68 ohm resistor across 7805 regulator in early Atari 2600. Why?

[Yoshi]

I've been working on an early revision Atari 2600 recently, and found a 2W 68 ohm(measured 70) connected across the 7805's input and output, but the rest of the circuit was what I expect for an old linear regulator circuit.

I was wondering why they would do that from a design standpoint. I've checked the service manual and it shows a 2W 680 ohm resistor, so that's even more confusing! Never seen that before.

I'm a second/third year Electronics Engineering Technology student, and I repair/modify old game consoles. I've never seen that in any of the other consoles I've worked on before, and none of my textbooks show that, and it just goes against what I've learned about 78xx regulators up to this point.

When I went to mod it, part of the 2600RGB mod I'm doing is replacing the 7805 with a modern switching regulator, and when testing the regulator on the switchboard before reattaching the motherboard I found that the regulator didn't work properly, and I got about 0.7V lower on my output than what I put in on my scope when it was supposed to be outputting 5v. I found that by removing the 70 ohm "short" resistor, the regulator than starting working properly and I got 5V. So what was the point of it being there in the first place?

[tszaboo]

Hehe. The 2600 had a 9V 500mA wall wart, at maximum load, the 7805 would dissipate 2W. This was obviusly too much to handle without SMD mounting and using the PCB as heatsink. And they did not wanted to add a heatsink, as it is made of metal, so it is very very expensive. As far as I remember, it has a screw mounting it on the PCB, so realistically, you only have 1-1.5W maximum dissipation.

So I guess some engineer measured, that the minimum current used by the machine is some 60+mA. What happens, there is always 5V on the regulator output, there is always 60mA flowing through the resistor, and that 0.25W is dissipated in the resistor, not the 7805, so they can get away without a heatsink.

[SingedFingers]

Cheapness measure. They do it in my frequency counter:



Passes way over the current of a single 7805 can handle without them having to invest in a TO-3 canned regulator.

Unfortunately it does this to the output (bottom is 5v...)





cheap asses 

[David Hess]

It is a cheap way to raise the available regulated power without a more powerful regulator when the output current has a minimum level.  The resistor is providing some of the load which would otherwise have to be supplied by the regulator.

[2N3055]

Offloading some of the thermal dissipation to resistor.. 

[Yoshi]

Thanks guys! I figured it had some sort of purpose. I take it this isn't exactly a good design practice?

[SingedFingers]

Thanks guys! I figured it had some sort of purpose. I take it this isn't exactly a good design practice?

If you draw a Venn diagram of "cheap", "fast" and "quality", this solution is grazing on some grass somewhere a few miles away.

It works, until it doesn't, and then it gives me a headache

[2N3055]

Thanks guys! I figured it had some sort of purpose. I take it this isn't exactly a good design practice?

If you can guarantee that current consumption doesn't fall below certain point ( so the regulator doesn't fall out of regulation) it is actually quite ok to decrease thermal stress on regulator. If resistor is rated for dissipation, it can be very reliable.

[james_s]

The HV board in the Amplifone vector monitor Atari used in Star Wars has resistors across the 7824 and 7924 regulators for this same purpose. The earlier monitors did not have this and it was found the regulators would overload during the deathstar explosion scene.

[David Hess]

Thanks guys! I figured it had some sort of purpose. I take it this isn't exactly a good design practice?

There is nothing wrong with this.  It just has limited applicability.