Atari Color Vector Power Brick & AR 2 Board - Acceptable Voltage Ranges

mattmatt

Member

Donor 2024
Joined
Nov 1, 2024
Messages
57
Reaction score
43
I'm going to be bringing up monitors on a couple different games: Star Wars, Tempest, Black Widow. I'm following @andrewb 's guide. So, I'm starting with the power supply: 034482-02D

Since I'm not a EE, schematic isn't my first language, and I like using checklists, I'm making a cheat sheet as a companion to the schematics to help me do these machines, and to be a quick reference in the future.

I'm attaching a shot of my cheat sheet, with the voltages I measured with the brick not hooked up to anything.
Question 1: Are the "Measured" voltages I have in the lower center table - right hand column ok?
Question 2: I couldn't find what the acceptable voltage ranges are for each pin when the brick isn't hooked up, so if I can get some help on those values I would appreciate it.
I'm assuming this is the correct way to test it prior to hooking up the AR board? Then next step is to test the AR board.... If this is wrong, please let me know.

Power brick checklist.jpg
 
Last edited:
I get an error when I try to view the attachment.

The Tempest schematics shows the voltages for the main brick connector, J5:

1742439873250.png


The 10.6V DC should actually measure between 13-15V at the brick with no load. It runs high, that's normal.

All AC voltages should be measured BETWEEN the two pins above. (i.e., one DMM lead on each pin, direction does not matter. You do not measure each pin to ground.)

All AC values should be within +/-10% of above. It's ok if they are not spot on. If they're off by more than that, check to make sure which configuration jumper is plugged into J3, the 9-pin connector on the brick. (For USA 120V operation, it should be the one with yellow wires.)
 
I get an error when I try to view the attachment.

The Tempest schematics shows the voltages for the main brick connector, J5:

View attachment 807467


The 10.6V DC should actually measure between 13-15V at the brick with no load. It runs high, that's normal.

All AC voltages should be measured BETWEEN the two pins above. (i.e., one DMM lead on each pin, direction does not matter. You do not measure each pin to ground.)

All AC values should be within +/-10% of above. It's ok if they are not spot on. If they're off by more than that, check to make sure which configuration jumper is plugged into J3, the 9-pin connector on the brick. (For USA 120V operation, it should be the one with yellow wires.)
Thanks, I re-attached the photo.
 
Those voltages are fine.

Next, connect the AR, then test the 10.3V DC, +5V DC, 36VAC, +22V DC, -22V DC, +12V DC, and -5V DC on the AR, with the game board NOT plugged in.

Also, you should test the AC ripple on the 10.3V test point, which will tell you how healthy your big blue filter cap on the brick is. You should check this any time you test voltages in your cab. If it's ok, your big blue is fine and you do not need to replace it. See this thread for how to measure:



Additional Notes:

- The 10.3V DC test point on the AR is the same voltage as the 10.6 on the brick (Atari named it both ways in the schematics, but it's the same voltage). However it might measure a volt or two less at the AR, that's normal. You should see at least 12V. If not, the fuse block tabs on the brick are usually in need of cleaning.

- Use any GND test point on the AR for all DC measurements, and the AC ripple. (Any one, they are all the same.) In general, always use a ground on the same board you're measuring voltage on.

- There is no test point for the +5. However you can use the positive (+) side of the C1 cap on the AR. Set the 5V to 5.00V using the pot on the AR. Note that you will need to bump this up when you connect the game board, as there will be around half a volt lost in the wiring and connectors. So if you wanted to set it to 5.5V at the brick with no load, that's fine too, you'll still need to tweak it later.

- The 36V AC is the same as the one on the brick, so it should be about the same. (However you still want to test it, to make sure there are no breaks between the brick and AR.) You also need to measure BETWEEN the two 36V AC test points on the AR. If you measure each test point to GND, you'll get half of the voltage (18V AC).

- The +22 and -22V DC run high. They'll be around +/-24V. That's nornal.

- The +12 and -5 are actually not used by Tempest (or any other color vector games except Star Wars). So they technically are not needed, and you can use an AR-II-04 in any of these games (which don't have those voltages). Tempests shipped with AR-II-02's from the factory (which have the +12 and -5V sections populated), even though the +12 and -5V are not used.

- If you do measure the -5V, it's normal on some AR's for it to read artificially high. This is because some 7905 regulators measure high when there is no load. Not all 7905's do this, it depends on the manufacturer, so it may not be high on some AR's. If it is high, you can check this by connecting a 1k resistor between the -5V test point and GND, and the voltage should come down to -5V. But again, this is somewhat moot for most color vector games, because they don't use the -5V.


Once all of the above are ok, make sure the monitor is UNPLUGGED, plug the game board in, then power up again and measure and tweak the 5V on the game boards. For Tempest, test it on BOTH the main and aux boards. The aux board gets its 5V power from the main, so it will always be a little lower. You want the aux to be at least 4.90V (a little lower than 5.00 is ok), and the main should be around 5.10 give or take, which is ok. The issue is if you just set the main to 5.00, the aux can be too low, so it's important to test both. (And always use the GND test point on the board you're measuring the given voltage on.)

