- Joined
- Feb 22, 2008
- Messages
- 8,889
- Reaction score
- 529
Article on the Atari Audio Regulator SENSE mod
- Thread starter bit_slicer
- Start date
Remember, the Atari engineers were trying to ship a product that had an expected lifetime in the 5 or less year range. The fact is that within the initial 5 years or so, the sense circuit minimized the amount of effort that an OP had to put into keeping the thing working. 30 years on, we now think of it as more trouble than it's worth, but back in the day I imagine it saved lots of time and effort on the production line as well as in the field.
This scenario above would not be a gameboard problem. While a gameboard can go bad and drag down the voltage.....it would drag down the voltage across the entire circuit. Meaning, the voltage would also be low at the Audio/Regulator PCB.
If the Audio/Regulator PCB is known good with correct voltages and the edge connector is known good.......this only leaves two possibilities of where the voltage loss is happening.....the connectors at the Audio/Regulator PCB, or the wiring itself.
Edward
D_Harris
Well-known member
I forgot to mention the connectors/wiring. But I assume that if they get warm to the touch, then that is an indication they are a problem.
So your saying that if the following voltages are off on the game board:
-22VDC +22VDC +10.3VDC +12VDC -5VDC
Then each of those same voltages would be off on the AR PCB, and vice versa, whether the sense had been disabled of not?
(Again, I see no +5VDC test points on my AR PCBs).
Thanks.
Darren Harris
Staten Island, New York.
The sense circuit (and sense mod) only affects the 5VDC line. It has nothing to do with the other voltage lines. It's not uncommon to have a voltage drop from a power supply....to it's final destination. The big question....how much of a voltage drop is acceptable? Ideally, as little as possible
. We're dealing with 20-35 year old machines....some voltage droop is going to happen. I try to keep mine under .15VDC......from beginning to end of a power circuit. A minor drop will not cause a connector or wiring to be noticibly warn.....or, depending on the amount of voltage dropping, it could take some time for the heat to be noticible. There's only so many places for voltage to drop......connectors and wiring. In the case of an old Atari machine you've got Audio/Regulator board connectors, wiring, edge connector....and that's it.Atari never put 5VDC test lugs on their Audio/Regulator boards (who knows why), but you don't need a test lug to check voltage. Midway never put test lugs on any of their power supplies. You just chech the voltage in the circuit....
Audio/Regulator board 5VDC at positive side of capacitor C1.
Edward
Edward
D_Harris
Well-known member
Don't you mean that the resistance is what causes the heat and not the voltage drop itself? I only mention this because I thought that a faulty AR PCB could also be the culprit in a voltage drop.The sense circuit (and sense mod) only affects the 5VDC line. It has nothing to do with the other voltage lines. It's not uncommon to have a voltage drop from a power supply....to it's final destination. The big question....how much of a voltage drop is acceptable? Ideally, as little as possible
Ok, so if the test point on a PCB show inadequate voltage, then the problem is not the PCB, but the wires and connectors. (You're not including the AR PCB as a possible problem?).
Thanks. That's one of three questions I've been trying to get an answer to for the past several days. I just tested the one in my Millipede at 5.63V, so I brought it back down to 5.V.
But I don't think that is the reason I blew a cap on the game board. My +22V is at about 24.77V, so I assume that since there is no way to adjust this voltage(unless the problem is in the wiring harness and/or connector) I'll have get an AR repair kit and shotgun it.
But I still need to find out if when to bypassing the edge connector am I supposed to attach the *5V lead from the game board to the positive side of C1. (And if a "Big Blue" can be the source of a voltage problem if it is dented).
Thanks.
Darren Harris
Staten Island, New York.
Yes, resistance (at a connnector) causes heat and voltage drop. A faulty A/R board won't cause voltage drop, it will cause voltage loss. Maybe we should make sure were on the same page. An example of voltage drop: your power supply is reading 5VDC, at the power supply, on it's 5VDC line....the motherboard reads 4.5VDC at the 5VDC line......thus, you have a .5VDC voltage drop through the entire 5VDC circuit. A faulty A/R board would cause 2.5VDC at the A/R board's 5VDC line and at the motherboard. A bad motherboarrd can also cause this same problem. The key here is....is the voltage the same at the power supply AND at the motherboard, or is it different.
Didn't you say the A/R board was good? Inadequate voltage can be caused by the A/R board, motherboard, wiring, or connectors.
