[Jam_Master_J]

Hi,
I'm building a 24V STEPs (in it's own enclosure) that I intend to use to power headphone amps (MH for now) and a DAC (Monica 2, project next in queue). The Monica 2 requires 12V DC, and thus I was thinking of using a simple voltage divider. I've seen a few posts though claiming that a regulation circuit is superior but it seems kind of silly to burn off 12V on the regulator.

2 main questions)

1. Is a simple voltage divider acceptable in this case, or are active components the way to go? I want to do this right, so if an LM3xx circuit is a better choice I'll do it.

2. If I use a simple voltage divider, how shall I select the resistors? Looking at the circuit, it seems to me that lower value resistors will reduce the effect that R-load has on the output voltage.

 

[amb]

If it seems silly to use a regulator to "burn off" 12V on the regulator, then it's just as silly to do so with resistors. The same amount of heat has to be dissipated either way. What's worse, with resistors you lose load regulation, the "12V" output won't stay at 12V -- it will vary with the current draw. And how much current does the DAC draw? If it's significant then you're much better off using a separate PSU with a lower voltage transformer.

 

[Jam_Master_J]

Thanks. Big goof on my part thinking that the resistors wouldn't burn the 12V too (bad EE student, bad). I'm just gonna throw a 5V TREAD into there as well and know that it's getting done right.

 

[Paragon]

Since the step voltage is set via two resistors, why not have a switch to switch between 24V and 12V? Can't view the step page but it should work. A few wire jumpers and a DPDT on-on switch should do it.

 

[Paragon]

R3 can remain on the board as 120Ω
R4 is what you need to change.

All caluclation are theoretical according to the LM117 datasheet. Vout=1.25(1+R4/R3)

The total resistance with the VSET trimmer for 24V should be around 2184Ω and for 12V, around 1032Ω
The way you have it set now R4 is 2KΩ so the 184 is taken up by the RSET trimmer. If you had a switch to change R4 to a 1KΩ resistor, you would end up with around 13.6V output without changing the VSET trimmer.

You could get a 1.05KΩ resistor for R4 and remove the VSET and jumper it to get around 12.2V and a 2.2KΩ for R4 to get 24.2V

Since you only need to switch one resistor, you can get a SPDT on-on switch.
There are two pads for the R4 resistor. The R4 pad that connects to the LM317 would have two jumpers out of it. Each jumper would have a resistor in series before it gets to the outer spots on the switch. The one that connects to ground should be jumpered to the center (Common) tap of the switch (either the other pad of R4 and jumper VSET or the ground pad of Vset or the closest circuit ground to the switch)

[three switch positions below for SPDT]
------
*1.05K
Ground
*2.2K
------
Where * connects to the pad to the LM317.

<*> You will still burn the 12V of out of the LM317 but it is an easy way to get 12V out of somethign that you already have set up with addition of a switch, few resistors, and some wire. Just a matter of finding the resistors. For people in the US, 2.2K and a combination of 1K and 47K 1/2W resistors from radioshack would work. Not sure what you have around in Canada to get resistors to test it out.

 

[Jam_Master_J]

Quote:

Originally Posted by Paragon /img/forum/go_quote.gif Since the step voltage is set via two resistors, why not have a switch to switch between 24V and 12V? Can't view the step page but it should work. A few wire jumpers and a DPDT on-on switch should do it.

But I want 12V and 24V output at the sametime.

 

[AndrewFischer]

You have a few options.

Voltage divider using resistors. Simple, cheap. Unregulated. Low efficiency.

Use a second linear regulator. A TREAD would be an easy way to get this. Leave out the diode bridge and big input capacitor. Put the your +24 into the test point. Simple. Not too expensive. Low efficiency.


How much current does your DAC require?

You need to drop 12V. You will be dumping as much power into the regulator as you use in your DAC. If the DAC uses a significant amount of power the 12V regulator will get very hot. Maybe too hot.



If that's the case you can either use a switch mode regulator or give in and build a second linear power supply.

 

[Paragon]

Quote:

Originally Posted by Jam_Master_J /img/forum/go_quote.gif But I want 12V and 24V output at the sametime.

Doh.. yeah.. I see that now. Amp and DAC.. my bad.

How about a rail splitter so you have +12 and -12 from the 24V. You shouldn't be burning much power since you still have 24V and you can just use the positive rail. (can technically use the negative rail for positive voltage also) The only thing is that I don't know how splitters are if you are using only one rail. (stability)

rail splitter can be done with an opamp or one of those TLEs depending on how much current you need. I wouldn't think the DAC would take too much current.

Just another shot in the dark.. probably wrong.

 

[Jam_Master_J]

Quote:

Originally Posted by Paragon /img/forum/go_quote.gif Doh.. yeah.. I see that now. Amp and DAC.. my bad.

No problem, I appreciate the help though.

Andrew,
I'm going to do the TREAD option. I don't anticipate the DAC drawing too much current so I think the heat can be managed.

Is there any reason not to voltage divide to say 15V and then regulate that? That way more of the heat is diverted through resistors than the regulator. Not too worried about efficiency, performance is more important.

 

[AndrewFischer]

Quote:

Originally Posted by Jam_Master_J /img/forum/go_quote.gif Is there any reason not to voltage divide to say 15V and then regulate that?

You are increasing the input impedance for the regulator. Worst case the regulator will drop out on a transient. If the DAC is a mostly constant load, this is not likely to happen.

To drop a volt or two, you can put couple of TO220 package diodes in series.

I'd only do the resistors if the LM317 on the TREAD gets really hot.

 

[Paragon]

What about splitting the 24V into a dual +-12V rail? Would that be the same as using a resistor voltage divider?

 

[Jam_Master_J]

Quote:

Originally Posted by Paragon /img/forum/go_quote.gif What about splitting the 24V into a dual +-12V rail? Would that be the same as using a resistor voltage divider?

I think the rail splitter is doing the same thing internally as a voltage divider circuit does.

 

[Paragon]

I had a feeling as such.

http://tangentsoft.net/elec/vgrounds.html

 

[AndrewFischer]

I just assembled this for work. 2 Gel Cells glued together and wired as a split supply. It powers a 32 channel EEG amplifier.