0-30V Stabilized Power Supply

[liquibyte]

Thank you guys for your help.  I made sure that I got sockets as well, just in case.  I've been doing measurements for a few hours off and on and I'm getting a steady 44.6 V out of both sides now.  I feel better now that I know you ran one at 60 V.  I'll go ahead with the build and see how things work out.  Worse case scenario is I'll have to replace op amps and work out some sort of voltage drop later on.  There's room in the case so everything's good.

 

[audioguru2]

Add another 2V for the negative rail on U2 results in 43V across U2.

Using a TLE2141 opamp for U2, a negative supply is not needed. The inputs will work slightly below ground but go down to only 0V and the output will go almost to ground but goes only as low as about +1.1V in this circuit.

 

[audioguru2]

.....work out some sort of voltage drop later on.

The high supply voltage problem has already been worked out for opamp U3 because it uses a -1.3V negative supply. So its positive supply pin is powered through a 10V zener diode which reduces its supply voltage by 10V.
Power your U1 and U2 from this zener diode too to reduce their supply voltage. Check the zener diode for heat.

 

[liquibyte]

The high supply voltage problem has already been worked out for opamp U3 because it uses a -1.3V negative supply. So its positive supply pin is powered through a 10V zener diode which reduces its supply voltage by 10V.
Power your U1 and U2 from this zener diode too to reduce their supply voltage. Check the zener diode for heat.

Like this?

Edit: removed pic to avoid confusion with other schematics.

 

[audioguru2]

Like this?

Absolutely NOT!
You have more than 3A passing through the zener diode because it also feeds the driver and output transistors which do not need their voltage reduced. With a current of more than 3A then the poor little zener diode will blow up like a bomb.

 

[liquibyte]

Yeah, I just realized that mistake.  Leave the transistors on the original input and split off the U1 and U2 through the zener right?

 

[audioguru2]

Yeah, I just realized that mistake.  Leave the transistors on the original input and split off the U1 and U2 through the zener right?

Correct.

 

[xristost]

Another solution of this problem is to wire a couple of power diodes in series after the rectifier and smoothing capacitors. Each diode will reduce voltage with ≈0.7VDC.

I started asking about hooking up the TO-3's over at EEVBlog because it's a much more active site and I was looking for an immediate answer to my question.  I am getting a bit of negative feedback regarding the fact that someone thinks that this design isn't sustainable due to the fact that with a 28V transformer after rectification and smoothing that the voltage into the opamps is going to be around 50V.  Has anyone that's built the 3 amp version run into issues with this?  If so, what sort of solution did you use to fix the issue to get the voltage to less than 44V so as to not overload the max limits on the TLE2141's?.

Edit:  Nevermind.  I hooked up the filter cap and measured 45.1 VDC.  It's a little over the 44V max of the op amps which worries me a bit.  So I guess a new question would be, is there a recommended way to drop that to the 44V?  Voltage divider to ground?

 

[liquibyte]

Another solution of this problem is to wire a couple of power diodes in series after the rectifier and smoothing capacitors. Each diode will reduce voltage with ≈0.7VDC.

I had thought of doing that with some ER502's I have.  I'd like to keep the original circuit if possible and had thought that might be a viable workaround to get just under the 44V.  As it stands now, I get right around 44.6V so it might be a good solution.  Audioguru's suggestion to run through the zener sounds like a good fit too if the thing doesn't get too hot.  I'm assuming that if it does, I might be able to find a component with a higher wattage rating that could handle things.  I was looking at the 1N5925BRLG which is rated at 3 watts but I'm not sure what else to look for in a zener besides voltage and wattage rating.

 

[redwire]

Using a TLE2141 opamp for U2, a negative supply is not needed. The inputs will work slightly below ground but go down to only 0V and the output will go almost to ground but goes only as low as about +1.1V in this circuit.

You are correct.  I  must have been thinking of another circuit.
   
On another note, there is no reason for U1 to need a 41V input rail.  It only needs about 13-14V in order to supply a constant 12.2 V.  Why didn't we connect the positive rail of U1  to the 10V zener like U3 and drop 10V?

 

[audioguru2]

On another note, there is no reason for U1 to need a 41V input rail.   It only needs about 13-14V in order to supply a constant 12.2 V.  Why didn't we connect the positive rail of U1  to the 10V zener like U3 and drop 10V?

Like opamp U2, U1 also does not have a negative supply. With a good 28V transformer, with no load its supply is +40.5V which is less than its 44V maximum.

U1 can be powered from the 10V zener diode together with U3 if you want and if the zener diode is rated at 1W then it will get only fairly warm.

 

[liquibyte]

Like opamp U2, U1 also does not have a negative supply. With a good 28V transformer, with no load its supply is +40.5V which is less than its 44V maximum.

U1 can be powered from the 10V zener diode together with U3 if you want and if the zener diode is rated at 1W then it will get only fairly warm.

If I were to drop in two seperate 12V 1W zeners for U1 and U2 instead of all three going through the 10V would that be ok?  I have two sitting here that could be used for this if it's possible without disrupting the circuit too much.

 

[audioguru2]

With a 12V zener diode reducing the raw positive supply to opamp U2 then your project will not produce regulated +30V at 3A. It might produce regulated 27V with no load, 25V at 1.5A and 23V at 3A.
Its output will have lots of ripple at high voltage/high current.

 

[liquibyte]

So U1 and U3 through the 10V zener.  Maybe a voltage drop with diodes through U2?  I've tried to get a good drop after the rectifier but even with two power diodes I can only get it down to 44.3 volts.  I think the 0.7 volt drop talked about a lot is theoretical and not practical, either that or I have some really efficient diodes.

 

[audioguru2]

The forward voltage drop of a diode depends on its current because it has some resistance.
Without a load on this project its opamps current is only 9mA for U1 and 3.4mA for each of U2 and U3.
At a few mA the voltage drop of a 1N400x rectifier might be only 0.5V.

 

[liquibyte]

I've got it planned out now to run U1 and U3 through the 10V zener and I'm going to drop a volt point something through a couple of 1N4005's in series which should get me under my 44V max on U2.  I've been consistently running tests for the voltage after the smoothing caps and I get between 44.7 and 45 volts so I think this may work.

 

[redwire]

liquidbyte ,  I  I personally think you are splitting hairs on getting to exactly 44V or less  on U2, but if it helps you sleep better, you could  put a 5.6V zener between the positive rail and pin 7 of U2.  Since you need at least one 5.6V zener for the project, it is likely you have a spare.  Hey, if thing don't work out just put a piece of wire across the Zener. 

 

[liquibyte]

I guess I am kind of splitting hairs but I figure it can't hurt.  I already soldered the diodes in so we'll see how it goes.  I appreciate everyone's help though, it is much appreciated.  Actually, the fun part really comes next when I try and design my own displays for this.

 

[redwire]

liquibyte  are you going to surprise us with something like this http://www.4dsystems.com.au/downloads/microLCD/uLCD-43/Docs/uLCD-43-Product-Brief-REV1.2.pdf

 

[liquibyte]

I wish I were that smart.  Just simple 4 digit volt and 3 digit current meters using led displays with max7219's or as1100's and atmega328p's.  I'll just be doing lego block designing, no need to re-invent the wheel.  I'd like to do more advanced stuff but I'm still just learning the basics.