The 22VDC (and the -22VDC) circuits are unregulated. An unregulated circuit will not be exact. In the case of Atari, they will generally be in the 24-25 volt range. This is fine.
Edward
D_Harris
Well-known member
One of the two AR PCBs I tried tested as follows, with the other putting out similar values :
10.3V = 12.50V on pin #5.
+22V = +24.94V on pin #3.
-22V = -25.12V on pin #4.
Also,
ARII PCB #1: For +12V and -5V I get -11.76V and -5.20V, respectively.
ARII PCB #2: For +12V and -5V I get -11.97V and -8.61V, respectively. (?!?)
Anyway, I assume that I'm not supposed to get much of any reading on the game board at pins A, 1, and 22, which are "Sense Returns". (I now have to learn how to check voltages at the chips legs).
I've decided that I'd rather not have the sense circuit adjusting things, so I'll do the necessary mod.
Thanks.
Darren Harris
Staten Island, New York.
OP
OP
bit_slicer
New member
I just tried it, and all the pics come up
Sorry
Just a note, this article needs to be overhauled, especially the math portion. Darren gave some excellent feedback and I still need to incorporate the changes. I have designs to move this page and all the others to a more-permanent server. It's on the to-do list!
-5.20VDC is fine for the -5VDC line.
-8.61 is definately too high for a -5VDC line
The other two readings (-11.76 & -11.97) are strange. If they were positive, that would be fine for a 12VDC line......but being negative voltages.....something's definately not right.
Edward
D_Harris
Well-known member
Just a note, this article needs to be overhauled, especially the math portion. Darren gave some excellent feedback and I still need to incorporate the changes. I have designs to move this page and all the others to a more-permanent server. It's on the to-do list!
Great! Will you include specifics on bypassing the edge connector also? (I've been trying to get an answer for several days on exactly what two points on the ARII I'm supposed to connect leads from the +5V and GND test points on the game board).
That was my error. Several days with Millipede on the brain and little sleep will do that. That. Those are actually positive voltages.
It should have read as follow:
ARII PCB #1: For +12V and -5V I get +11.76V and -5.20V, respectively.
ARII PCB #2: For +12V and -5V I get +11.97V and -8.61V, respectively.
Even though I blew a game board cap with the first ARII connected, it is obviously the second ARII which has a problem and will have to be rebuilt.
Darren Harris
Staten Island, New York.
Last edited:
DarrenF
Well-known member
Well, if this thread weren't confusing enough already... now we've got two different Darrens involved
Well, if you were to go the route of by-passing the wiring harness, and installing soldered-on +5V wires (which makes it a PITA to remove/repair just the PCB or just the ARII)...
You'd likely want somewhere on the top of the ARII, rather than having wires coming from the back, where you can't see them (for when the break loose later... and you'll want to see them shorting against who-knows-what). I'm assuming you don't have test points on your ARII, or you wouldn't be asking. I'm sure there are quite a few choices of locations that will work... Looking at a schematic, one that pops out to me is at C1. The +5V output exists across it. Of course the GND is on the "-" leg of the cap, and the +5V on the other leg. CR1 is another candidate. You could also tap in at a variety of resistors, but you'd need to be carefull about which end you connected to.
-8.6V on the -5V line... There are really only a few components in the -5V supply. But before that, lets discuss how and where you measured it. You put your red DMM lead on pin 9 of J10, and your black DMM lead on either pin 5, 8, 11 or 12 of J10? With the meter in DC voltage mode?
Well, if you were to go the route of by-passing the wiring harness, and installing soldered-on +5V wires (which makes it a PITA to remove/repair just the PCB or just the ARII)...
You'd likely want somewhere on the top of the ARII, rather than having wires coming from the back, where you can't see them (for when the break loose later... and you'll want to see them shorting against who-knows-what). I'm assuming you don't have test points on your ARII, or you wouldn't be asking. I'm sure there are quite a few choices of locations that will work... Looking at a schematic, one that pops out to me is at C1. The +5V output exists across it. Of course the GND is on the "-" leg of the cap, and the +5V on the other leg. CR1 is another candidate. You could also tap in at a variety of resistors, but you'd need to be carefull about which end you connected to.
-8.6V on the -5V line... There are really only a few components in the -5V supply. But before that, lets discuss how and where you measured it. You put your red DMM lead on pin 9 of J10, and your black DMM lead on either pin 5, 8, 11 or 12 of J10? With the meter in DC voltage mode?
