it would be the argument of 'all original' vs. 'venting the monitor'.
 
Looking at the wiring diagrams, it shows an AC fan is meant to be in the cabinet so you're idea of needing more cooling tracks.

1744253704370.png
 
Looking at the wiring diagrams, it shows an AC fan is meant to be in the cabinet so you're idea of needing more cooling tracks.

View attachment 812126
There is one, I've got it installed. But since I have the pole position cage and not the more mesh-y cage, I might try to see if I can get some more air going over the PCB specifically.

Won't matter for an hour or so of gameplay, but if I leave it on all day that might get toasty
 
You're overthinking the cooling thing.

As long as there's a 12V fan running (even running at 5V), anywhere near the top of the cab, blowing up and out, it's going to suck enough cool air into the bottom of the cab, to cool everything in the cab.

These games don't actually generate a ton of heat. It might SEEM like they do, if you open one up and feel like the cab is warm inside. But that's because if there's insufficient airflow, the cabinet acts like a huge winter jacket, trapping heat.

However just the slightest amount of moving air is going to pull heat away from everything inside faster than it can be generated. It really doesn't take that much flow. One 3" 12V PC cooling fan is all you need (which you can even run at 5V to be quieter), as long as it's drawing the air up and through the entire cab, and blowing it out the top.

Also, if you're having a problem with interference as the game is warming up, I'd still be less inclined to think it's a cooling issue, but rather something else not being right, and the temperature is just bringing it out more. Tens of degrees shouldn't make a difference to a healthy game. You want to focus on finding the root cause here, not treating the symptom.

FWIW, switchers generate noise. That's probably the thing I'd focus on first. Do you have a scope?
 
You're overthinking the cooling thing.

As long as there's a 12V fan running (even running at 5V), anywhere near the top of the cab, blowing up and out, it's going to suck enough cool air into the bottom of the cab, to cool everything in the cab.
Even with the relatively non porous Pole Position EMI cage? Doesn't seem like much air is moving through that at the moment. You had mentioned a while back that a cage + fan is better than running without a cage, which is better than just a cage without a fan (for Pole Position).

1744255185975.jpeg

I did order a bunch of 12v fans to try to change over some of my other noisier games (lookin' at you Qix and Firefox).

Also, if you're having a problem with interference as the game is warming up, I'd still be less inclined to think it's a cooling issue, but rather something else not being right, and the temperature is just bringing it out more. Tens of degrees shouldn't make a difference to a healthy game. You want to focus on finding the root cause here, not treating the symptom.
What I saw when the game warmed up previously was weirdness with the monitor.

So, I'm trying to hunt down the remaining issues with the display. To that end, I rigged up a mini-me test.

1744255425701.png

This has revealed the following tidbits of information:

1) All the pulsing and jitteryness of the disco is from the disco itself. The other monitor was unaffected.
2) The interference to the monitor is coming in through the game board wiring/video feed. It isn't present when I use the TPG on the disco while the game board is running, regardless of if the monitor frame is grounded or not. Probably because there isn't a ground loop in that case. (Nor was it present on the little monitor, though I didn't try tying an earth ground to it)
3) The screen blanking issue I described earlier is definitely a gameboard problem. It happens on both monitors.
4) The video ground is critical. I can't just use the earth ground (I tried that one for funsies)

To address the pulse/bloom on the disco monitor, I'm looking at replacing the voltage regulator, the filter cap, and one other crusty cap I Identified on the disco... (When I had ordered the cap kit for the Disco, it was missing 5-6 electrolytics that are present on the PCB. )
This includes this crusty fellow.

1744255777650.png

I didn't have the appropriate cap on hand when I initially capped this and forgot about it. Not anymore. Ordered the proper cap.
 
Last edited:
Even with the relatively non porous Pole Position EMI cage? Doesn't seem like much air is moving through that at the moment.


The cages are porous. They are vented on the top and bottom, which is in the direction of convection airflow. (Which is aided if you have an all-cabinet exhaust fan, helping pull more air up through the cab.) Heat rises, which is why the cages are designed with a solid front. It creates a draft up and through the cage, over the full surface of the boards.



2) The interference to the monitor is coming in through the game board wiring/video feed. It isn't present when I use the TPG on the disco while the game board is running, regardless of if the monitor frame is grounded or not. Probably because there isn't a ground loop in that case. (Nor was it present on the little monitor, though I didn't try tying and earth ground to it)


This is why I'm suspecting the power supply. Switching noise could be coming through the game board.

If you had a scope, you could see if/how much noise is on the various power rails.
 
This is why I'm suspecting the power supply. Switching noise could be coming through the game board.

If you had a scope, you could see if/how much noise is on the various power rails.
I do have scopes... I'll take a look tomorrow. I'm not sure if the noise was coming from the switcher how I would resolve it.
 
I do have scopes... I'll take a look tomorrow. I'm not sure if the noise was coming from the switcher how I would resolve it.


We'll cross that bridge if we come to it.

