7905 voltage regulator

melchman

Member
Joined
Jan 11, 2008
Messages
841
Reaction score
3
Location
Denville, New Jersey
Why would a 7905 put out -14.5 instead of -5?

I always change them out when rebuilding AR-II boards but it happens so often I just want to know why.
 
Why would a 7905 put out -14.5 instead of -5?

I always change them out when rebuilding AR-II boards but it happens so often I just want to know why.
Because it's not regulating properly because there's no load on the output? A lot of atari games have nothing connected to -5 so you'll get a garbage reading.
 
No Load?

No Load? I am silly. Yes, the load is missing, wanted to test voltage prior to connecting the harness so as not to blow out the boards.

Should I make a molex plug with a resistor across the the -5VDC pin?
 
No Load? I am silly. Yes, the load is missing, wanted to test voltage prior to connecting the harness so as not to blow out the boards.

Should I make a molex plug with a resistor across the the -5VDC pin?


You can. This was discussed indepth a few years ago when Dokert was designing a test harness for the ARII boards. He may have some on hand or you can look up his thread and see what parts they used for the loads.
 
A Voltage Regulator shouldn't need a load. If it's reading -14 it's bad just replace it.
You are incorrect. The datasheet for the 7905 is vague because it specifies a 5mA minimum current and doesn't say what happens. In reality the component works like he describes. Test it and see what happens.

This is an FAQ where somebody plugs an AR2 into some game, maybe tempest, which doesn't use the -5 voltage and sees exactly this. For Bob's page see: http://therealbobroberts.net/7905.html
 
1. If it is reading way high -14vdc or better and it is unloaded don't sweat it, this is normal. If you have the circuit loaded and it is putting out that high, then you have a problem. Don't think just because you plugged the A/R II back into the game that the circuit is loaded. Some games simply do not use the -5vdc, and it will still read high like crazy.
2. Using a higher quality 7905, usually doesn't have this problem when unloaded. It is the difference between spending $0.22 and $0.38.
 
I work with these every day.

Here's a picture of me supplying -25v to one, and measuring the output unloaded. It's internally regulated.

IMAG0311.jpg
 
More fuel

I have more fuel for this fire.

We pulled some AR board together and tried to get to the root of the problem.

I tested six 7905's out of load and got these results:

Reading
No. Name New? Out IN
1 Battlezone Old -5.00 -5.00
2 03 Replace New -14.9 -14.7
3 02001 New -14.9 -14.7
4 02002 New -16.0 -15.4
5 NIB1 New -14.9 ---
6 NIB2 New -14.9 ---
7 OLDStock Old -5.00 -4.98

I am using a Missile Command Cocktail to test. The Battlezone board is an AR II-02 which is known working in Battlezone. 03 Replace was a recapped board as were 02001 and 02002. NIB1 and NIB2 are 7905's I pulled from Bob Robert's bags. and Old Stock was stolen from an old AR scrap board.

The Out of load reading were take using Talon2000's method shown earlier in this thread. The in readings were taken when hooked to the MC Board.
 
Could also be because of a problem with oscillation. From what I remember, you need a small cap on the outputs of the 78 and 79 series regulators to make 'em stable.
 
Could also be because of a problem with oscillation. From what I remember, you need a small cap on the outputs of the 78 and 79 series regulators to make 'em stable.

Yes, typically a 10uf cap on the output. However, I have bench tested many many of these and never had an issue. The voltage is spot on without any external components. If I get some time im going to pull some out and test them.
 
At electronics school we were taught to always use the caps across input and ground and output and ground because of possible oscillation.

It's no use measuring them without any load.
 
Did u contact Bob Roberts?
I already provided Bob's FAQ explanation in previous responses: http://therealbobroberts.net/7905.html

The datasheet I saw says that behavior is undefined below 5mA output. Maybe there is something more sinister going on there, but the datasheet is pretty clear (in that undefined way) that you cannot expect any specific behavior for the unloaded tests described above. If somebody really cares, throw a large resistor on there and load it very lightly and see if it works better.
 
Back
Top Bottom