Also, test and make sure the -15V DC and +15V DC are good on the main board. (They both should be within +/- half a volt of target.)

Once voltages are ok, make sure the game is playing blind. If it is, do not plug in the monitor yet. You'll want to do that in stages. Details for how to do that are explained in my 6100 bring-up guide. Download here:


Sorry for the extra detail, but I wrote it all out so I can just bookmark this thread, and refer people to it when they want to assemble an Atari vector cab from complete scratch (i.e., from the brick up), as this is everything you have to do to make sure you don't blow anything up. (And your diagram is useful as well.)
 
@andrewb ,
This is wonderful. Thanks so much for the detailed explanation!

I have just a couple clarifying questions:
  1. So the +5, What is the min/max voltage you should see on this when it's not hooked up to the gameboard (to know if its a problem)?
    1. Could you please explain why we don't test the +Sense? Is it redundant or something?
  2. In general, is there a rule of thumb on all the voltages (in electronics in general) ? Like a percentage once it is on the AR board, or are the min/max dependent on the application? (seems like some things would need a tighter tolerance).
  3. I saw a YouTube with a guy who had a sense mod jumper plug when he did the test, but it doesn't seem like that is necessary? Or is it a way to dial in the +5VDC on the bench without plugging in the gameboard?
 
Good questions!

1) For all AR's EXCEPT Asteroids, the full range of the pot should be from about 4.4 to 5.6V, with no game board attached, give or take a tenth or so. (Asteroids AR's will be about a volt higher than that, because their slightly different design makes them read artificially high with no load). For Asteroids (not Deluxe) AR's, you want to set them to 6.2V with no load, which will come down to 5.0 once you plug the game board in.

Generally speaking, as long as you can dial in 5.0V with no load, you should be ok, since most of the common failure modes for AR's will result in a way too high (~10V) or way too low (~0.5V) voltage with no load. So the no-load check is really just to make sure your AR isn't fried, and make sure it's set roughly in the ballpark before you connect the game board. The main thing you want to watch out for is to make sure the AR isn't measuring anything like 8-12V, which can happen for various failure modes (like if the 2N3055 is failed short, or if R29 or R30 is burned up or cracked).

1.1) You can use the +SENSE test point to test and set the voltage with no load. With no load, the result will be the same as the C1 cap. The +SENSE and -SENSE points will measure differently when the game board is connected (as they are indicating the voltage seen at the game board, even though the test point is on the AR), so I don't recommend them, just to not confuse people. (Plus the fact that there are +SENSE and -SENSE test points, and many of the old farts around here can't see the difference without their reading glasses, which could potentially confuse things more.) In general you want to measure voltage on the board you care about. So measure AR voltages on the AR, and game board voltages on the game board.

The C1 cap gives the true 5V reading at the AR, which is what you want to set with no load. (Or if you want to see what he voltage is at the AR, when the game board is connected and you've adjusted the pot to give you 5.00V at the game board, if you want to see how much voltage is being lost in the wiring and connectors). You should see roughly half a volt drop between the AR and game board, under normal conditions. If it's significantly more than that, it means it's time to clean and DeOxit the edge connector fingers, and DeOxit and reseat the Molexes on the AR, because too much voltage is being lost between the AR and game board.

2) No rule of thumb is universal, so take this with a grain of salt. But all electronics are designed to tolerate some variation in supply voltage, which is often on the order of +/-10%, for typical consumer electronics. Some things need higher tolerances, and sometimes tolerances are mentioned in schematics, etc. But when people design circuits, one of the things they need to account for is the fact that every component in the circuit will have some tolerance from the manufacturer (be it 5%, 10%, 1%, whatever is specified for the parts used). So they will run what are called 'monte carlo simulations' during the design process, which will vary all of the parts by random values within a stated range, to make sure the circuit still functions properly under any possible combination of parts tolerances.

Most of this arcade stuff was manufactured very quickly, and companies used a wide array of suppliers. So the measurements you will see when comparing two Asteroids boards for example can vary quite a bit depending on the tolerances of the parts on that specific board. In general, the easiest way of knowing what the ballpark should be for any given measurement is to compare it to a working version of that same board. But even then, no two boards will be exactly alike.

3) A sense jumper is not necessary, if you are testing the board with no load (either in the cab, or on the bench.) This is why it's ok to test an AR in a cab without the game board plugged in, and nothing bad will happen. If you DO want to test with a load on the bench, that's when you need a sense jumper. But there's really no benefit to testing with a load on the bench, unless you're repairing a stack of AR's, and you have an appropriate test load. For a single game you'll need to test with the actual game board you're using anyway (as different game boards draw different current). So it's sufficient to do the initial test with no load, just to make sure the AR isn't outputting 10V because it's fried. Then test with the actual game board.
 
Back
Top Bottom