HOW does this 24V regulator put out 48V?

[jaunty]

" alt="" class="bbc_img" />

doing a repair on a power supply (in this case a 48v phantom power rail going to a mixing board). IC3 is an LM7824 (capable of 24V output regulated). I'm not getting the expected 48V output but rather more like 17V. Obviously there's a problem somewhere (already replaced all the capacitors as a matter of general service) but in order to diagnose the problem i need to understand how this LM7824 is generating 48V. Anyone?

[bdunham7]

(Attachment Link) " alt="" class="bbc_img" />

doing a repair on a power supply (in this case a 48v phantom power rail going to a mixing board). IC3 is an LM7824 (capable of 24V output regulated). I'm not getting the expected 48V output but rather more like 17V. Obviously there's a problem somewhere (already replaced all the capacitors as a matter of general service) but in order to diagnose the problem i need to understand how this LM7824 is generating 48V. Anyone?

The R11-R12 divider would be what set the voltage.  What AC voltage are you seeing at across the transformer leads and what DC voltages do you measure across C11 and across R11 and R12, or, from the R11/R12 junction to ground?

[HackedFridgeMagnet]

What is the voltage at the pins of the regulator.

Measure the DC voltages around the circuit and update.

The regulator will give 24V from its output to common pin if its input to common pin is  > ~25V

the common pin is lifted from ground which means you will get something greater than 24 V out.

the gain is something similar to this page.
http://www.learnerswings.com/2014/07/adjustable-output-voltage-regulator.html

inside the lm7824 there already is a resistor divider, we have to take this into account.

[jaunty]

thanks - I will reply later with test data when i get access, though i'm a bit ambivalent about supplying it only because SOMETHING isn't working properly and it might confuse issues - i was just hoping someone could give me a pointer if I supplied the basic circuit topology so i could understand how such a thing would be possible. my best guess is that this circuit is 'riding shotgun' on top of the neighbouring circuit providing pos 18V (indicated as 'V+' in the rail across the bottom, where one might ordinarily see a ground)

[ledtester]

What is the voltage at the input pin (pin 1) of IC3, and what are the values of R11 and R12? That will say a lot about how it is supposed to work.

[bdunham7]

thanks - I will reply later with test data when i get access, though i'm a bit ambivalent about supplying it only because SOMETHING isn't working properly and it might confuse issues - i was just hoping someone could give me a pointer if I supplied the basic circuit topology so i could understand how such a thing would be possible. my best guess is that this circuit is 'riding shotgun' on top of the neighbouring circuit providing pos 18V (indicated as 'V+' in the rail across the bottom, where one might ordinarily see a ground)

The LM7824 is going to try to maintain 24 volts between its output terminal and its ground or sense terminal.  Less than 24 volts it will turn on more, more than that it shuts off more.  The LM7824 has no way of knowing what the rest of the circuit is doing.  In this case, adding R11 and R12 causes current to flow in those resistors which means a voltage drop.  In normal operation, R12 will have 24 volts across it and R11 will have some other voltage that I can't determine without knowing its value and the current--and that the LM7824 has no way of knowing about.  So if you measure from the output to the bottom of R11, which is the same as your V+, you should get 24 volts + the voltage drop across R11.  If your V+ line has a potential to another point (lets call it ground) --you say 18 volts--again this circuit doesn't know or care about that.  If you measure from the output to ground in that case you will get 24V + the R11 voltage + 18V. 

If there is another power supply in series like this and your circuit is an 'add-on', and the output is 17V measure to ground, while V+ is 18V, the most likely cause is that there is a complete failure, perhaps no power from the transformer, and the 17 volts is just your 18V from the V+ bleeding through the resistors and the unpowered regulator.  Or, the regulator could be open, with the same result.

[floobydust]

I think the IC is better a LM317HV or TL783. To increase a 78xx output voltage I have seen a zener used in the GND terminal.

[jaunty]

What is the voltage at the input pin (pin 1) of IC3, and what are the values of R11 and R12? That will say a lot about how it is supposed to work.

well not with a failed bridge rectifier - which seems to be the issue with the circuit. i just thought someone might recognize the topology. getting 0.919V DC across C11 but 32VAC coming in from the toroid winding. input to the rectifier is 17.x volts and output 15.x volts ... so i guess it partly answers our question. it seems the 17.x (18V really by design) is supposed to be borrowed from the other leg and added to whatever is generated by this particular leg of the circuit by the rectifier (?) - still makes no sense to me - thought this would have been a big 'no no' adding voltages like this ...

