Can I do this to make bridged DC supplies?

[agent420]

I'd like to have two seperate potentials relative to ground using a 12.6V CT transformer. One will be used for 5V with a 7805 and the other 12V using a 7812. The ground is common to both circuits.

I've never seen a center tapped transformer used in this manner, but in the simulator everything appears to work correctly without issues. The center-tapped lead won't require a diode because it obtains recitifcation through the outer winding legs going to ground.

Will this work in reality or am I missing something?

**broken link removed**

 

[Nigel Goodwin]

It's no different to a normal bridge rectifier giving a spilt supply, which is all you need to use - you've just drawn it differently!

 

[agent420]

It's no different to a normal bridge rectifier giving a spilt supply, which is all you need to use - you've just drawn it differently!

But in most cases I can think of the split supplies with a bridge result in positive and negative rails, here I am trying to get two positive rails. Or is that the same?

[edit]

Also, aren't the split rails normally the same potential? This is what I think of split bridged, but I don't think it's the same as the circuit above...

**broken link removed**

 

[Roff]

I think what Nigel is saying is that you have just redefined the reference node (GND). Draw the schematic without a GND first. You can then put GND on any node. The voltage difference between any two nodes will remain the same.

 

[Nigel Goodwin]

I think what Nigel is saying is that you have just redefined the reference node (GND). Draw the schematic without a GND first. You can then put GND on any node. The voltage difference between any two nodes will remain the same.

Yes, it's all exactly the same circuit, for the last diagram simply re-label Neg as ground, ground as 5V, and Pos as 12V (obviously NOT the actual voltages, but the regulators they feed).

 

[agent420]

I see. Thank you both for the insight.

 

[Roff]

I see. Thank you both for the insight.

Give yourself credit for figuring out how to do that, though. I don't think it was an obvious solution, except in hindsight. I have never seen it done that way before. Perhaps Nigel (or someone else) has.

 

[Nigel Goodwin]

I see. Thank you both for the insight.

It's just a question of understanding what's going on, and it's quite a difficult concept to grasp for beginners - that you don't have any absolute voltages, they are all relative to something else, when you're measuring with a meter it's relative to the second meter lead, so it depends where you put it!.

Generally you measure with reference to ground or zero volts, but not always, and it's only a convention - nothing else.

 

[Nigel Goodwin]

Give yourself credit for figuring out how to do that, though. I don't think it was an obvious solution, except in hindsight. I have never seen it done that way before. Perhaps Nigel (or someone else) has.

Yes, I've seen it before, but not drawn the way he did, it took a couple of seconds of mental 're-drawing' to realise what it was.

 

[agent420]

Yes, I've seen it before, but not drawn the way he did, it took a couple of seconds of mental 're-drawing' to realise what it was.

Maybe I can patent the idea before anyone else notices!

It's always interesting figuring out on your own a law or circuit approach. Often I have a much better understanding of it that way compared to simply reciting it from a textbook.

[edit]

Just thought of it, this situation is just like the 7V trick some pc mod people use to obtain a lower fan voltage from referencing between the 5V and 12V rails.

 

[zevon8]

One guy I worked with used to draw schematics like that alot. It would drive you crazy trying to work it out sometimes. For him it was natural and made sense, but I was used to "seeing" it drawn like the second diagram. The trick was, as Nigel pointed out, finding the node that measurements were relative to. What complicated it ( for me anyways ) was that this coworker drew the AC buss as a parallel lines across the page, with branches coming off either vertically up or down with the diodes in all sorts of odd directions feeding the various secondary ciruits.

 

[Nigel Goodwin]

For 'old' people, the early transistor circuits were usually +ve ground, and drawn with the negative rail at the top - basically because the only transistors were PNP. But I always found it hard to look at them!.

 

[audioguru]

For 'old' people, the early transistor circuits were usually +ve ground, and drawn with the negative rail at the top - basically because the only transistors were PNP. But I always found it hard to look at them!.

Here is a transistor circuit that is 41 years old. They have it drawn wrong with a positive supply instead of a negative supply.

 

[Nigel Goodwin]

Here is a transistor circuit that is 41 years old. They have it drawn wrong with a positive supply instead of a negative supply.

Yes, drawn it wrong - and even the resistors look strange

But OC71's were fine transistors, general purpose PNP audio, bit small for an output transistor though? - later on the OC72 became available, and was capable of more power.

 

[audioguru]

I liked the powerful (for small transistors) complementary germanium AC127 and AC128 transistors. They were in a glass case, surrounded with thermal grease and had an overall small metal case. I guess in those days they didn't know how to bond the chip to the metal case then seal it.

 

[Nigel Goodwin]

I liked the powerful (for small transistors) complementary germanium AC127 and AC128 transistors. They were in a glass case, surrounded with thermal grease and had an overall small metal case. I guess in those days they didn't know how to bond the chip to the metal case then seal it.

But they were replaced by the higher spec AC187 and AC188, there were also AC187K and AC188K versions, these had a square metal casing with a hole through so you could bolt them to a heatsink!.

 

[agent420]

Perhaps another question... The original intent of this circuit was to minimize the inefficiency of the 5V regulator by reducing the input from the 17-18V to 8V or so. Currently I am using a 7805 on the 18V rail and it is getting hotter than I think it should, though the current draw is only 150ma or so. I suspect this is because of the high voltage drop of 18V-5V.

I have seen other regulator circuits where the dc is rectified by a simple bridge and then fed to a 7812; which in turn provides the input for a 7805... Would that be a better or worse solution to the 1st schematic firgure I posted?

 

[Nigel Goodwin]

Perhaps another question... The original intent of this circuit was to minimize the inefficiency of the 5V regulator by reducing the input from the 17-18V to 8V or so. Currently I am using a 7805 on the 18V rail and it is getting hotter than I think it should, though the current draw is only 150ma or so. I suspect this is because of the high voltage drop of 18V-5V.

I have seen other regulator circuits where the dc is rectified by a simple bridge and then fed to a 7812; which in turn provides the input for a 7805... Would that be a better or worse solution to the 1st schematic firgure I posted?

Depends on what's been drawn from the 12V rail, the heat HAS to go somewhere, in this case it would be shared between the two regulators, but the 12V one already has it's own heat to deal with.

If you've got a centre tapped transformer then using it to give two rails is a good idea, it reduces the heat loss - at the cost of adding a single electrolytic.

 

[agent420]

Thanks for the reply and confirming my suspicions. I think I'd rather have the 12V rail seperated from the 5V logic, because I'm using it to power 2 small relays and the pwm power source for an incandescent lamp dimmer circuit. Even with a 3 terminal regulator and snubber diodes, I'm sure the 12V rail will have transients better left off the logic supply line.

 

[Nigel Goodwin]

Thanks for the reply and confirming my suspicions. I think I'd rather have the 12V rail seperated from the 5V logic, because I'm using it to power 2 small relays and the pwm power source for an incandescent lamp dimmer circuit. Even with a 3 terminal regulator and snubber diodes, I'm sure the 12V rail will have transients better left off the logic supply line.

Sorry for the thread 'drift', but yes you are better doing it that way, and it's certainly a more elegant and superior design.