Power supply 5v regulator - what's wrong with this picture? (mr. do)

TheShanMan

Well-known member

Donor 2011, 2024
Joined
Sep 9, 2008
Messages
5,657
Reaction score
725
Location
Hillsboro, Oregon
Power supply 5v regulator - what's wrong with this picture? (mr. do)

I'm trying to get my Universal power supply working. Right now the regulated 5v supply is putting out 10v, which is the same voltage as the input to the regulator. So initially I thought the regulator was bad, but I got a new one and the same problem is happening.

Here is the circuit for the 5v supply. The input is 9vac.

attachment.php


There is a 20 watt 3.9 ohm resistor connected between the input and the output. Seems kinda screwy to me, though I'm sure that circuit diagram is correct in that regard. But the effect is that it is pulling up the output to 10v. If I disconnect the resistor, the output is 5.03v. Bear in mind I do not have this connected to my mr do board (I don't want to fry it with 10v!).

I would love an explanation of what that resistor is supposed to do (the datasheet for the 78H05 regulator doesn't have such a resistor in the sample circuit).

Even more so, can anyone tell from this circuit what might be wrong with my PS?
 

Attachments

  • mr do 5v supply.jpg
    mr do 5v supply.jpg
    31.1 KB · Views: 304
Last edited:
Not really sure what the purpose of that resistor is. I've seen them used as bleeders across big filter caps, but not usually that low of a resistance. That resistor basically bypasses the regulator.
 
Early Atari boards like Sprint 2 use a very similar setup. I can't explain what the purpose of the resistor is, but I remember on my Sprint 2 it was very important that the resistor be within specs or the voltage would not be correct. The Sprint 2 board uses an Lm323 for a voltage regulator.
 

Attachments

  • Sprint2_5v.JPG
    Sprint2_5v.JPG
    29.4 KB · Views: 17
Its pulling your voltage up to 10v because you have no load. Connect a low impedance load across the outputs and you'll have more voltage across the resistor. Most likely its either to balance output impedance or current limit the regulator.
 
I wondered about that! What would be a good test load so I don't have to risk my board? I tried a coin door lamp and that pulled it down to a little more than 9v, but obviously not enough of a load.

Thanks guys!
 
I would love an explanation of what that resistor is supposed to do (the datasheet for the 78H05 regulator doesn't have such a resistor in the sample circuit).

Even more so, can anyone tell from this circuit what might be wrong with my PS?

The shunt resistor provides extra current beyond with the regulator can supply.

Assuming everything's regulating properly and there's a sufficient load in place, there's 10V on one size of the resistor and 5V on the other, so at 3.9ohms, you've got an extra 1.25A being supplied to the load by the resistor, so the regulator isn't being stressed.
 
Great explanation, Mark. I appreciate that!

So what would be a good test load? Maybe an R902 off of a g07 chassis (2 ohm, 7 watt)? That would make it 2.5 amps at 5v and the fuse is 8 amps. If I remember my wattage equation, that would be 12.5 watts, but I assume that would be OK if I only conduct the test for like a second (before the resistor can overheat)?
 
The shunt resistor provides extra current beyond with the regulator can supply.

Assuming everything's regulating properly and there's a sufficient load in place, there's 10V on one size of the resistor and 5V on the other, so at 3.9ohms, you've got an extra 1.25A being supplied to the load by the resistor, so the regulator isn't being stressed.

Learned something new today. Thanks for the explanation.
 
The shunt resistor provides extra current beyond with the regulator can supply.

Assuming everything's regulating properly and there's a sufficient load in place, there's 10V on one size of the resistor and 5V on the other, so at 3.9ohms, you've got an extra 1.25A being supplied to the load by the resistor, so the regulator isn't being stressed.

Great explanation. I never understood why the circuit was set up like that, but that makes sense. I assume the same explanation applies to the Atari circuit in the Sprint 2 example I posted above?
 
@TheShanMan - yes I know this is an old post but I am in the exact same boat with my Zero Hour restore. Getting 10V on the power supply. Was going to replace the regulator but maybe that is not the issue. What did you end up doing?

Yes, I have 6 other Universal games but they are all in storage right now so I don't want to pull them out just to see how their power supplies behave.
 
I ended up going with a switcher ultimately. I don't remember if I did so because of the above issue or something else. Perhaps I never settled on a good test load since I got no answer for that, and therefore never had the courage to hook up my board and cross my fingers that it would regulate to 5v properly.

The switcher has worked great, but I still have the original power supply mounted in my cab so if there's a good way to ensure that it'll regulate properly when hooking it up to the board, I could make use of the original finally. I guess figuring out a good test load could be done by measuring the current drawn by the board when hooked up to the switcher.

Please let me know if you get yours working!
 
You will need a load that pulls > 1.25A for the power supply to regulate properly. The resistor supplies the first 1.25A of current. Over 1.25A, the voltage drop across the resistor will be more than 5V so the regulator will start passing current to pull the output voltage back up to 5V. You need a load resistor that is less than 4ohms. A pair of 5W, 5ohm resistors in parallel would do. That will be 2.5ohms and pull 2A@5V (10W). Warning, they will get hot -- like burn your finger prints off hot -- if you leave them connected for any significant amount of time. A pair of 10W would be better.

 
  • Like
Reactions: DZA
Yeah that makes sense in light of what Mark said way back when. I'd still probably want to measure the current going from a switcher to the board just to be sure. The potential for overvoltage is a little concerning without knowing what the board actually draws first, to me at least.
 
Thanks all. Hmmm.

Most of all my games run the original power supplies when possible. So much so that I have re-built more than a dozen. So one part of me wants to fix this.

The other part of me knows if I fry a Zero Hour board, it is going to be a PAIN to fix it if I ever do and no one really repairs these.

I could leave all the original stuff in it, pull the 5V lines out of the connector and add the switcher on the side and just run the 5Vs to the board.

Gotta sleep on it.
 
Yeah me too. The only games I can think of that I don't have running original supplies are Mr. Do! and Crazy Climber (never was successful getting that one working either). I've been content though knowing that the original supply is in there and I could switch back if I ever want to. Good luck in your decision!
 
You will need a load that pulls > 1.25A for the power supply to regulate properly. The resistor supplies the first 1.25A of current. Over 1.25A, the voltage drop across the resistor will be more than 5V so the regulator will start passing current to pull the output voltage back up to 5V. You need a load resistor that is less than 4ohms. A pair of 5W, 5ohm resistors in parallel would do. That will be 2.5ohms and pull 2A@5V (10W). Warning, they will get hot -- like burn your finger prints off hot -- if you leave them connected for any significant amount of time. A pair of 10W would be better.

I'll at least get these 10W resistors and see what I measure. I'll go from there...

Thanks for doing the math.
 
Previous discussion with more detailed description of how this type of regulator circuit/mod works:


 
Back
Top Bottom