UAV and power mod installed, but stops after a few seconds.

[Zilch]

Had fun installing the mods, my first attempt at such a thing. 

 

It's thrilling to see the Atari logo light up my screen...

 

But after about 5 or 10 seconds, my poor 5200 stops working. It'll go to a glitchy screen, black, or usually a green screen. 

 

Google-fu suggests it may be a bad GTIA, memory, something loose...

 

Anyone know that might be causing this? Ice cleaned the cart slot using the credit card method and swabbed the carts. 

 

 

[+-^CrossBow^-]

Was it working properly prior to any modification work being done? If it wasn't or that wasn't tested, then it could also be a flaky voltage regulator that isn't doing its job properly etc. There are a number of things it could be. On the IC side of things, I'd start with the CPU first. Did you keep the 4050 IC in place when you installed the UAV or did you remove it completely? If you left it in place, I was going to suggest pulling the UAV out and just testing on RF output to help isolate the UAV setup as being part of the issue.

 

[Zilch]

Unfortunately, the system was untested before I installed the UAV, aside from powering on. My only TV didn't recognize signal from the power/RF switch box even when I routed it through a VCR to RCA output, so I suspected the visibly oxidized switch box was the culprit (hence the UAV mod.)

 

I did pull the 4050 out and attempted to run the system with only the UAV card in its place. I got the same results you see in the video, so I soldered the original 4050 to the UAV card and put it back in place together. Same behavior, unfortunately. 

 

I did remove the coax cable as part of my mod (and ease of work) but I can reinstall it and try running through the VCR to see if that works. 

 

Edit: I case it's relevant, I seem to get a few more seconds of function time if I wait longer between power-ons. I waited overnight and got a good 20 seconds of the Star Raiders opening screen before it noped out. After that, power cycling tends to go straight to black or green screen 

 

Also, may be coincidence but when I tap the console or swap controls, it seems to shut out, too. Wondering if there's a physical looseness contributing. 

 

Hey, thanks for the idea. I'm new to the world of electronics, but am absolutely *loving* it so far. One thing that I always appreciate is people willing to help with knowledge and know-how, so, thank you. 

Edited October 5, 2023 by Zilch

[+-^CrossBow^-]

Well the 4050 is only still needed if you plan to still use the RF output. If you don't care to keep that working, then the 4050 isn't needed at that point. but as you already have it reattached to the UAV that is moot.

 

Okay, so given that if you leave it off for a bit it will work briefly indicates that something is getting hot very quickly and shutting down power to part of the circuit. Issues like this, I always follow the ASAP rule...

 

Always

Start

At

Power

 

So you need to get a voltmeter and verify that input voltage and output voltage from the twin voltage regulators in the system. Those are the small squares that are screwed down onto the large black plates near the power area. You measure the voltage with a meter by putting the black probe of the meter along any ground point on the board. Can even touch the RF modulator housing as that is always attached to ground. The red probe when attached to the left most leg or pin of the voltage regulator is the INPUT voltage. That will be somewhere around 10 - 12v on average from a 5200. The right most leg or pin of the voltage regulator is the output voltage that is provided to the rest of the 5200 logic. That should read around +5v. 

 

So the VR readings need to be checked first and then we work our way down from there to see if the culprit can be found.

 

 

[Zilch]

I think you got it! VR2 is giving me voltage readings, but VR1 on the backside of the heat sink is reading nothing at all. I'm assuming it's toasted and will replace it. May as well get VR2 while I'm at it...

 

Optimism! Thank you! 

 

Edit: I noticed quite a bit of heat from the heat sink with power on for a minute. I assumed this was the heat sink doing its job, sinking heat. Is it possible the faulty VR was making it cook a bit too hot, also? 

Edited October 9, 2023 by Zilch

[+-^CrossBow^-]

30 minutes ago, Zilch said:

I think you got it! VR2 is giving me voltage readings, but VR1 on the backside of the heat sink is reading nothing at all. I'm assuming it's toasted and will replace it. May as well get VR2 while I'm at it...

 

Optimism! Thank you! 

 

Edit: I noticed quite a bit of heat from the heat sink with power on for a minute. I assumed this was the heat sink doing its job, sinking heat. Is it possible the faulty VR was making it cook a bit too hot, also? 

You should still be showing input voltage to both VRs? If you aren't showing input voltage then something else prior to the VRs is faulty and will voltages checked along the way to see where the issue might lie.

 

[Zilch]

If I'm reading the schematics correctly, it looks like C43 should have voltage, using the method you describe. However, it doesn't. Might this be the culprit, or is it more likely something further back in the series? 

[+-^CrossBow^-]

Check for input voltage on both sides of the MJE210s. They are just lower left of the RF modulator. Also check for voltage on the power resistors just after them that are side by side.

 

 

[Zilch]

Thanks again. The MJE210's have voltage on both sides, and resistors 55, 56, 57, and 58 all have input voltage. 

 

Just to see, I soldered in a new 7805 for VR1. It reads 9 volts going into it. It gets too hot to touch, as does the VR1 side of the heat sink. Meanwhile, VR2 is cool to the touch. 

 

I have continuity through the new VR1 and the black L9 that follows it. 

 

This is definitely educational for me. 

 

My testing, in case seeing anything is helpful.