9v regulator

[the dude]

so I've got these regulators from electronics goldmine, they are marked  LM78L09, I use them on the dc jacks of the pedals I build, very strange thing is happening with them, after a few days they just go bad, they basicly stop putting out voltage, so I suspect they are a bad batch or just cheaply made regulators, anyway's just letting everyone know Dont buy them, I think they were 4 for a 1.00 (good deal), but now I've got headaches from them, so stay away
JR
PS: Anyone else have that problem with them?

[gez]

Exceeding the power rating?

[seanm]

An LM7809 needs at least a 12V input. Are you providing 12V in?

[the dude]

The power supply that I'm using is a 9vdc, I checked it with a meter and its more like 13-14 volts so its getting at least 12volts, and not exceeding the max voltage,
thanks,
JR

[gez]

The power supply that I'm using is a 9vdc, I checked it with a meter and its more like 13-14 volts so its getting at least 12volts, and not exceeding the max voltage,

If it's DC without a load then the measurement is pretty meaningless.

[gez]

PS  If you've already got a 9VDC PSU, why do you need further regulation?  Or is it some cheapo thing with a ripple problem?

[the dude]

I like to use the 9v regulators for 2 reasons.
1. The pedals seem to have less unwanted noise in the circuit.
2. To protect the circuit if say a 24vdc was accidently plugged into the jack.
I'm also filtering the power using a 47uf electro cap.
JR

[AWelti]

My first thought - the regulator you mentioned is 100mA low power version, make sure you're not simply burning it out with too much demand
My Second thought - If you're using filter caps in the circuit you will need two diodes to be absolutely safe. 

D1 points from the regulator's output leg back to the input leg - when you disconnect power, the regulator's input goes to 0V while the cap tries to hold the output at 9V; this diode drains the cap around the regulator instead of through the regulator.

D2 points from ground to the output of the regulator - especially if you are switching power with the ring of a TRS jack, the charge on the cap could float negative; this diode won't allow it to go too far negative.

If they still blow up, then maybe they are a bad batch...  As far as under powering (less than 12V) I don't believe that would kill the regulator, but I have seen power from an under powered regulator and it's actually dirtier than unregulated power - nastier than ripple alone.

[robbiemcm]

I had my 12v/DC unregulated power supply and I attatched my 7809 like the picture above shows, and this seems to give me a nice ~9v supply. However, I havn't tested it out on any pedals yet as I don;t have a working one But my question is, what's this 47uF filter cap? Do I just attatch that between the two rails one after the 7809? But wouldn't that make it permanently on?

[George Giblet]

- Make sure the DC input is greater than 10.7V (that is to the minimum point on input when ripple is present)
- Make sure the input DC rail is sufficiently filtered.  You might be pushing it with 47uF if that's the only input filter cap.
- Make sure the regulator is not too far from the input filter cap.  If it is you have to add a cap right at the input
  terminals of the regulator: minimum = 330nF, preferred is 4.7uF or so.  If you don't the device could oscillate.
- Make sure you have a cap right across the output.  I don't trust these regulators with anything less than 100nF,
  this also prevents oscillation - for some reason the L devices tend to be friendlier than the TO220 tab types.
- Make sure the power dissipation is below the specified limits.
- Make sure the current draw is not exceeding the output protection limit.

If you don't stick to these conditions the device may cause you grief.

[gez]

I like to use the 9v regulators for 2 reasons.
1. The pedals seem to have less unwanted noise in the circuit.
2. To protect the circuit if say a 24vdc was accidently plugged into the jack.

1. Why don't you just build/buy a decent PSU? 
2.  You already know what you're plugging into it - 9V - so what's the problem?  If you're building stuff for other people there are other ways to protect a circuit that won't cause any problems.

[Paul Perry (Frostwave)]

I like using regulators too, but they aren't very good as protective devices.. because, it is harder to protect a regulator, than it is to protect the circuit itself WITHOUT a regulator!

And, here is something that can go wrong.. if you have a big cap on the output of a reg, and then short the INPUT (or have a leaky or small input cap and turn off the supply) you can momentarily have more volts at the output of the reg than at the input,  which is fatal. You sometimes see a diode connected backwards from the in & out terminals of the reg, that is what it is for, so the regulator can't try to run backwards.

In any case, if your regs are failing, you should measure the in & out voltages while under load and let us know what they are.

[robbiemcm]

Wait.. so what if I took what I showed you above in that picture and added a (x)uF cap from top rail to bottom rail directly before the 7809, then a (y)uF cap directly after the 7809 once again from top rail to bottom. Would that work? (Also, if it will work, would you mind helping me out with the values )

[gez]

Wait.. so what if I took what I showed you above in that picture and added a (x)uF cap from top rail to bottom rail directly before the 7809, then a (y)uF cap directly after the 7809 once again from top rail to bottom. Would that work? (Also, if it will work, would you mind helping me out with the values )

Regulators usually require a cap at input and output.  The input cap is often supplied courtesy of the smoothing cap used in a PSU, or (if using something like a 5V reg) the filter cap across the rails which often, though not always, gets used in stompbox circuits.  The data sheet for the reg you use should have the details of what is required.

[AWelti]

I noticed in your diagram above you pictured a heatsink - that would be a beefier TO-220 package, good for about 1.5 amps - the 78"L"09 doesn't have a heatsink it's a TO-92 package and is good for about 100mA - as for hooking it up, here is the datasheet... they show some typical circuits - I think you would want to go with a combination of fig. 1, fig. 6 and fig. 7 (about 9 or 10 pages into the document):

http://rocky.digikey.com/WebLib/Texas%20Instruments/Web%20data/UA78L00%20Series.pdf

Good Luck

[Paul Perry (Frostwave)]

If it helps, I usually have 1000 uF at the input of the reg, and 100 uF at the output. reading the fine print in the National Semiconductor application notes, and the data sheets, it isn't always the case that the more cap on the output  the better, you can get some values that tend to increase the noise a bit.

[robbiemcm]

Thanks for that, this is the one I used: http://www.altronics.com.au/index.asp?area=item&id=Z0509

[gez]

If it helps, I usually have 1000 uF at the input of the reg,

Obviously you know this already Paul, but just to avoid confusion you don't need such a high value if the input to the regulator is already regulated, ie coming from a DC wall wart.  The large value mentioned would be the smoothing cap following a rectified AC signal.

Yeah, I've noticed that some large value caps do introduce noise.  If I need a large value at the output of a regulator I often shove a small value resistor before it to form a simple RC low pass and this usually does the job.

[Paul Perry (Frostwave)]

Obviously you know this already Paul, but just to avoid confusion you don't need such a high value if the input to the regulator is already regulated, ie coming from a DC wall wart.  The large value mentioned would be the smoothing cap following a rectified AC signal.

True enough, but more than once I've found someone using a crap adaptor they found somewhere that has a dried-out cap, or one that was "good enough for the original job" (like a DC power drill!) but not quite up to the audio application. But yeah, 99% of the time it's cool......

[robbiemcm]

So a 1000uF cap before it, what value after it? It seems people use 4.7uF, 47uF and 100uF.. how do I know which one?