These 6-pin switching power regulators in a TSOP package deserve mention IMO: Part# ADP2301.

I currently use Microchip's MCP-Series of Linear Voltage Regulatorsfor low voltage output .8-5V (<6VDC input) ~ 1Amp regulation for the simplicity of the circuit, i.e. low parts count. The drawback is the input voltage has to be relatively close to the desired output voltage.

The Analog part has an input voltage of 3-20V, but requires a few more LRC connections. Also it can drive up to 1.2Amps...

Always a tradeoff, but the Analog switching IC is very impressive.

_________________
65Org16:https://github.com/ElEctric-EyE/verilog-6502

From the end of the Non-typical power-supply circuits section of the circuit potpourri page of my 6502 primer:

Switching regulators (which use inductors), whether buck or boost, are typically a better way than charge pumps to do power supplies, due to efficiency, particularly when regulation is needed; but they tend to do poorly on breadboards, and their PC-board layout is not for beginners. (The one I've used most in our commercial products is the MAX732 which allows you to start with anywhere from 4V to 9.3V input and get 12V out. Another is the MAX669 which let us start with a pair of AA batteries and get 12V out.) One way around the difficulty that switching regulators present for the hobbyist is to use integrated switching regulators which put all the difficult stuff in a small pre-made module, sometimes with the same pinout as the popular 7805 for example. Power Trends (now unfortunately taken over by TI) is one supplier whose 78SR112 module I designed into a product in 1993 and they continue to have a wide range of offerings. Pololu Robotics and Electronics and MicroPower Direct are two other such suppliers.

Getting good, quiet behavior out of an IC like you mention won't be easy for the hobbyist-- although it may not be as important for a digital circuit where you don't plan to ever add analog circuitry like A/D and D/A converters. If you do want to try it, make the connections to the VIN and BST capacitors and the diode absolutely as short as you can, and keep the connection from the inductor to the VOUT capacitor short.


That's only to hold the dissipation down at high currents. If the current is low enough, you can have maximum input voltage and minimum output voltage at the same time. "Dropout" voltage is where it begins to lose regulation, ie, there's not enough difference for it to be able to do its job; so the input voltage needs to be higher than the output voltage by at least the dropout voltage to maintain regulation.

LDO regulators can go unstable (oscillate) under certain conditions. There was an article in Electronic Design, Nov 4, 1996, starting on page 99, on keeping LDOs stable, similar to dealing with the phase margin on op amps and watching your poles and zeros in a feedback system. Since the details are well beyond a lot of users' understanding, LDO manufacturers normally recommend types and values of output capacitor to keep the LDO happy since it can cause trouble to just put any ol' capacitors on there. If your input is always at least about 1.75V above the output and the dissipation of a linear regulator is ok, you can always use something like the LM317T instead of an LDO, and you can be sure that it won't go unstable.

_________________
http://WilsonMinesCo.com/ lots of 6502 resources
The "second front page" is http://wilsonminesco.com/links.html .
What's an additional VIA among friends, anyhow?

I did not mention a link to the project I was referring to in my original post. Here it is. They do provide gerber files in addition to the schematics for board layout which utilizes 2 of the switching regulators I mentioned, so one could see how the Analog engineers made it work.

Great point about the special needs of an analog power supply...

Can you comment on the 2 parts Analog has: The ADP2300 and ADP2301. One has a switching frequency @ 700kHz and the other @ 1.4MHz. What the purpose of these 2 frequencies? and which would be better, although maybe not optimally suited for audio IC's?

BTW, very nice write up on Non-Typical Power Supply Circuits on your 6502 primer page! I just checked it out.
I've seen more simple schematics to use diodes in a similar fashion for voltage doubling, but it was not explained nearly as succinctly as your site.

_________________
65Org16:https://github.com/ElEctric-EyE/verilog-6502

It looks like you have to register to see it (which I rather not do). Did I miss something?


It's about LDO (low-dropout) regulators in particular, not normal NPN ones which require more input-output differential but whose configuration does not have such potential to oscillate.


The higher the frequency, the smaller the inductors and capacitors you can get away with. That's valuable in things that have to be super small, but probably not an issue for the kind of stuff we're doing. However higher frequencies also make it more of a challenge to get the layout right and get good performance though, and that's all the more true with higher current.

The first switching regulator I used was the MAX732 which uses PWM at only 170kHz. This was in 1994 and we had thru-hole assembly equipment, so I designed the product to use the 8-pin DIP, and since it was for audio and music, we needed it much quieter than the spec.s say you can get. It wasn't easy. Although you can't hear 170kHz, excess inductance in certain connections made it produce sub-harmonic noise which definitely was audible. We met the goal though. Later, after a couple of years of production, Maxim changed the wafer process and made the rise times faster, and then we had to modify the circuit again because it was producing RF noise that was getting into the aircraft radios. I used the same MAX732 on other products after that, but applied what I learned to subsequent designs so they were very quiet. The rise-time issue is related to why Dr. Howard Johnson said he would like manufacturers to specify a minimum rise time also, not just a maximum, IOW, that they won't let it get so much faster than it needs to be that it causes problems.


Thanks. I would like to give more info on them, but it gets overwhelming. For a given application, you'll probably have to breadboard it and take a lot of measurements to see if you're going to meet the goals, before laying out a board.

_________________
http://WilsonMinesCo.com/ lots of 6502 resources
The "second front page" is http://wilsonminesco.com/links.html .
What's an additional VIA among friends, anyhow?

I don't know, I can't view the gerber or board layout files presently, just schematics and BOM.

Wasn't the primary advantage of a switching power supply the availability of much more current than a linear power supply?

I started this thread on these switching power regulators because I was impressed at the low parts count for a switching power supply, however the current is not that much above the LDO regulators, only 200mA... When I look at the switching power supply I use from an old PC that says 30Amps @5VDC, that's impressive for it's size compared to linear which would require a huge step-down transformer only for 10 Amps.

The fact that they can allow a greater input voltage is not a reason to use them. I think that's what you were saying when you suggested an LM317.

Maybe I was impressed on the progress of these devices.

_________________
65Org16:https://github.com/ElEctric-EyE/verilog-6502

You probably know most of this stuff, but I'll post it for the benefit of others too.


A primary advantage is efficiency, and that can also mean greater current or smaller size when you don't have to get rid of much heat in the process. Heat sinks take room. In a "buck" configuration that steps the voltage down, the output current may actually be higher than the input current, as it acts kind of like a gear ratio which exchanges torque for speed from one end to the other. Related to that, another advantage is that a switcher can, in a "boost" configuration, be used to step the voltage up (unlike linear regulators).


The parts count for something like a 7805 regulator is hard to beat-- just a .1uF capacitor at the input and one at the output. Integrated switching regulators like I linked to have all the parts on a single hybrid module, relieving you of the design and assembly work.


The tiny switcher ICs with integrated switching MOSFETs are pretty limited in current. The ones that can do the higher powers will require external MOSFETs, adding to the parts count and design complexity. They definitely exist though.


The PC power supplies start the switching business at the AC mains, instead of first using a step-down transformer and rectifier and smoothing capacitor circuit as a starting point and then regulating. Myself, I have not looked into how to design that kind of thing.


I'm not sure which way you mean here. The LM317 is an easy solution for situations where:

• the input voltage is in the range of 1.75V to 40V above the desired output voltage, and
• the heat produced by throttling the output to the desired output voltage is acceptable (the operation is basically like dropping the voltage by way of a resistor that is constantly adjusted to give the right output), and
• you have power available to burn, which may exclude battery-powered situations. I have used linear regulators (LP2940 and LP2950) in battery-powered situations where the current used by the regulated load is a small percentage of the overall load on the battery, so using a linear regulator did not impact battery life significantly.

If you can go with a standard output voltage like 5V, 12V, or 15V, the 78xx (or LM340 series which is somewhat improved over the 78xx) family reduces the external parts down to just an input and an output capacitor. For low-power situations, the 78Lxx, LM317L, and LP2950 regulators are available in TO-92 or even smaller packages.


Some consumer items are driven heavily by cost and size (including being as low-power as possible to get the battery size and weight down). The market of the two companies I've worked for since 1985 is different, and although I've benefited from various new parts on the market, it can also be very frustrating finding a fit when so many of the parts are not compatible with legacy equipment we have to interface to in the aircraft. Hobby stuff is a mix, but I realize you're going for the low-voltage FPGA stuff.

For convenient, inexpensive, complete fixed-voltage power supplies, even Jameco has a lot of attractive offerings in wall-wart power supplies with switching regulators in them, and for more power, table-top supplies.

_________________
http://WilsonMinesCo.com/ lots of 6502 resources
The "second front page" is http://wilsonminesco.com/links.html .
What's an additional VIA among friends, anyhow?

Great info Thanks! Give me some time to consider...

_________________
65Org16:https://github.com/ElEctric-EyE/verilog-6502