Three Terminal Regulators - 7905, 7508 and 7512.

[gbowne1]

I'm working on the power supply of my radio receiver project and looking at the 78xx series three terminal regulators in the TO-220 package. 

Any ideas on which ones work the best?  I want to keep the power input fairly clean.  Bought a Hammond transformer to take the 110-120v input from the computer type connector (IEC 60320-1 Type C13).  Found a bunch of Nichicon electrolytic caps, I'll use. 

The supply has to turn a bi-color LED from red into green when the unit is turned on (or comes from standby power).  I also bought a white/blue one for posterity. :-)

Greg

[Bored@Work]

Any ideas on which ones work the best?

Here we go again. There is no such thing as "the best". You don't need the afreaking best, you need good enough.

If you are really concerned about your power supply you are putting the cart before the horse with your search for "the best". Quantify how good the power needs to be for your circuit. Then you know how bad the power supply can be allowed to be (with a certain margin to stay on the safe side). Then check potential power supply candidates (of which the regulators are only a part) if they are good enough.

[FreeThinker]

If all you want to drive is an led then they will be fine, but I suspect you really want to do a little more perhaps power the device? In which case we need more info before we can make a judgement

[Kiriakos-GR]

I want to keep the power input fairly clean.

Greg

Greg did you made the wild thought to set those three in line as triple filtering ?

[ejeffrey]

The 78XX series are all pretty much the same, it is just a question of what voltage you need/want.  The 317 adjustable regulators have slightly better specs, especially if you add a bypass capacitor to the adjust pin, but the difference is unimportant in most cases: you pick the 317 because you need adjustable voltage.

At high frequency, none of the standard regulators do much at all.  They provide neither particularly good regulation, nor very good supply rejection.  At the high frequency, you are really depending on your input/output filters to provide the performance you need.  For instance, low ESR capacitors are key to providing good load regulation -- but watch out for stability if you switch to a LDO regulator.

However, what you need, or what is best for your application depends on your actual requirements.

Greg did you made the wild thought to set those three in line as triple filtering ?

I wouldn't recommend this, I expect the performance will be less than if you implement multi-stage RC filtering, but the idea is right.  It is hard to get tremendous rejection in a single stage, cascading filters generally improves your performance much more than building a single super-awesome filter. The reason I don't suggest chaining multiple 78XX regulators is that they are mostly all good at DC where it is easy to get really high supply rejection in a single stage.  Passive filters and capacitance multipliers are much better at high frequency.

[gbowne1]

Well, it needs to be adequately sheilded so that it keeps out of the receiver circuitry, and doesn't induce AC hum, eespecially into the 500kHz to 30MHz region.  I do also have some LM3xx regulators, but not as many in stock as I do the 78xx's.

Yeah it has to do more than just power the device.  Have to consider that there is a large GLCD 128x64 display with backlight. 

The receiver is designed for 6v DC, but the mains input is 110-120v AC.  Maybe by using a diode bridge rectifier.

It's also a fused input up to 250v 25A.  The power supply should be about 200-220w.. no more than 350w.

I figured i would let the power go down to 108v AC before it drops out. and a surge of no more than 121-122v AC.  We get more spikes than brownouts around here.. tho area pole transformers and wires are aging.

I think I've got a few supply circuits lined up that could do this but i'd like to hear your ideas before I trudge ahead.

I'd also thought about buying a TDK-Lambda power supply module/brick.

[Kiriakos-GR]

Ok got ya , SSB modulation etc etc ...   

I just wish you good luck.

[gbowne1]

Yeah it has SSB.. LSB, USB.  The original receiver circuit used AA cells and some C cells, or altenately a 6v wall wart AC adapter.  I don't really like wall warts.  I made some upgrades to the receiver circuit and did some breadboard R&D using a simple power supply so I could build and test each stage till it was working optimally.  Mind you the receiver.design is 20-22 years old.

[ejeffrey]

Well, it needs to be adequately sheilded so that it keeps out of the receiver circuitry, and doesn't induce AC hum, eespecially into the 500kHz to 30MHz region.  I do also have some LM3xx regulators, but not as many in stock as I do the 78xx's.

Passive filters are the way to go here, but what you really need is a specified level of noise and ideally a way to test it.

Yeah it has to do more than just power the device.  Have to consider that there is a large GLCD 128x64 display with backlight. 

Step 1 is to have separate analog and digital supplies.  With a linear supply you would usually derive them from the rectified transformer output, but have separate filter and regulator sections.

It's also a fused input up to 250v 25A.  The power supply should be about 200-220w.. no more than 350w.

  I assume this is a typo since you were running on a 6V wall wart or some alkaline batteries before. 

I'd also thought about buying a TDK-Lambda power supply module/brick.

I strongly recommend this.  A supply module will have the appropriate safety, fusing, and filtering requirements for connecting to mains power.  You may still need to add RF filtering and post-regulation, but best to start from a safe low-voltage DC supply.

[gbowne1]

Yeah, I thought it would be better to have a modular power supply I could swap out in case of failure.

Yeah, that's a typo.  Was working on some other calculations.

Ok, so ideas on where to go from here? 

[Zad]

With RF systems in general and any other noise-sensitive systems, I tend to put a voltage regulator(s) in each module, each with their own filtering. They really do cost pennies. Power spike protection generally goes on the HV side of the transformer, but a choke in the output and some local filtering on each module does no harm.

[Harvs]

This is actually one I've read about (not in terms of RF, but audio freq stuff for input stages.)  Unfortunatly it was too long ago to give you any definate answers, but i can give you a few tips in your search for info.

What you are trying to do is minimize the output impeadance of the regulator ciruit at all frequncies of load, and ensuring the regulator remains stable under all conditions.

Below is a refence that seems to cover most topologies and gives good discussion.
http://www.onsemi.com/pub_link/Collateral/HB206-D.PDF

However do some Googling of lowering linear regulator output impedance (most info will probably be from the audio world.) And also read up on your regulator stability...