Thermal Camera Recommendation?

[ArcadeTechGW]

Based on that observation further review of the logic board power and typically faulty components is required.

(Inadequate test wiring causes voltage drop from source to load, not excessive current).

OBSERVATION: Ah, that large power resistor in parallel with the LM323. 2ohm 20W. There's some good Arcade Classic Repair vids on this circuit. The resistor is there to help regulation and startup inrush. But I believe if there is a regulation issue, the LM323 should be doing alot of the work (regulating voltage and passing current) under normal conditions. (Though if you work thru Ohms Law and assume the input is 12V (10Vunreg), the output is 5V and the current is 2amps, then the regulator has an impedance of 3.5ohms and it and the resistor are sharing the load current and both are getting hot, but the regulator is less hot than if the resistor was absent from the circuit. This being said it is possible for the 5V circuit to operate at higher current than the rating of the LM323 since some of the load current is supplied by the parallel 20W resistor). Perhaps excessive power issue is occurring if either the resistor is out of spec AND/OR the LM323 is faulty. "Power and Connections"

Some more trivia. If using a 2amp 250VAC 3AG slow-blow fuse and noting that this fuse takes "20 seconds to open", this implies by the timing curve that almost 4amps is being drawn on the AC side.

Good use of engineering here. Good explanations.

 

[jlweatherly]

Perhaps excessive power issue is occurring if either the resistor is out of spec AND/OR the LM323 is faulty.

Power resistor measures 2 Ohm in circuit, and smoked a little bit the first time I had the power brick plugged in to everything. Smoke never came back enough to see when I had a camera on it.

Also, I was dumb at some point and killed the original LM323 by reversing voltage going into it (was confused by readings I was getting and thought my wires were backward in the connector, they weren't.)

Currently have a new LM323 installed from APAR that the fuse issue persists with.

 

[VectorCollector]

Yes, the power resistor, at startup, is managing the inrush current into the logic board's 5V load. It will have a spike of power (and perhaps "smoke" for a second) and then it and the LM323 will share the load theoretically well within their specifications.

Reversing the input voltage might have damaged more than the LM323 as the 20W low-ohm resistor is still in circuit. Thus,, the load side will see the inverted voltage.

Got anymore confessions to make? LOL

 

[jlweatherly]

Reversing the input voltage might have damaged more than the LM323 as the 20W low-ohm resistor is still in circuit. Thus,, the load side will see the inverted voltage.

Got anymore confessions to make? LOL

Lol, well I think that was the worst of them so far, at least from what I can recall.

After I fixed (aka unfucked-my-fuckup), I tried the game with only 5V from bench supply and things seemed to be the same as before, semi-working with garbage on screen. So I may have got lucky with my mistake?

I figure if the voltage reversing had done further permanent damage then I wouldn't have been able to even any normal operation after it was corrected.

Edit: And to be clear, reversing the voltage definitely killed the og regulator, board wouldn't work on the bench until I replaced it.

 

[ArcadeTechGW]

Power resistor measures 2 Ohm in circuit, and smoked a little bit the first time I had the power brick plugged in to everything. Smoke never came back enough to see when I had a camera on it.

Also, I was dumb at some point and killed the original LM323 by reversing voltage going into it (was confused by readings I was getting and thought my wires were backward in the connector, they weren't.)

Currently have a new LM323 installed from APAR that the fuse issue persists with.

Lift a leg on that resistor and check it out of the circuit. The smoke may have been nothing, or it may have damaged the resistor.

I know, I've done it.

 

[XorLor]

Thank you for the example pics. I'm curious how quick the refresh rate is on the screen? Does it react quickly to changes in temp or where it's aimed?

The frame rate is only 9 Hz but it's fast enough to see components heating up in real time.

 

[jlweatherly]

Reversing the input voltage might have damaged more than the LM323 as the 20W low-ohm resistor is still in circuit. Thus,, the load side will see the inverted voltage.

I am pretty concerned about this too, so will be checking C24 and any other electrolytics closely.

More importantly, for a simple linear regulator the input current will only be slightly higher than the output current.
Conversely, if the input to the regulator is fused at 2amps, then one surmises that the output current will be less than 2amps.
If only powering the +5V rail, this can be confirmed.
But when reviewing the schematic, the 10V unreg rail is also used by other components directly. So any fused 10V unreg rail will have even more amperage than that alone for the +5V rail.

EDIT: The 2A fuses are on the AC input side and not the 10VDC unreg output.

Just saw this edit,

The manual only refers to 4 fuses total, two on power brick, two on monitor. My power brick has 5 total, the two it mentions in the manual, then 3 more spots which are fused as follows (looking from the top, left to right): 25A, 2A, 2A


Can't find mention of these three fuses in the schematics, and the fact one of them is 25A seems a bit silly. Here's where I think they exist in the circuit:

Pretty sure my fuse that's popping is on 10V unregulated just after the MR750's because I lose 5V on the game when it happens. It measured near 12VDC last time I checked with the connector unplugged. No significant AC voltage detected there.

It's the last 2A here that is for 10V unreg that we've been discussing. Red wire, green stripe:

 

[VectorCollector]

So, the power brick does not match the Starship manual.
What to do? ANSWER: research other games around the same vintage that use that brick and have similar functionality.

OBSERVATION: Canyon Bomber (1977) uses that brick with five fuses. The fuse in question is noted to be 8AMP 125V FAST BLOW.
So, we must ask the question -> why are you using a 2AMP slow blow in that location, assuming that fuse location is wired the same.
The use of an 8AMP would make more sense for the 10Vunreg rail, no?

https://www.arcade-museum.com/manuals-videogames/C/CanyonBomber.pdf

 

[VectorCollector]

Note, fuse values are not exactly the same as in the Starship manual, but are similar. Atari did slighly modify values from game to game. Each year, as well, Atari fixed errors/typos/missing-info. But even the Canyon Bomber manual is inconsistent with the schematic drawing showing the fuse strategy.

Another observation is that most "large" Atari PCBs draw about +5V at 3-4A (this includes vectors like Asteroids and rasters like each Pole Position PCBs).
For a linear power circuit as we have here, the amperage on the +5V output is the same (more or less) as the input current off of the +10Vunreg (neglecting the other circuitry off of this rail). So, fusing typically is defined at 150%-200% of nominal. This confirms that the 10Vunreg fuse should be in the 8A range.

So we conclude the following,

1. wrong fuse value being used (at F5)
2. logic board is faulty but not the cause of the fuse opening
3. the observed 3.6A off of the +5V is consistent with other Atari boards
4. the need for a thermal camera is not needed here to fix the logic board - TBD (laughs)

 

[jlweatherly]

Good points about researching games of similar era. My other thought was to ask for anyone else who has a working one to check their current fuse setup.

This game has never worked in my possession, this is my first serious foray into trying to get it working after it's sat in my garage for 4 years 3 years. So either it came with a 2A and that's what I've always replaced it with, or I screwed up at some point and swapped it around. Usually I'm pretty good about over-documenting with pictures.

I'll need to go back and review the pics I took back when I picked it up, which have since been archived to the computer so not easy to get to until I'm home tomorrow.

 

[VectorCollector]

Looking at VAPS, there are 12 instances of Starship_1 original cabinets listed.
Some members are consistent KLOV contributors, so you might get lucky and one of them will check the fuse values for you.

Game Census for Starship 1

This Starship 1 page is a list of all the collectors who own or want the coin-operated Videogame machine entitled Starship 1, manufactured by Atari in 1976

www.arcade-museum.com