D_Harris
Well-known member
The idea is to add solder two leads to the test points at the top left of the Millipede game board and two leads to the AR II PCB, and connect them via a molex connector.Well, if this thread weren't confusing enough already... now we've got two different Darrens involved
Well, if you were to go the route of by-passing the wiring harness, and installing soldered-on +5V wires (which makes it a PITA to remove/repair just the PCB or just the ARII)...
I want to solder only on the component side of the AR PCB. There are test points, but not for +5V.
Looking at the image, the idea was to solder one lead to C1, which is the positive leg of that cap, and one lead to the ground. (Both marked with a red "X").
I measured at the two indicated locations in the image below.
Thanks.
Daren Harris
Staten Island, New York.
Attachments
DarrenF
Well-known member
That should work fine to get +5 from the ARII to the game PCB outside of the OEM harness.
Looks to me like you were measuring the -5VDC properly (no offense intended, I've just seen too many cases of people making voltage measurements incorrectly).
Personally, I'd check for AC content on the -5V output, and also look at the input to the regulator as well (negative end of C31, both DC and AC measurements). Depending on what I saw, I might suspect the regulator (7905, Q9), or it could also be one or more bad cap (i.e. C24, C31, C19).
D_Harris
Well-known member
I assume that you have to have a game board plugged in before you check voltage on the ARII, so I put the Millipede PCB that I blew a cap on back in.
I don't know what the "the input to the regulator" is or "7905, Q9", but I did check the voltages below:
ARII PCB #1:
At C31 I got 10.3(AC) & -24.8(DC)
At -5V I got 1.5(AC) & -5.2(DC)
ARII PCB #2:
At C31 I got 8.3(AC) & -24.8(DC)
At -5V I got 3.1(AC) & .002(DC)*
*That probably would have reached zero, but I heard a pop accompanied by smoke when the 7815 blew on the game board, which is right next to the cap that previously exploded.(See pic). (Now the red light no longer comes on on the Millipede game board).
So I turned off the machine and then the meter started reading -8.61V. (?!?) So I unplugged the game.
Thanks.
Darren Harris
Staten Island, New York.
Attachments
Last edited:
DarrenF
Well-known member
Not at all. In fact, I usually make sure all the outputs to an AR/ARII are good before I connect any game PCB to it. Just because its outputs are all good w/o a game doesn't mean the ARII is fine, but if they're jacked with no PCB, they certainly aren't going to get any better.
By "input to the regulator" I meant the negative end of C31; which is the voltage that goes into the -5V regulator, which is a "7905" (part number) and is identified on the PCB and schematics as "Q9". You should find it mounted to the heat sink.
You readings indicate quite a bit of AC ripple on the -5VDC output. And the blown cap and voltage regulator indicate that there was probably an assload of AC on the +22VDC output of the ARII as well.
I'd just replace all of the electrolytic caps on the things (any that aren't bad yet probably will be soon). I'd also check the diodes (CR5, 6, 7 & 8) for shorts too (even though that should cause a fuse to blow on the xformer ass'y instead).
Oh, BTW, looking at the Millipede schematics... the -5VDC isn't even used. All it pull from the ARII are +5, +22 & -22.
I just did a pair of these for my Pole Position. I used the test point for ground, and got the +5 from the resistor circled in green. This seemed like the most logical point to hang a heavy-gauge wire for me. I soldered a wire to the leg that has a quick disconnect on the other end so the board can be removed.
Attachments
D_Harris
Well-known member
Ok, then I don't need to have a game board connected.
Ok, since I checked the negative leg of C31 that as covered.
What would be considered normal AC readings?
No shorts that I can see.
I notice I saw nothing to do with -5V on the game board, but I did see -15V, +15V, and -28V.
Thanks.
The pics below are of my Pole Position. which was fixed by Dick Millikan. There are two leads soldered to both sides. If I remember correctly, the red wire is +5V and the blue wire is ground. I guess I can just tie the blues together and the reds together and run them to the same spots on the ARII as you did. I have some two position Molex connectors I can use.
Darren Harris
Staten Island, New York.
Attachments
Instead of tying them together, I ran each board's +5 and ground to its own ARII so I could adjust the voltage for each one.
DarrenF
Well-known member
Well... it depends. Less on a regulated line than an unregulated one. But typically it should be in in millivolts (10s or maybe 100s of mV), usually not in volts.
I guess I'd be concerned about the filter caps if the AC ripple is in excess of, say, 5% of the DC value on a regulated supply. For unregulated, perhaps up to 10% or 15% would be OK.
