5V or 12V TPS?

In article , Clint Sharp=20 says...>=20

From the very same website...

Linear regulators

A dangerous precedent was established by the first linear regulator=20 semiconductors sold commercially like the LM7805 type devices: they=20 require no input or output capacitor and are completely stable under=20 virtually any operating conditions. Some of the newer LDO regulators=20 require careful attention to external capacitors to operate in a stable=20 mode. The main reason the 7805 is unconditionally stable is because the=20 power pass transistor was made up of an NPN Darlington as shown in=20 Figure 4. They are usually referred to as ?NPN regulators? for this=20 reason, and are comprised of an error amplifier, voltage reference, and=20 NPN Darlington power transistor.

Note the words "unconditionally stable" and "completely stable".

--=20 Conor.

"Some of you may be anxious about finding a new job, or a new place to=20 live. I know how you feel." President Bush, 2008

In article , Conor says...>

Just to add, that was written by the very same person who wrote the PDF you linked to.

Just a note...you don't need the capacitors on an NPN but you do on a LDO. Your typical LM7805 is a NPN.

And the first PDF mentions MC7800's which are a variant, specifically that they're equipped with thermal overload. It also mentions that they're only really needed in high current and high frequency applications, to provide stability, not to make them work.

Next....

In message , Mark writes

Admittedly it's not likely to happen on a mains powered board but it's not for reverse polarity protection in that way anyway.

Basically, the reason for the diode across the regulator input/output;

If the input voltage drops below the output voltage of the regulator (during cranking perhaps, hence the battery/wiring shot comment) and there's a charged cap on the output, the cap can dump its charge back through the regulator damaging it in the process.

When the input voltage comes back to spec the now faulty regulator could potentially pass the unregulated input voltage to the load.

Put a diode anode to the regulator output and cathode to the input and your regulator will survive, so will its load.

In article , Clint Sharp says...>

...the cap which you don't need on a NPN regulator and I wouldn't have put in?

Look, I'd be unlikely to put one in on a circuit where the load consists solely of a voltage divider.

What I was remarking on was your apparant belief that Clint was the one without a clue, when you are the one who exclaimed "Bypass cap? WTF is that?".

In message , Conor writes

Even the company that says they don't oscillate recommends you fit them for other reasons but the point is, some manufacturer's parts *do* oscillate and you cannot guarantee that you will always have a nice National semiconductor part available.

Some of the nasty no name ones you can buy will oscillate with no load if you're unlucky in the way you lay it out.

I've seen one 'brand' oscillate with only a power indicator LED on the output and others in truck 'droppers' when used to drive BFO TO-3 transistors. Both cured by replacing or fitting the recommended caps respectively.

I've seen 250mV of noise on a power rail 'disappear' (according to the Tek 465B I was using at the time) when the caps were fitted.

If you're designing something you'll probably get away with it 99 times out of a hundred, probably 999/1000 but long and short of it, fit the caps, fit the diode, it's good advice, cheap insurance (under a quid for all the parts) and you're unlikely to have weird problems caused by noisy power.

FWIW, they oscillate in the 40-60MHz region when they do oscillate depending on the length of cables, circuit board layout, I've measured it.

In article , Clint Sharp says...>

Ah, that explains why I've not come across it. I've not really worked with anything up to that involving them.

Perfect example just today. An elderly DAB tuner which showed several problems - including dying when the KBPS changed on R4, intermittent display, and ending up not working at all. But always ok at switch on from cold.

Two rail power supply - 5 and 12 volts using 78 series regs. Both measured fine. Just about to give up as it's double sided board and mainly surface mount. But put the scope across the power supplies in turn to find the 5 volt one very dirty. Clean at switch on. Replacing the 100µF cap on the output sorted it.

Oscillating by how much? I built an oscilloscope calibrator using one once. I think I'd have noticed if it was. A few millivolts here and there isn't going to make much difference ITRW. Capacitors wouldn't stop the regulator oscillating - they'd just smooth the output, surely?

The only results I can find mentioning oscillation involve using them in a circuit with high frequencies. You're not going to get that in a car.

FYI, From National Semiconductor:

By Chester Simpson, Member of Technical Staff, Power Supply Design Group

Linear regulators

A dangerous precedent was established by the first linear regulator semiconductors sold commercially like the LM7805 type devices: they require no input or output capacitor and are completely stable under virtually any operating conditions. Some of the newer LDO regulators require careful attention to external capacitors to operate in a stable mode. The main reason the 7805 is unconditionally stable is because the power pass transistor was made up of an NPN Darlington as shown in Figure 4. They are usually referred to as ?NPN regulators? for this reason, and are comprised of an error amplifier, voltage reference, and NPN Darlington power transistor.

On the one I mentioned earlier it was about 100 mV of 'noise' - the scope wouldn't trigger so fair to assume made up of lots of different frequencies. That was more than enough to completely f**k up the tuner. A new cap sorted it.

Could you explain? They're normally fed from a DC supply which should be filtered at the input.