18v power supply regulator overheating

[TelmoJackson]

I've tested it one more time and the voltage readings are good, so I think the transformer is ok. I don't think it overheated enough to burn, it reached 55ºC on the enclosure.

I just tried the "3sec test" and I can touch it with my fingers for like 15-20sec. If redneck science doesn't lie it might be ok?

I'll put bigger caps anyway, there's room for them and I don't mind spending like 1€ more on this building.

[Rixen]

check there are no shorted diodes in the rectifier...

[antonis]

Heat is produced by power dissipation and can't be present without any current (no load)..

A small amount of power dissipation is due to internal losses, like core and winding coil losses but it doesn't rationaliaze your case temperature rise by itself..

55^oC on the enclosure of transformer could be acceptable in case of reaching it's 2.4VA power rating..

So, somewhere exists current consumption (leakage, faulty or anything else..)

[TelmoJackson]

check there are no shorted diodes in the rectifier...

You got it! Thank you so much I was getting so frustrated... One rectifier diode (the one that drains negative voltage to ground) was shorted, so was loading the transformer.

Now it works fine and everything is colder than Kimi Raikonen. Just for security I put 63v 2200uf cap.

Thank you all guys!

[duck_arse]

not so fast! what is the temperature of the transformer, now it is running tikkety-boo, please? and the tran temp full(-ish) load?

[TelmoJackson]

not so fast! what is the temperature of the transformer, now it is running tikkety-boo, please? and the tran temp full(-ish) load?

It is going ok, good output, good voltage in everything and 0 heating. Temperature on the cover of the transformer (and everything else) is less than 10ºC over the ambient temperature.

[Rixen]

great that it's working !

just for the record, there are issues with going for higher than necessary smoothing capacitors in power supplies: The peak charging currents can get large, causing stress on components, higher losses, higher pulsed magnetic fields (with more harmonics) and potentially saturation of the transformer.

The reason for this is a larger capacitor has less voltage droop at a given load, therefore when the next half sine wave comes from the rectifier only the very tip of the wave is available to charge the energy drained from the capacitor. Same energy in less time will require more current.

I normally use 1000 uF per amp of current, but one should normally calculate the ripple voltage at full load to make sure the regulator never drops out of regulation.

[TelmoJackson]

Well... I need more of your help please

I just tested it with pedals to an amp and I'm getting soooo much ground loop noise. When I plug only one pedal: silent, when I plug two pedals to different transformers: silent but when I plug two pedals to the same transformer: BBBZZZZZZZZZZ 50Hz hums like a beast.

How can I solve it? This is the schematic for the isolation of the outputs:


(Polarity of the bridge rect. is reversed, but it's just the drawing)

EDIT: I just thought this looking at my schematic: could it be that every channel has a rectifier? Shouldn't it be one rectifier and then all the channels in parallel?

[antonis]

I normally use 1000 uF per amp of current, but one should normally calculate the ripple voltage at full load to make sure the regulator never drops out of regulation.

There are many "rules of thumb" (even "rules of blind eye") but all of them deal with the amount of V_p-p ripple in a mean of "absolute" voltage value..
(e.g. for 1000μF per Ampere, you have a 1.77V (60Hz) to 2.12V (50Hz) voltage ripple which may be acceptable for a 33VDC but forbitten for a 5VDC..)

[antonis]

<This is the schematic for the isolation of the outputs>

I see NO isolation at all..!!

You need SEPERATE secondary windings..
(One for EACH individual supply..)

Like in your own schematic:

[TelmoJackson]

<This is the schematic for the isolation of the outputs>

I see NO isolation at all..!!

Wow, and what's isolation? Now you really knock me out, just one rectifier would do the job or should it be a transformer for output?

[TelmoJackson]

You need SEPERATE secondary windings..
(One for EACH individual supply..)

So it is a transformer per output right?
But I have a power supply that I opened and reverse engineered it and it was just one transformer, one reftifier, one regulator and rippling caps and then all the outputs in parallel just like a daisy chain and it is fairly silent. (More than mine for sure )

[TelmoJackson]

<This is the schematic for the isolation of the outputs>

I see NO isolation at all..!!

Wow, and what's isolation? Now you really knock me out, just one rectifier would do the job or should it be a transformer for output?

EDIT: sorry for this post, missclicked I was trying to edit the previous one 

[antonis]

Wow, and what's isolation? Now you really knock me out, just one rectifier would do the job or should it be a transformer for output?

Wow, delete your query before R.G. comes and slap your hand..!! 
(read yesterday - not a.s.a.p - about Spyder at Geofex..)

[antonis]

So it is a transformer per output right?

If you slow down your post response, we'll be able to tell you...!!!  

Yes, a transformer per output should be ideal but a multi secondary transformer is also OK for the job..
(and more money & space saver..) 

In brief, ONE secondary winding for EACH ONE supply..!!!

But I have a power supply that I opened and reverse engineered it and it was just one transformer, one reftifier, one regulator and rippling caps and then all the outputs in parallel just like a daisy chain and it is fairly silent. (More than mine for sure )

One PS silence doesn't guarantee the same for another one..

[TelmoJackson]

Sorry for the multiresponse, I'm getting a little frustrated with this...
So the ideal is one transformer or, at least, winding per output, that's for sure.
Now, can I try to fix my schematic with the transformers I have now?
Would it be better if I rewire to this schematic?


Sorry again for the multipost and thank you for your help!

EDIT: the power supply I took the idea from was more like this than one rect per channel, but my brilliant mind thought it would be better to have a rectifier per channel.

[antonis]

 

It doesn't make any difference if the outputs share the same GND at their common rectifier point (- at common bridge ) or at the their own one (- at each bridge) because all these gounds are (alternativelly but in phase) interconnected to the same wire of transformer secondary..
(suppose you know how full wave rectification works..)

To get a little further, exactly the same stands for different secondarys of a center tapped transformer..
(some people are confused about e.g.  2 X 12 VAC CT considering it as a dual isolated secondary one - which, of course, doesn't apply..!!)

[TelmoJackson]

It doesn't make any difference if the outputs share the same GND at their common rectifier point (- at common bridge ) or at the their own one (- at each bridge) because all these gounds are (alternativelly but in phase) interconnected to the same wire of transformer secondary..
(suppose you know how full wave rectification works..)

To get a little further, exactly the same stands for different secondarys of a center tapped transformer..
(some people are confused about e.g.  2 X 12 VAC CT considering it as a dual isolated secondary one - which, of course, doesn't apply..!!)

So I don't bother trying to put a single rectifier? Just go for single transformers right?
Daisychain the outputs (like I saw on the one I was copying) won't work neither, or even make it worse am I right?

(I basically know how rectification works, but not so deeply, just know how to follow schematics and barely know what every part does)

[antonis]

Yes, you can go for single transformers if you insist..!!

Just think of also use multi-secondaries type.. 

[TelmoJackson]

Yes, you can go for single transformers if you insist..!!

Just think of also use multi-secondaries type.. 

Well... I don't insist but if there's no other way to make it silent I'd have to spend like 40€ more
I thought of multi-secondaries but at the local shop there's only symetrical secondaries, and I think I'm gonna have the same problem.
Soooo... Gonna spend some more money and hopefully I don't have to ask anything else (fingers crossed) hahahha
Thank you so much!!