Negative Vreg questions..

I've rigged up a differential supply and all voltages are fine WITHOUT loads except the -18v supply. The diff voltages are fed +-30VDC, the 48v is through a doubler, so 60v through a floating lm317.

I have:

48v= 47.5
+18=17.99
-18=-27.24(edited, sorry!)
+15=14.99
-15=-14.95


the negative 18v tames under a 8k load or around 3ma of current drawn to keep it regulating..

the part is an st L7918cv.

Datasheet is sparse on info, all testing was done with 500ma..



anyone else see problems with STmicro parts?

the Moto 7915 regulates without a load though..

Is that negative regulator well decoupled ? For some reason those neg-regul. sometimes need a higher cap-value compared to the pos-regulators.

There was some discussion over here (or at the old place), a search may find you something more detailed.

regards,

Peter

yeah I figured that it might and definately used larger values with no real change..

Hmm. Never heard of the requirement of a minimum load for the 79XX series, although the bypassing is an issue as remarked.

Maybe the parts are defective---it wouldn't be the first time that something slipped through the test procedure net.

The one that floored me once was a bunch of LM358 op amps (kind of an off brand---GL358 was the P/N, I forget who that is/was) that weren't unity-gain stable!! Talk about things you come to take for granted.

The guy was using them in bootstrapped twin-T notch filters at very low frequencies. I assured him that he was doing something wrong when he told me the filters oscillated. Not so.

Yes minimum load requirements have been dicussed here due to a chap with the exact same symptoms. They need a few mils to get the transistors to wake up.

Watch your 79 series pinouts. They are different from 78's, as iZ found out not long ago. :sad:

They are different from 78's, as iZ found out not long ago.

Click to expand...

You're not alone... :evil:

Samuel

The '79 series of regulators are generally much weaker than the '78 series when it comes to both stability and protection. Only execption to this is the TS78xx/TS79xx series, where the '78 is not easy to work with.

The high positive voltage you have at the output of your '79 regulator must come from somewhere - and it's not from the regulator's input (I hope). These regulators lock up if they see a wrong-polarity voltage at their output when they start - often that is prevented by diodes antiparallel with the regulator - see the data sheets.

Try finding out where the positive voltage at the negative regulator's output comes from - and fix that.

Jakob E.

Oh oh oh oh, sorry sorry!

Jakob and all,

I forgot the - before the 27v, I was in a total rush last night and screwed that up, SORRY!

the regulator is getting -35v input and the output is -27vdc!

I'll edit that now.

however, I went back and tried a handful of the ST L7918cvs and they all exhibit the same symptoms. It's obvious they need a load of at least 3ma to regulate properly.

I've just never seen this before so it worried me a bit but after careful testing it seems they are fine, they just HAVE to have the load on them.

I've always used other brands before ST because I constantly have problems with their other types of regulators dying while running an idle load for no reason. But alas it was all I had laying around.

I'll be letting you know if I blow something up using these.. :green:

I've ALWAYS had that issue with 79xx parts, the 78xx parts will regulate fine with no load, but 79xx always show a way negative output until you load them slightly.

Doesnt seem to affect anything

I've been using On, TI, and Moto 79XX versions for some time with no need for output loading for proper operation. It seems to be just the ST parts that need the load.

thanks guys!

All those Italians went to a three-hour lunch that day and forgot to fix the problem ;-)

HA HA HA!
:green: