5vdc power supply

jims

Senior Member
Picaxe power supply question. I have powered all of my previous projests from either 3x AA alkaline batteries in series; or a 9vdc wall wart through an off-the-shelf “buck boost” module. These have worked reliably with no problems. In order to move-on I have made a 5vdc regulator using the attached circuit from the Picaxe manual . I am using a 9vdc 300 mA wall wart. The regulator chip is an LM7805C; and the supply output is 5.6vdc (with a load consisting of a yellow Led & 330 ohm resistor ).

Is this typical?? Is it safe to use this supply to power my project (project draws about 120 mA and consists of an 08M2; a Picaxe 133Y serial OLED display; a PIR; and a Picaxe ERF module )?
Thank you, JimS5vdc supply.JPG
 

Dippy

Moderator
I haven't used one of those 'old faithfuls' for years but I seem to recall much better results than that.
Typically within 0.1 or 0.2 volts from memory with maybe 10 to 200mA load.

Is it a proper one or an Ebay 'special'?
i.e. what manufacturer? Have you checked manufacturer's data sheet?
 

Reloadron

Senior Member
I don't quite understand the 5.6 VDC output? The LM7805C should regulate at 5 Volts +/- 0.2 Volt so the output should range between 4.8 Min and 5.2 Maximum and most I have worked with do much better than that. You mention the LM7805C which is a 1 Amp regulator with proper heat sinking and your supply max is 300 mA so that should be fine. Without knowing the LED forward voltage & current the Yellow LED with a 330 Ohm resistor really doesn't tell much. The 7805C input voltage must be at least 7 Volts because the LM7805C has a 2 Volt dropout voltage. In the attached drawing they used a 78L05 which is a lower current (100 mA) 5 Volt regulator and placed a 1N4001 diode in the line. I can only assume or guess the diode is in there to provide input protection and it will drop about 0.7 Volts if the circuit were connected reverse polarity the diode would protect the regulator input. Anyway, I don't understand the higher output voltage you are seeing, I would not expect a voltage exceeding 5.2 volts maximum. Make sure the regulator input voltage exceeds 7 volts.

Ron
 

Technical

Technical Support
Staff member
5.6V would imply a diode in the path of the 0V pin of the regulator, giving an extra 0.6V more than expected? We've done exactly that with 2 diodes to give 13.2V from a 12V regulator in the past.
 

hippy

Technical Support
Staff member
It is surprising the output is so high at 5.6V. The only time I have seen 5.6V or higher is when using the trick of putting a diode from regulator 0V with cathode to 0V to deliberately up the output voltage.
 

jims

Senior Member
Thank you to each of you for your help. I replaced the Fairchild LM7805C chip and now have 4.98VDC. Must have fried the chip and let the "smoke" escape. I'll have to be more careful with the soldering next time. JimSnext
 

srnet

Senior Member
Jims,
Why a 9v wall wart plus a 5V regulator? Why not just a regulated 5V SMPS wall wart?

Ken
A linear supply will (normally) be a lot quieter and have less supply noise than a SMPS supply.

SMPS supplies can cause problems with some circuits.
 

Goeytex

Senior Member
Sometimes I use an LP2950 or MCP1702 dedicated to the Picaxe and something like an LD50V dedicated to an LCD module. I may have 2 to 3 regulators on a PCB or breadboard each providing local regulation for a device or group of devices. I also like to use a separate regulator for RF transceiver modules, rather than use the same supply that the Picaxe is using.

Sometimes hanging one regulator on a board to supply everything doesn't always work out too well. Especially true on breadboards with lots of stray capacitance, high resistance contacts, and wires snaking all over the place.
 

Dippy

Moderator
Circuit, depending on the app I tend to go for LP2950, LP2980, LM2940, LM2936.
Some others too including something from Torex which I can't remember the number.
Remember to be aware of the max Vin rating too, so always read Data Sheet.

There's a lot of things to consider when designing so don't work 'blind' and don't simply cut'n'paste without some thought.
 

Circuit

Senior Member
Sometimes I use an LP2950 or MCP1702 dedicated to the Picaxe and something like an LD50V dedicated to an LCD module. I may have 2 to 3 regulators on a PCB or breadboard each providing local regulation for a device or group of devices. I also like to use a separate regulator for RF transceiver modules, rather than use the same supply that the Picaxe is using.

Sometimes hanging one regulator on a board to supply everything doesn't always work out too well. Especially true on breadboards with lots of stray capacitance, high resistance contacts, and wires snaking all over the place.
Dippy, Goey, thanks, informative and interesting.

Goey, I notice that both Digi and RF systems place quite a range of decoupling/bypass caps on the power lines on their demo boards (for xbee and for Zulu respectively). Digi stick 100uF; 10uF; 100nF; 47pF and 8.2pF on theirs. RF systems place 4.7uF; 100nF; 47pF and 47nF onto the VCC on theirs - they use a LM317 as the regulator. I have seen several references to the LM317 as being particularly stable in its output. Like you, when I am using a transceiver, I use a separate regulator for the transceiver supply and at the moment I am using the LM317. I particularly like this regulator because it is so easy to adapt to produce any of the voltages that I commonly require; I just keep a boxful of them.
 

Goeytex

Senior Member
The LM317 is a great regulator and I also have quite a few handy. But it is an older device and even though sometimes touted as an LDO, it cannot compare to some of the more modern devices. It is quite inefifcient by modern standards.

Admittedly, since I have (semi) retired I have not kept up with many of the newer offerings. In my previous positions I always seem to have samples available for new products as they came out. Many of these sources have dried up a bit as vendors generally only throw samples where they can see the potential for larger volumes in future orders. Such is the business.

In looking at only TI offerings, I see several newer products that can outperform the LM317 in most areas and with a much lower dropout voltage. Some are even specifically recommended for RF devices, such as the TPS7A4501-SP. But I do not have the time or inclination to try every new thing when what I have handy is "good enough". However, if I were still designing commercially and trying to stay on the leading edge, I would certainly be looking for and trying out the newer stuff.
 

Dippy

Moderator
It's not that certain regulators are 'bad' or 'wrong', it's just that designers (proper EEs) will choose a particular device as it is more apropriate to the circuit and application.
There are, as you will have seen in Data Sheets, a whole lump of parameters that are there for consideration.

Some of these choices may be based on personal experience as well as Data Sheet specification. Cost can play a part in many cases - often causing accountants and PMs to make strange noises :). (I speak from experience).

Certain types of circuit (e.g. digital or RF) require extra care. Extra bypassing/decoupling with specific values/types of capacitor are usually required to ensure optimum perfomance.
A lot of this is down to the REAL performance of capacitors. Sometimes, the paralleling of caps is used.
This aspect and physical circuit layout is whole separate subject which I'll leave.
Ready-made modules will tend to have already done all or most of this for us already.

Many demo boards will also have done this for you, but they are not designed with your application in mind. So, thanks to the accountant, they may bung on a cheap but adequate regulator which may not suit your micro-power battery application. In many cases, it is up to you (the EE designer) to take knowledge from the reference design and adapt it to your requirements.
 

wapo54001

Senior Member
I read about the 'newer' regulators with interest since I've been using the LM317 for almost everything. I was looking at the ld50v but was put off by the maximum input voltage of 10 volts which is not safe for connecting to something like a wall wart. Looking further on the TI website I came across the TL780 series which has the 1~2% output tolerance and good regulation numbers, and wondered if there was a reason it was not included as a good regulator for Picaxe applications? I also like the price -- 88 cents in single quantity -- which is not far above other regulators with lesser specs.

One thing I don't like about using the TO-220 is how far it stands above the pcb when using it for applications where some higher current capability is desirable but only at around 100 milliamps. I wonder how much current the KTT package would handle if used standing up as a through-hole device.
 

rq3

Senior Member
I'm also partial to TI, specifically their TL720M05-Q1. It tolerates over 40 volts on it's input, and is over-temp, over-current, and reverse polarity protected.
Drawbacks are that it's surface mount only (though very easy to hand solder), and needs at least 22 uF (preferably high quality tantalum) on the output.

Rip
 

Goeytex

Senior Member
... Looking further on the TI website I came across the TL780 series which has the 1~2% output tolerance and good regulation numbers, and wondered if there was a reason it was not included as a good regulator for Picaxe applications?
Likely because it is not very efficient, has a high dropout voltage, and no reverse bias protection. This regulator, like the LM317 will not do well without a heat sink or thermal pad on the PCB, especially with high current and/or a high input voltage relative to the output voltage. It is 30 year old technology, but can still be "good enough" in many applications.
 

wapo54001

Senior Member
Likely because it is not very efficient, has a high dropout voltage, and no reverse bias protection. This regulator, like the LM317 will not do well without a heat sink or thermal pad on the PCB, especially with high current and/or a high input voltage relative to the output voltage. It is 30 year old technology, but can still be "good enough" in many applications.
Are there better, more modern linear regulators? Can you recommend a better regulator for the following requirement?

1. Input voltage 10~16 vdc unregulated
2. Output voltage 5V, steady 50ma, occasional 1 minute peak of 100ma without heat sink.
3. Output voltage tolerance of 1% desired, 2% acceptable.
3. Lowest possible noise on the output (for driving sensitive linear circuits).
4. Through hole very desirable but not mandatory.
5. Simple topography with minimum external parts.
 

premelec

Senior Member
Probably not what you want to hear :) but I'd use a TO92 317 with resistor in series with the input voltage source [before the filter capacitor on 317 input]. The resistor scaled to drop 4 volts at .1 amp [40 ohms .01 x 40 = .4watts ] - No I haven't checked the TO92 dissipation - if it doesn't quite work add a zener diode before the 317 set to 7 volts or so - and get the power into the resistor... You are asking for a basically inefficient power design. Not clear why you want that... Using a switching regulator to get to 7 volts and then the linear regulator could result in very quiet system. Recently I've seen attention paid to switching regulator chips that are reportedly very quiet - I don't remember who the advertiser was. Closer examination of data sheets may give you your answer [e.g. voltage drop of 317 at various currents etc and dissipation allowance - left to the reader as an exercise :) ]
 

wapo54001

Senior Member
Checkout the following:

Micrel MIC2954
ST LP2954
ST L4931CZ50-AP
TI LM2930
@Goeytex, thanks for these additional suggestions. I'm going to put together a bit of a spreadsheet to look at the devices that have come up; definitely something here will be better than the LM317. I do like the possibility of 200~300 ma from a TO-92, that would be a significant benefit to me.
 

Dippy

Moderator
If you are doing a spreadsheet then I would suggest that you include some power calcs.

E.g. 12Vin to 3.3Vout @ 200mA = ?? Watts
That should have a TO-92 glowing nicely.

You must consider the thermal calcs after reading the headline specifications.
It could save a lot of time and some nasty smells.
 

Goeytex

Senior Member
While a device may be able to source 250 - 300 ma in a TO-92 Package, that does not mean that you can ignore heat dissipation considerations. There are small clip on type heat sinks designed specifically for T0-92 packages and they are pretty cheap.
 

premelec

Senior Member
@wapo# - you are changing specs - 100 ma for 1 minute is a lot different than 200 ma which is not likely for a TO92 but might work in a Dpak - in any case as Dippy says you've got to consider POWER. Using zeners and resistors can externalize some of the power dissipation before the regulator but the regulator itself will have current restrictions. It has been suggested on this forum that there be a tutorial on power supply design and this would still benefit a lot of beginners. When you have plenty of excess voltage to work with there's nothing wrong with 317s etc as they just require larger input/output differential to work... and protect themselves [theoretically ] from overheating failure - they turn off - which is not a failure of the part just of the design.
 

fernando_g

Senior Member
Many moons ago, I wrote a "Picaxe power supply tutorial" with some help from WestAust.
I submitted it, if I remember properly, in the finished projects section.
It may require an update, nevertheless is a good place to start.
 

premelec

Senior Member
Thanks - I remembered someone had done this - somehow it needs to be made more obvious to users as there are so many questions having to do with PS.
Maybe it can get into a manual....
 

Mary Rose 1545

New Member
make sure you use the decoupling capacitors and they are mounted as close as possible to the regulator, and are high quality, preferably tantalums.
 

jims

Senior Member
Many moons ago, I wrote a "Picaxe power supply tutorial" with some help from WestAust.
I submitted it, if I remember properly, in the finished projects section.
It may require an update, nevertheless is a good place to start.
fernando_g....
This sounds like it would be a good reference. I've done a FORUM search but can't find it. Can anyone help?
Thank you, JimS
 

wapo54001

Senior Member
I have always wondered, if two identical regulators are physically close together can they share an input capacitor of the value specified in the datasheet, or does each regulator need the value specified?
 

Goeytex

Senior Member
On many regulators the input capacitor (usually a small value) is optional, or is recommended if the regulator is more than 6-12 inches from the power source. ( refer to the datasheet). While I see no compelling reason to share a small value input cap, I don't imagine it could cause much harm.
 

geoff07

Senior Member
If they are close together, then two caps in parallel is almost electrically identical to one larger cap. The only differences might be caused by a) different types of cap with different ESR or resonances etc (e.g. ceramic vs tantulum), or b) some inductance in the connection between them. But the critical cap is the output cap, in the cases where one is needed at all (an LDO for example), as it is that which prevents oscillation. The input cap simply helps carry the load and is not critical. If two connected regs are taking a lot of power then too-small an input cap might affect regulation and allow some cross-modulation.

