[sdiy] 7915 regulators

[Seb Francis]

Jay Schwichtenberg wrote:

> In my experience the pin outs are the same for the same type. But the pin
> outs are different between the 78XX/79XX. Check that out.

Yep, and more annoyingly the metal tab is not connected to ground on the 79XX, so you have to insulate-mount to share the same heatsink with 78XX.

> Yes put a small
> load on them for testing. Some will work others won't without a load. Also
> with 30 volts in put a big heat sink on it or a small (in resistance, not
> current) power resistor in series to dissipate some of the power going into
> the regulator.

The transformer is rated 20VAC 1Amp, so I'm assuming the voltage across the smoothing cap will drop under load.  There's a 15VAC tap on the transformer, but I figured this might be too low under full load.  In any case, this is only a bench power supply for powering my breadboard, etc.. so the duty cycle should be quite low.

>
> I read a lot on 78XX/79XX regulators and gave up using them on anything that
> is remotely serious. There was an article on power supplies for audiophiles
> in Audio Amateur by Walt Jung and Jan Didden (sp?) that made me give up on
> them. Basically it said the 78/79 were junk, LM317/LM337 were better and
> LT1086/LT1175 were tolerable in audio. They were looking at ripple rejection
> and noise. Each set was about -10 db quieter than the others. Anymore I use
> LM317/LM337 for misc things and the Linear Tech parts for audio.

I tend to agree with you (going by what I have read).  In my modular PSU I'm using LM723s with power transistors to get the current.  I did try measuring the noise amplitude of the 7815 vs LM723 with a digital scope, and actually it seemed pretty similar .. but I think it was right on the edge of the scope's resolution at that voltage offset.  My cheap+easy 3 rail bench-PSU based on 7XXX seems ok though .. now that I know I have to put a bit of load across the 79XX to get the right voltage.

Cheers,
Seb