[jaunty]

thanks - I will reply later with test data when i get access, though i'm a bit ambivalent about supplying it only because SOMETHING isn't working properly and it might confuse issues - i was just hoping someone could give me a pointer if I supplied the basic circuit topology so i could understand how such a thing would be possible. my best guess is that this circuit is 'riding shotgun' on top of the neighbouring circuit providing pos 18V (indicated as 'V+' in the rail across the bottom, where one might ordinarily see a ground)

The LM7824 is going to try to maintain 24 volts between its output terminal and its ground or sense terminal.  Less than 24 volts it will turn on more, more than that it shuts off more.  The LM7824 has no way of knowing what the rest of the circuit is doing.  In this case, adding R11 and R12 causes current to flow in those resistors which means a voltage drop.  In normal operation, R12 will have 24 volts across it and R11 will have some other voltage that I can't determine without knowing its value and the current--and that the LM7824 has no way of knowing about.  So if you measure from the output to the bottom of R11, which is the same as your V+, you should get 24 volts + the voltage drop across R11.  If your V+ line has a potential to another point (lets call it ground) --you say 18 volts--again this circuit doesn't know or care about that.  If you measure from the output to ground in that case you will get 24V + the R11 voltage + 18V. 

If there is another power supply in series like this and your circuit is an 'add-on', and the output is 17V measure to ground, while V+ is 18V, the most likely cause is that there is a complete failure, perhaps no power from the transformer, and the 17 volts is just your 18V from the V+ bleeding through the resistors and the unpowered regulator.  Or, the regulator could be open, with the same result.

yes that seems to be what's going on mostly ... 32VAC out of the trafo and nil coming out of the diode bridge. but still scratching my head over the philosophy behind this - how the two legs are added together through the LM7824, if that's what's going on.

the resistors are R11 = 470 ohm, R12 = 3K7 and R13 = 3K7

[jaunty]

posted just above. nil (0.9v across C11) and 32VAC out of the toroid (thank goodness). dead diode bridge i reckon - but thats an easy fix. knowing the potentials across R11 and R12 isn't going to tell us much it would seem. Guess the thing to do is put in a BY164 (closest thing i have to a B80C1500) and fire it up again and figure out how this thing is working. I may disable this circuit or mod it if it's taking a current draw from the +18V leg as i'd rather not compromise the feed to the op amps in the mixing board. if that's what's going on it seems it could be a much better design!

[jaunty]

this may help - the full power supply for confirmation (should have posted earlier but i was trying to keep things simple)

" alt="" class="bbc_img" />

it does not appear to be drawing power from the +18V circuit however (according to this)

[tunk]

The voltage across R12 is 24V and the current is 6.5mA.
The current through R11 is this plus the "Quiescent Current"
(max 8mA, typically lower, maybe 4-5mA) giving ~5V.
The total: 24 + 5 + 18 = 47V

[bdunham7]

this may help - the full power supply for confirmation (should have posted earlier but i was trying to keep things simple)

(Attachment Link) " alt="" class="bbc_img" />

it does not appear to be drawing power from the +18V circuit however (according to this)

Don't do that.  The complete picture is always helpful.

This is not a complicated issue.  If the power supplies are truly isolated, then you can connect them in series and add up voltages just as if you were using 9 volt batteries.  Nothing odd or wrong about it.  I'm assuming the 48V supply is very low current.

As for the bridge rectifier failing in that manner, I would take it out and test it thoroughly before coming to that conclusion.  There might be some other strange issue causing a loss of isolation, or something.  It would have to have failed open somwhere for it to behave this way and open is not the typical failure mode. Hopefully it is the failure in your case.

[jaunty]

well after a ridiculous amount of running around and fooling myself - thinking the fault must have been a short somewhere - i discovered that the circuit had a blown fuse - so i popped a new one in and blew that immediately too (!). Having checked over nearly everything I finally resorted to my dim bulb tester - and yep sure enough that thing lit up BRIGHT when i put a fuse back in! So i decided to second guess myself and pull the secondary from their solder connections and GUESS WHAT? 1-2 megs is what i measured!! 

well at least i feel some satisfaction at least of having found the issue. But here's my last question then (i gave up trying to understand how an  LM7824 generates 48V) - how the hell does an open secondary generate a fault condition??? the fuse is located between the secondary and the diode bridge! Shouldn't an open secondary behave like any open circuit?? not making sense to me - but it sure explains the earlier symptoms!!

[jaunty]

also - is it dangerous to operate the transformer with the open secondary? i was reading something about this condition causing abnormally high core EM saturation and therefore messing everything else up (?) or no?

[bdunham7]

Not making a lot of sense here.

On your schematic, is it F1, F2 or F3 that is blowing?  Or is there another fuse?

An open secondary would be open, no current would flow and your fuse would not blow.  And no, your open/disconnected secondary shouldn't cause any serious issues--as in nobody will die.