The first thing to do would be to very closely scrutinize every part on the PS (against the schematics) to make sure no components are wrong or missing.

Next, I'd look at the cab wiring, to make sure there aren't any grounding situations that are different from the stock wiring, that could be causing PS noise.

Lastly, capacitance could be added to combat noise, if the source can be pinpointed. But again, these games were designed to work from the factory, so if anything was needed along those lines, it should already be there. But that's why you'd want to make sure everything that's supposed to be there on the PS is already there, and the parts are good (as it's possible to have caps go bad that aren't electrolytics.)

But the next step would be to use a scope to find where the noise is coming from, and what the frequency is, which might give some clues as to what's causing it.
 
We'll cross that bridge if we come to it.

The first thing to do would be to very closely scrutinize every part on the PS (against the schematics) to make sure no components are wrong or missing.

Next, I'd look at the cab wiring, to make sure there aren't any grounding situations that are different from the stock wiring, that could be causing PS noise.

Lastly, capacitance could be added to combat noise, if the source can be pinpointed. But again, these games were designed to work from the factory, so if anything was needed along those lines, it should already be there. But that's why you'd want to make sure everything that's supposed to be there on the PS is already there, and the parts are good (as it's possible to have caps go bad that aren't electrolytics.)

But the next step would be to use a scope to find where the noise is coming from, and what the frequency is, which might give some clues as to what's causing it.
1000000555.jpg

There is some noise, but it doesn't seem like a stable frequency that my scope can actually lock on to.

 
It isn't going to be something you'll be able to measure with an automatic scope measurement.

Do what you did above for all of the voltages. Zoom in on the noise, and look for the highest peaks. If you zoom in far enough, you may see that the central spike in the noise repeats with some regular pattern (separated by other random stuff). Take the cursors, and manually measure the time between highest peaks.

There is noise on the case you showed above, and you can see more if you zoom in. But you may find even higher noise on some of the other voltage outputs.

Also, keep the sampling rate at maximum (1GS/s). Set the trigger level so it's just above the noise you see on screen. (Set it to edge trigger, on rising AND falling edges. These are under the trigger settings). Then slowly lower the trigger level until you get regular acquisitions. There may be larger spikes that you can lock on to, but you might not be triggering on them if the scope is just in auto mode.
 
Last edited:
I Robot is officially running in a stock configuration!

I wasn't able to get the proper choke for the monitor DC power supply, but I reached out to Coilcraft with the specs of the original choke Atari used, and they recommended another model which I used in the Monitor DC line. It's not the exact same (roughly half the inductance) but it's over the inductance Atari documented for the Firefox choke, so I suppose it's good. (Anyone want any repros? lol)

I've linked the monitor and the degauss to the original switcher and officially unhooked the auxiliary power transformer, which seemed to be the cause of the brutal noise on the monitor. I'm now able to ground the monitor! And while there is a wave present, it's more of a shadow and difficult to see. I think I prefer to have the monitor properly grounded and have a faint wave than no wave and a hazard.

1745339687166.jpeg

I modified a pole position EMI filter PCB to have capacitors in the spots indicated in the I Robot schematic, and it seems to be working A-OK.

Also, thanks to @dutchman an original back door for the I Robot monitor housing arrived, and it fits perfectly! Thank you!

1745339772160.jpeg

I still think the machine has a problem with VBLANK somewhere, but overall I am very very happy with this. I think if I try to go chase VBLANK, I'll probably fuck something up big time....
 
Last edited:
Oh, figured I'd post a tiny little bit of repro fun.

I got some properly sized translucent heat shrink and finished my repro atari choke.

1745967796153.jpeg

It's a simple part, but it looks (IMO) like the real thing, and I'm pleased.

ALSO. I did some experimenting, and I think the RFI shield is crucial for reducing the noise on the monitor, but not necessarily for what you'd expect.
I noticed that if I removed the shield and changed nothing else, the noise would come back viciously. I experimented with ferrite cores on various signal cables to no avail. But putting the RFI shield back? The monitor picture clears right up. Why?

I suspect it's a closer path to earth ground for the logic pcb. Because the cage itself is grounded, and the ground plane of the RFI shield contacts the cage, it provides a more direct path to earth ground than going through the switching power supply. Same thing with the monitor. Direct path to earth ground currently. If I unplug the monitor ground but leave the RFI cage on - noise. Because the grounding isn't "equal."

So. Grounding is fun.
 
I think it's simply because your PS is generating noise, and the RF board just adds extra capacitance to every line. (See previous comments about reducing noise). However I think you're just masking the actual issue here.

You should go back to using your scope to pinpoint the source of the noise. If you need help understanding the scope triggering to do this, just ask.
 
I think it's simply because your PS is generating noise, and the RF board just adds extra capacitance to every line.
The ps is definitely generating noise, but it's not the capacitance that's masking it I don't think. I really think it's the grounding.

If I leave the RFI board on and separate it from the cage by 1cm, the noise comes back full force.

If I then take a clip without moving the shield and clip the shield to the cage, noise is gone.
 
Back
Top Bottom