low noise power supply - advice needed

[amptramp]

Your latest schematic does not show the capacitor across the voltage adjust pot.  This capacitor gives you a lot of extra ripple rejection and is the reason for the diode from the adjust pin to the output pin on the LM338.

The resistors between the diode bridge and the first capacitor have to carry the momentary pulse currents.  Since dissipation is R*I^2, these will take a lot more power than the resistor between the two caps at the bridge output.  They are going to run hot and should be rated for plenty of wattage.  These resistors can be put in series with the transformer secondary.  A capacitor between the output side of the resistor and the other rail will make a very effective EMI filter for garbage coming in on the AC line, but of course, this will also add to the dissipation.

The LM338 requires a heat sink to get anywhere near its rated current output.  The heatsink must be insulated from ground.

[blana]

First of all, thanks for your help!

Answering in the incoming order:

The LM338 is supposed to get hot but might need a heat sink to avoid overheating.

I am using a huge heat sink.

What did you have connected to the output as a load when the resistors got hot?  Are the resistors properly rated for the power that you expect them to have to dissipate?

First, the load was the ~1.5k resistor suggested by Kesh, but i removed it afterwards - and i think it was a bad idea, because all the drop was probably taken by the internal resistance of C3, which got swollen. I think i'll put that resistance back.

I had no other stuff connected to the output of the source.


The resistors before the LM 338 are rated 5w and after the calculations done with the help of Kesh, the three of them should sustain a possible of 10v drop on all of them. V=I * R, I = 3.3 / 2.7 = 1.2. P = V*I = 4w.

Are the capacitors rated well above the voltages they'll see?

C4 is 200v, C1 is 25v (oops, its above, but not "well above"), C2 is 100v, C5 is 100v, C3 is 25v.

Your latest schematic does not show the capacitor across the voltage adjust pot.  This capacitor gives you a lot of extra ripple rejection and is the reason for the diode from the adjust pin to the output pin on the LM338.

Oh, s***. I forgot this one in the schematic, somehow it was an older version of the file. I'll add this one to the circuit asap. Thanks for spotting this, i was sure it was there.

The resistors between the diode bridge and the first capacitor have to carry the momentary pulse currents.  Since dissipation is R*I^2, these will take a lot more power than the resistor between the two caps at the bridge output.  They are going to run hot and should be rated for plenty of wattage.

Yes, i think the resistors can dissipate the required wattage, if my previous calculations are correct. "The resistors before the LM 338 are rated 5w and after the calculations done with the help of Kesh, the three of them should sustain a possible of 10v drop on all of them. V=I * R, I = 3.3 / 2.7 = 1.2. P = V*I = 4w."

These resistors can be put in series with the transformer secondary.  A capacitor between the output side of the resistor and the other rail will make a very effective EMI filter for garbage coming in on the AC line, but of course, this will also add to the dissipation.

I'm not quite sure i understand this correctly. Are you suggesting the placement of the resistor chain should be before the rectifier bridge? And also have a capacitor in series at the end of the resistor chain to help the filtering of AC spikes?

The LM338 requires a heat sink to get anywhere near its rated current output.  The heatsink must be insulated from ground.

Yeah, i have a huge heatsink, and its siolated from the ground.


Thanks a lot for your help!

[Kesh]

Hello again.

I just built the circuit, but it has some issues.

The resistances and the lm338 get hot fast. The output capacitor got swollen after i managed to do seme measurements. One note is that the output resistance (1.5k) got hot also, and i removed it in the idea of having no consumer and being able to measure the voltages without the circuit getting hot. It partly worked, meaning that it gets hot slower, but it still gets hot.

The current schematic is this:

https://www.dropbox.com/s/g7sdxzj4w4oxh0w/lm%20338%20source%20v2.jpg

The only diferences are: the LM 338, C4 - 470u at 200v and high temp, and the control resistances which are fixed.

