Quick question - voltage regulators in parallel?

Nope. I would definitely not do that. They never track each other exactly.

--
Regards, Joerg

http://www.analogconsultants.com

Alright, thanks Joerg. Do you know of any 2A-rated regulators like the trusty 7805 series?

There is a version up to 3A but I don't know the part number.

- Henry

--

"lain" schrieb im Newsbeitrag news: snipped-for-privacy@16g2000cwy.googlegroups.com...

There used to be 5A versions but they might have been obsoleted. There is no big market for 5V anymore and today everyone uses switchers. When I designed big 5V supplies in the old days I did it in discretes since the higher power regulators were hard to get and quite expensive.

--
Regards, Joerg

http://www.analogconsultants.com

They might argue, but try putting a small R, say 0R2 in series with the o/p of each regulator, it might even the balance, maybe, possibly, if they are from the same batch.

Of course this will affect the Vout, it depends on your requirements

But check that they have the same o/p voltage at the same load first

martin

The 78H05 was the 5A version but AFAIK it's long since obsolete. For a one-off hobby project it may be possible to find one somewhere.

--
Regards, Joerg

http://www.analogconsultants.com

"martin griffith" schrieb im Newsbeitrag news: snipped-for-privacy@4ax.com...

There is even an option to use a 7805 with a PNP transistor and the 7805 does regulation where the PNP powers. If the OP goes behind the 78xx part number there is surely enough equivalent parts around with higher current spec.

And there is a 7805-like switcher chip made by TI.

- Henry

--

Unfortunately if this works I'll be making a lot of them, so the one-off solution won't work.. however after digging around the datasheet for this 7805 I see it says it's fine to go over 1A so long as you provide adequate heatsinking... so I'll just slap a nice heatsink on it & some thermal paste and hope for the best...

- Eric A.

Well, not a great idea.

What's going to happen is one of them will have a setpoint a few millivolts higher than the other one. The one with the higher setpoint will try to pull the voltage up to its desired level. Since it's unlikely to put out 2 amps, it will go into current limiting and effectively become just a say 1.25 amp current source.

Then the one with the lower setpoint will become the voltage regulator, putting out 0.75 amps at it's setpoint voltage.

This might be a stable situation, for small values of the word "stable".

All of course in theory. In reality they might oscillate, or the higher one might go into a full thermal shutdown. Some of them have a designed in thermal hysteresis-- so it might go into wild on-off cycles.

There is an adjustable one, I think it's the LM338T

You will need an extra two resistors, and I suggest metal film, 1% or better.

Chris

Why not look into the PTH series from TI? Switching regulator modules in a variety of packages and input voltages. Plus they don't heat up that much.

Awesome, looks like that would be my best option, as I will also be producing a few other voltages.. now for another quick question, I'll have 12volts coming in to the circuit, but it's going to run from a car so I'm expecting a LOT of noise on the line. Would, say, five 6800uF Electrolytic capacitors in parallel with the incoming 12V (about 0.034 Farad) work well to filter it? I've been playing with the idea of using "super capacitors", but they're pretty pricey and because most of them are rated around 2.5volts that I can find, I'd have to use them in series and though the overall capacitance would be higher by a decimal place or so, I understand putting caps in series should be avoided.. Any thoughts from you experts out there? Thank again!!

-Eric A.

I remember the PNP trick, but couldnt remember which data sheet it was is, prolly Natsemi, a suitable place to retire old 2N3055s' maybe

martin

regulator modules

Ohhh neat, never seen these before. But! - I'm on an extremely tight budget and these linear regulators are only about $1 each, whereas the PTH-series seems to run about $11 to $30 each (mouser.com being my source), but I've got 'em bookmarked for future projects now, thanks!

-Eric A.

You can use the extra series resistors for a quick test / prototype, but for longer term you might as well get some 2A 5V regulators:

Also possible: use a high-current bypass transistor as shown in Figure

14 here: You will need a pnp transistor that can handle the extra 1 Amp.

Mark

Not gonna work for you. Your best approach is the one redbelly gave in his response: the 7805 with an MJ2955 PNP pass transistor - use a heat sink.

30 amps with a linear is possible with this method, but it gets physically large - you'd need a BIG heatsink and at least 4 pass transistors to be safe.

You mentioned an automotive environment & big caps. Forget massive capacitance. Someone mentioned LC. I'll add this - use a TVS diode like a P6KE24 to ground with a voltage dropping resistor. The general idea is like this:

+12 ---[L]---+----[1R]---+---[V/R]---[Load]---+ | | | [C] [TVS] | | | | Gnd ---------+-----------+--------------------+

The LC acts to limit transients. The 1 ohm 25W power resistor will drop 2 volts taking a little heat away from the regulator under normal conditions. If a transient gets past the LC and the TVS conducts, the power resistor will drop the voltage and dissipate power so that the regulator & the TVS are protected.

Ed

regulator modules

Eh, heatsinks aren't exactly cheap either. Unless you need the low noise of a linear, I'd go with a buck converter.

I remember building a switcher with a 7805, a transistor (PNP), coil and diode. Worked reasonably well for what it was. I think it was the 'current boost' circuit with the coil between PNP collector and 7805 output, and the diode to prevent the collector going negative.

Thomas

I vote for this one - I've done it, and it works like a champ. :-)

And don't forget the capacitors!

Cheers! Rich