Split Power with 2 7805's

[Littlebilly91]

I am making a split power supply with 2 7805's (see pic). Is there a reason that instead of getting an even -5, 0, +5 split, I get -6, 0, +5? If I put an additional regulator in parallel with the first (leftmost) one, I get a five volt difference (instead of 6), but I have read things saying not to do that.

Thanks,
Chet

[Terry McGee]

Hi Chet

Doesn't look like a legitimate use of the devices to me. It might just work under certain load conditions, but couldn't be relied upon. I think you might have to slip into the village for a 7905.

Hmmm, even then you may be in trouble. 12V is probably not enough to yield 5-0-5V - it probably won't leave enough voltage drop across the regulators (check their specs, I'm relying on memory here.)

Terry

[Elvee]

Originally Posted by Littlebilly91

I am making a split power supply with 2 7805's (see pic). Is there a reason that instead of getting an even -5, 0, +5 split, I get -6, 0, +5? If I put an additional regulator in parallel with the first (leftmost) one, I get a five volt difference (instead of 6), but I have read things saying not to do that.

Thanks,
Chet

Two problems there:

-For the circuit to work properly, the ground must always source current. It is a possibility, it depends on the kind of load the circuit is used with.
If the condition isn't met, the ground will wander unpredictably.
In fact, it will probably be already the case open-circuit: the quiescent current of the upper regulator will upset the lower one.
A bleeder resistor could be connected at the output of the lower regulator to make sure it is always sourcing current.

-The upper regulator only has a 2V regulation margin, which is exactly equal to the drop out voltage for the 78xx series.
It might work well at low currents, but at higher loads there is no guarantee.
A low drop out regulator might be a possibility, but since the ground current can be very high, this could cause more problems in relation with the first issue.

[Littlebilly91]

Originally Posted by Terry McGee

Doesn't look like a legitimate use of the devices to me. It might just work under certain load conditions, but couldn't be relied upon. I think you might have to slip into the village for a 7905.

Hmmm, even then you may be in trouble. 12V is probably not enough to yield 5-0-5V - it probably won't leave enough voltage drop across the regulators (check their specs, I'm relying on memory here.)

Terry

I had some trouble with the 7905 in there, too. That was my first idea. Another uneven split.

My load current should never be too large. I am powering two op amps, two comparator LM339 chips, and some LEDs. Would this cause a problem?

The reason that I regarded this as a problem was because I was getting a DC bias out of my voltage follower by about 1 volt. It may be a problem elsewhere, but I figured that the uneven rails was a likely source of that problem.

[Littlebilly91]

Originally Posted by Elvee

Two problems there:

A bleeder resistor could be connected at the output of the lower regulator to make sure it is always sourcing current.

Ok, I tried the bleeder resistance. It works brilliantly. I am using 100 Ohms. Large resistances (+10k) don't seem to work so well. Thank you very much.