What should i do? :S

If R7 is getting hot then you have done something very wrong. as 9V over 1.5K cannot overload even a 1/8W resistor.

I cannot see what R1 is. It should be about 120 ohms, and for 9V R2 should be 750 ohms. It looks like you have wildly out values there.

If the output cap swelled up then maybe it is in backwards.

Did you take any volt readings across the hot resistors?

[blana]

Hello again,

After i added the capacitor across the R2, i also noticed another grave mistake: the output capacitor was in reverse polarity. After fixing these mistakes, the circuit outputs a stable 9.03 V and stays extremely cool.

Tomorrow i will start building the box, and test the humming.

Thanks again for your invaluable help!

[blana]

Hey Kesh, you answered while i was posting and i did not see your reply! Thank you very much!

Indeed the output cap was reversed (how did you know? ) The R1 and R2 are as you say 120 and 750, I just forgot to change the values in the picture, sorry for that!

Awesome! Thanks again!

[Kesh]

glad it's working. and i knew because that's what they do   a few seconds longer and it could have exploded

by the way, in circuits i have seen and copied, it is more usual to have more resistance between C4 and C1 than the bridge and C4

[blana]

Hello again!

Due to computer problems i have not been able to post for a while.

I have built the source and its working fine, without hum.

Once again, I'd like to thank everyone for their help! You guys are great!

[blana]

Hello!

Me again, with another question / problem.

Today I have aquired a tube stompbox (silver dragon) that functions on AC current. Beside the dragon, I am using a whammy (also an AC pedal) and I am thinking of including some AC outputs on my power source, to supply all my effects from a single source.

Since i have a 18v toroidal coil already in the source, i was thinking there may be a way to create a voltage divider to obtain the necessary AC voltage drops. Looking on the net i found some articles about AC voltage dividers, but also some warnings that they should not be sued for circuits drawing more than a few mili amps. Since both my AC pedals draw more than this, I am not sure how to achieve this.

Any suggestions?

Thanks in advance.

[Kesh]

your toroidal is 2x9V. you could just tap one of the secondaries. but you may get some odd, possibly dangerous, grounding if you do it wrong.

[blana]

Hey, thanks for the reply!


I thought at that for the whammy, which requires 9v, but the "silver dragon" pedal requires an unusual voltage - 13v, so it wont solve the whole problem.


Thanks,

[Kesh]

Check if that 13V is 13V peak

9V rms = 13V peak

[blana]

The thing is that the dragon's original adapter was fried by its previous possessor, so i wont be able to measure it. Another useful info is that i managed to power the effect with the digitech ac adapter which is supposed to be a 9v one, but I'm not sure it functions under optimal conditions. I will do a check on digitech's power supply maybe i can find out if it is 9v peak or rms.

I did a net search for the original adapter and i could not find any info.

In the effect's manual, they only say this:

Maximum Input                  +8dBu
Input Impedance                470K
Current Consumption         25mA
Power Requirements         1A 13VAC

Thanks!

[blana]

Hello,

I kept testing the silver dragon with the 9v AC source and it works without any obvious problems, so i decided to go with the 9v ac supply instead of the 13 one, (which i have no idea how to build). I modified my source last night and now it has two AC sockets. I just need to test it at the rehearsal room with everything connected, but I hope it works just fine.

Thanks again for your priceless help!

[blana]

Hello again,

Unfortunately, the source does not work so well.

Whenever i chain an AC pedal to a DC pedal, there is a huge BRRR noise. If I break the connection between pedals everything works fine. I am guessing something happens when the "grounds" of the pedals are connected, but what?

Kesh, you said something about odd grounding if i wire the coil wrong, can this be the cause?

This is a rough schematic of what's inside the box:

https://www.dropbox.com/s/ep6zxpkvd46uvhu/ac-and-dc-source.png


What am i doing wrong?

Thanks in advance.