There is a very good explanation of regulator capacitors here: www.eetimes.com/document.asp?doc_id=1225555
 

Dippy

Moderator
That's a clear article, I haven't been through in detail.
Probably as said before a zillion times, for specific devices get the Data Sheets and App Notes from the Manufacturer of that device.
And if buying from Mr Unknown-Clone off Ebay fingers crossed :)
 

wapo54001

Senior Member
There is a very good explanation of regulator capacitors here: www.eetimes.com/document.asp?doc_id=1225555
The impression I'm getting from this document is that there are two reasons to have capacitors downstream from a regulator -- 1) one at the regulator to establish stability, and 2) more at locations downstream where abrupt changes in current demand can cause voltage to drop. Then, may I assume that distance from the regulator per se (without rapid change in current) does not require a capacitor, especially if current draw is steady-state?
 

rq3

Senior Member
The impression I'm getting from this document is that there are two reasons to have capacitors downstream from a regulator -- 1) one at the regulator to establish stability, and 2) more at locations downstream where abrupt changes in current demand can cause voltage to drop. Then, may I assume that distance from the regulator per se (without rapid change in current) does not require a capacitor, especially if current draw is steady-state?
Technically, yes. Practically, no. You would have to define "steady state". For a a very small incandescent light bulb, for example, you'd probably get away with it, as the filament is a thermally controlled current regulator (to some extent). But even a very low current logic gate switching at high speed may APPEAR to be steady state, but is asking for very large "gulps" of electrons every time it changes state. The local (large) capacitors provide those electrons, because they can't get there fast enough from the regulator due to wiring (or printed circuit trace) resistance (impedance).

And when the logic switches, taking its "gulp" of electrons, it creates a radio frequency pulse. On a good PC layout, the function of the small cap is to short that RF to ground, so that it won't radiate into your neighbor's television, cell phone, or AM/RF radio.
 

geoff07

Senior Member
So in practice, 1) put an electrolytic close to the reg, typically 1-2.2uF, to prevent LDO oscillation; 2) put 0.1uF close to each logic chip to absorb switching noise and a larger reservoir cap where the supply lead joins each subsystem to compensate for any voltage drop in the connection.
 

wapo54001

Senior Member
So in practice, 1) put an electrolytic close to the reg, typically 1-2.2uF, to prevent LDO oscillation; 2) put 0.1uF close to each logic chip to absorb switching noise and a larger reservoir cap where the supply lead joins each subsystem to compensate for any voltage drop in the connection.
Seems to me that the article in question is saying that the only reason to put a capacitor in the line is to either stabilize the regulator, or to be available to supply current for abruptly changing demand. Is the "larger reservoir cap" left over from the days when circuits demanded a lot of current or is it a belt-and-suspenders solution to a problem that doesn't exist? If it's required, how do you calculate the required size, how do you determine exactly where it is needed?
 

Dippy

Moderator
Even modern circuits can demand a lot of current in gulps so a 'reservoir' cap can still be needed.

Let's say, for example, your circuit which has a switched relay on the board.
Your PCB track line to the 'top' of the coil might be thin and a few cm long.
When that relay (coil) is triggered it will demand a current along that track.
Now you have IZ losses and radiation from that track.
Imagine you put a capacitor across the 'top' of the coil down to ground.
Now, when the coil is switched it can provide most of the energy to energise the coil.
This means a reduced pulse of current down the track as it has formed an RC.
(This can also apply with IRLED transmitter circuits).

Size?
Depends on the application and how long it needs to provide energy or hold the power up.
Have you heard of Time Constants?

Where is it needed?
As locally as possible to the thing.

An engineer from TI wrote an excellent and amusing article about the habit of novices "throwing" capacitors at circuits without proper consideration (I've done it myself before). You often don't need to chuck a zillion microfarads randomly into a circuit. And sometimes it can be harmful.
I have to add this is where simulators can help the grey matter :)

Summary.
No, 'reservoir capacitors' are here to stay where needed.
You, as the designer, need to work out the best position and value.
Though, quite often, guesswork and a little over-engineering works just fine ;)
 
Top