[jaunty]

not making sense to me either! not a great schematic sorry - it's a totally different fuse - let's call it F4 - located between the secondary for the 48V rail and the diode bridge! but it's creating a fault (via dim bulb tester) in the main line when that fuse is in. I disconnected the secondary and measure well over 1 megohm. I don't see how that could be creating the fault but there seems to be lots of little gaps in my knowledge like that ... what do you surmise from that? I'd love to know

[bdunham7]

You previously measured voltage at this transformer output.  The fuse is blown, indicating a short.  You've checked with a dim-bulb tester with the fuse in and it indicates high draw, indicating a short, assuming you have an appropriate bulb and all. 

I would guess that there is a short in the bridge rectifier and that you are making some error in measuring the resistance of the transformer.  First, try the dim-bulb test without the fuse in place--if it still lights up that means something else is going on.  If that is good--no short with fuse out--then pull the bridge rectifier out and test that. 

[jaunty]

yes! already done. The dim bulb tester behaves BEAUTIFULLY with the fuse out. just a slight glow for half a second and then off. earlier - suspecting the bridge rectifier i removed it and tested it (fine) and then put in a new one with higher rating anyway - just to be safe. I've also replaced the reservoir cap and the LM7824 with new ... but yeah there sure is something not making sense here! thanks. I'll double check the winding resistance - who knows maybe it was a lacquer conductivity issue - i did notice it was jumping around a bit at the time ...


amendment: dammit! you're right!! 1.3 ohms! gahhh ... i felt better when i decided the winding was open. hmmm
 ...  i guess the next step (after double checking the new bridge rectifier) would be to put the secondary leads back on the board and then cut the traces after the reservoir cap and see how things behave then? I was about to do that anyway. Sorry for the confusion.

[jaunty]

what is making this thing a bit complicated is that the +18V rail seems to be feeding the rail with the LM7824 on it to make up the 48V. Although it's not being shown in the schematics. But both the +18V and -18V rails seem to be performing nicely - which is making me think the fault isn't likely to be coming from that end of things ... i feel like i need fresh eyes on this.

[bdunham7]

yes! already done. The dim bulb tester behaves BEAUTIFULLY with the fuse out. just a slight glow for half a second and then off. earlier - suspecting the bridge rectifier i removed it and tested it (fine) and then put in a new one with higher rating anyway - just to be safe. I've also replaced the reservoir cap and the LM7824 with new ... but yeah there sure is something not making sense here! thanks. I'll double check the winding resistance - who knows maybe it was a lacquer conductivity issue - i did notice it was jumping around a bit at the time ...


amendment: dammit! you're right!! 1.3 ohms! gahhh ... i felt better when i decided the winding was open. hmmm
 ...  i guess the next step (after double checking the new bridge rectifier) would be to put the secondary leads back on the board and then cut the traces after the reservoir cap and see how things behave then? I was about to do that anyway. Sorry for the confusion.

Whoa!  Slow down!  More thinking, less cutting and soldering! 

Since your soldering iron is warm, take out the LM7824 completely and redo your dim bulb test.  That should narrow things down a bit.  If it still fails, disconnect the transformer secondaries again and test for continuity between those and F2 and F3.

[jaunty]

well i hadn't cut anything yet - but what i mentioned was previously tested.

Anyway- i just did what you recommended - pulled the 7824 and reconnected the secondary for that rail and yep ... bulb lit up like the sun - which would have been my prediction i suppose.

one really ODD thing going on here that i'm not sure has anything to do with the problem - but differs from every other PS i've worked on is that I'm getting open measurements IN CIRCUIT with the bridge rectifier - both the old one and new one i put in. when it remove them they test fine. But I've never seen that before - i've always had a fair amount of success testing them in circuit in the past. I can't think of any condition that might cause this to be the case.

[bdunham7]

well i hadn't cut anything yet - but what i mentioned was previously tested.

Anyway- i just did what you recommended - pulled the 7824 and reconnected the secondary for that rail and yep ... bulb lit up like the sun - which would have been my prediction i suppose.

one really ODD thing going on here that i'm not sure has anything to do with the problem - but differs from every other PS i've worked on is that I'm getting open measurements IN CIRCUIT with the bridge rectifier - both the old one and new one i put in. when it remove them they test fine. But I've never seen that before - i've always had a fair amount of success testing them in circuit in the past. I can't think of any condition that might cause this to be the case.

Perhaps you have ohmmeter/test lead issues or the capacitor is still charged and interfering.  From your description, either there is a cross-short in the transformer or the bridge rectifier is not installed correctly.  Can you post a clear photo of this thing?

[jaunty]

I dunno - could be the awkward angle i'm working at ... but i doubt there are test lead issues. more likely to be my error but if i'm able to successfully test them out of circuit every time but never IN circuit ...well who knows.

here are some pictures.

overview of rectifier, transformer leads, etc


ok seems i cannot post more than one upload at a time.

[jaunty]

heat sink that holds 7824.