[Jdansti]

If I could jump in before Kesh answers, how are the AC pedals grounded to the enclosure?  If the pedal using AC has one leg of the AC attached to the enclosure, you could be hearing AC hum (50 Hz in your country?).  The guitar plugs/cables create a common ground for the pedals and amp.  For the AC pedals, I think you'd need to connect the converted DC negative (0V) to the enclosure (where it's converted to DC inside the enclosure), but I'm not sure if that would help when mixing AC and DC.  Just some more thoughts to consider.

[blana]

Thanks for the reply Jdansti!

The DC sockets are indeed connected to metal enclosure, with the negative side as you say (or ground).

For the AC sockets i used all-plastic sockets, especially to prevent them from having the same "ground" as the dc circuit. To my mind was like having them powered by different warts, where there is no connection between the power supplies' grounds. Chances are I am wrong, but i dont yet understand why.
The AC pedals are in no way grounded to anything.

Regarding what you say  about connecting the inside ground of the pedals to the supply enclosure: I'm pretty scared of opening up my whammy pedal, i did once before and it looks like a computer mainboard - i dont feel so confident to mess up with that kind of circuit...

Regarding the noise: maybe my previous description was too vague - its not the usual hum, or loud hum, its a SUPER LOUD BRRRR-GRRRR-IIIOIII-WERYWRONGSTUFFGOINGON type of sound, thats why i suspect I'm doing something very wrong.

Thanks for your help!

[Jdansti]

Good information. You might want to do this test to help with the troubleshooting:

1) Connect guitar to a DC pedal, patch cord to AC pedal, patch cord to amp. Power up both pedals and describe noise with:

a) both pedals in bypass,
b) DC pedal on/AC pedal bypassed,
c) AC pedal on/DC bypassed,
d) both pedals on,
e) both pedals on, but disconnect the patch cord between the pedals and plug guitar into AC pedal,
f) keeping the configuration in e, unplug the power to the DC pedal.

2) Connect guitar to a AC pedal, patch cord to DC pedal, patch cord to amp. Power up both pedals and describe noise with:

a) both pedals in bypass,
b) AC pedal on/DC pedal bypassed,
c) DC pedal on/AC bypassed,
d) both pedals on,
e) both pedals on, but disconnect the patch cord between the pedals and plug guitar into DC pedal,
f) keeping the configuration in e, unplug the power to the AC pedal.
.

[blana]

Hello Jdansti,

Thank you very much for your help!

I tried doing what you so thoroughly described in your post. The conclusion is that everything works fine until i connect the two effects via a cable. When the are connected, the noise is present at all times. As soon as one of the power supplies is disconnected, the effect stops (doh) and the noise stops.

There are two worrying things i've noticed while doing these tests:

1. When the effects are powered, and connected, the toroidal coil starts humming - i'm guessing this is bad and is because of a short.
2. My digital effect (digitech rp250) restarts, or does strange things on the display.

My noob guess is that there is a short caused when connecting the grounds, but I dont know where.

The only solution i can think of is to separate the stabilizer circuit via a 1:1 coil, or somehow with a pair of condensers  (one for each of the stabiliser wires, connected in series with each wire - let me know if this sounds stupid).

Thanks!

[R.G.]

This could well be the updated version of why you don't want a power jack grounded to the enclosure on a Uni/Neovibe.

If the pedal has a bridge rectifier as the first thing in, it runs the incoming power through one diode to the signal ground all the time, and probably also hooks the signal ground to the chassis. If the incoming power is connected to signal ground by other pedals as well, you have the condition where the two bottom diodes of the bridge are trying to force the power supply negative side to be at two different voltages at once. For an AC power supply, this gives horrible hum because the signal ground is being wiggled around.

[blana]

Hello and thanks for your reply!

I understand. What can i do to prevent this? I mean, even if the DC supply does not touch the enclosure, if an AC effect is connected to a DC effect, the same will happen, since their grounds will be connected. I tried to scavange a suitable coupling transformer from an old ATX source, to separate the two supplies, but i couldn't find one, so i'm out of ideas.

Thanks!