redwire

0-30V Stabilized Power Supply

Recommended Posts

I finally have my power supply finished! It's on the shelf! I'm doing some full power tests.

I Started with a pair of the Chinese circuit board kits.  Lucky that I found this thread. I made the mods to the boards and added higher power components and  PWM fan control.

My transformers have multiple primary voltage taps. I Used a comparator to select a higher transformer input voltage tap when the power supply  output voltage is low. This greatly lowered the power Dissipation. With commercial power supplies this transformer voltage selection is done on the secondary of the transformer.  With the transformers that I got for free, I could only switch primary winding taps. This required triac switching circuits.

Pictures are attached.

 

PS4 001s.jpg

PS4 002s.jpg

PS4 004s.jpg

PS4 006s.jpg

PS4 007s.jpg

PS4 009s.jpg

PS4 010s.jpg

Share this post


Link to post
Share on other sites

Hello all,

A lot of reading here and much more to learn. Also have the Banggood power supply board, still in the bag though - will likely stay there.

Have ordered PCB's at Elecrow from Gerber files provided by Liquibyte. So, thank you very much Liquibyte.

10PCB + shipping = US$ 18.50 that is about AUS$ 25.00

Since I have a 24V Toroidal already I will have to settle for a 25V 3A power supply. That is fine for me as it is more important

to build it.

Thanks to all here.

Regards, William

 

Share this post


Link to post
Share on other sites
On 06.09.2015 at 8:27 PM, James Bond said:

Hi everyone

Sorry my bad.

There go the schematic and the error. 

I joined the Proteus file too, so you can test it if you have Proteus. 

Best regards 

 

P.S::

Can i change TLE2141 by MC34071 without any problem?

My goal is controll the PSU using arduino, but need put its to work first :)

Final.pdsprj

erro1.jpg

Schematic.jpg

 

On 06.09.2015 at 8:27 PM, James Bond said:

Hi everyone

Sorry my bad.

There go the schematic and the error. 

I joined the Proteus file too, so you can test it if you have Proteus. 

Best regards 

 

P.S::

Can i change TLE2141 by MC34071 without any problem?

My goal is controll the PSU using arduino, but need put its to work first :)

Final.pdsprj

erro1.jpg

Schematic.jpg

 

Share this post


Link to post
Share on other sites

@mars

Hi there,

Look at Page 87 and 88 posts from Liquibyte and you will find PCB, schematic and gerber files to make a nice power supply that will work.

How would I know you might ask. Glad you asked. I have had PCB's made from those gerber files and built a power supply. And it works well.

Look here for my post:

"electronics-lab.com/community/index.php?/topic/40835-0-30v-0-3a-latest-data/&page=2"

All credit goes to member Liquibyte. Thanks a lot Liquibyte for making all files available.

Many thanks also to those who were actively involved in this thread and helped improve the original design.

Hope that will answer your question.

PS. I f you want to built one like I just mentioned, I still have some PCB's left. I only ask AU$2.50 per PCB plus shipping to your country in an envelope.

 

Cheers, William

Edited by repairman2be
Added more info.

Share this post


Link to post
Share on other sites

Hi Liquibyte,

I have bought two of your PCBs from repairman2be (rev. 8.), awaiting them any day now. I have studied the schematic pretty thoroughly and tried to read up on the threads.
Very interesting, and good to have something educating to do with my time :-). So far I have one question that I can't find a comment on in the forums.

First, let me introduce myself. I'm a Swedish 30 year old mmWave/μWave/RF engineer (read: allergic to noise) and I just had a second son so I had to find something to do all the nights I stay up taking care of him. Designing a PSU seems like a good choice!
It started with me realizing that I today financially can handle some of the ideas I had when I was 20, now time is more of an issue instead. But hey, it doesn't matter if it takes time to reach the finish line.

I always wanted to build a boombox. I thought a good place to start was to design a PSU (to not waste battery when a socket is available) and a relay-circuit. I read up a lot on audio PSUs and actually designed a linear 10-18V, 10A supply, and hand drew the breadboard and everything. When I saw the end-price I reconsidered. This PSU is kinda useless for anything but the amplifer, it's better to have a 0-30V supply I can use for lots of things. Especially considering I have no use for or place to store a boombox. The reason I'm saying this is that I might be tainted by audiophiles in my way of thinking :-). That I'm tainted by my RF-world at work I don't see as a problem, RF is probably what I'm going to use it for.

Then I found this design and got interested. But I still wanted it to be able to run the amplifier I had in mind for the boombox, which is something like 14V/6A. I considered paralleling two of these or use a LM723 regulator. I ended up with the idea of two 0-15V LM723 PSUs which I can parallel connect for  0-15V/0-10A or series connect for 0-30V/0-5A. I was pretty happy with this idea and started designing. However, I couldn't get the current limit to work as nicely as I desired for a bench PSU.
Then I came back to this design, and here I am now.

I'll start with building your 3A design pretty much identical to yours. To be able to do some measurements and see how it behaves before I try to improve it :-). After that I will start designing my own board. I'll probably stick to 3A or 5A for a while, if I need an amplifier PSU I'll design it for that purpose. 
My ideas for improvements is to reduce the voltage drop a little, have a look at the negative supply (it hurts my eyes that it doesn't load both lines symmetrically). If I reduce the voltage drop and feel that I can't find an optimal transformer I might reduce the output voltage a little. I feel 30VAC is a little too high and 24VAC apparently is not enough so there is some unused power wasted there. We will see how it goes.

For later I might also take up the idea of LT1236 as reference. We will see. First I believe I have some startup transients to deal with!

So, now to my question:
Your R9 (original R17), is 68 Ohm. The original design says 33 Ohm but you changed it to 68 Ohm. Why did you change that? Have you found some issues with stability if it is too small? The reason I'm asking is that if I lower the current sense resistor I might want to increase this. Maybe I'm doing this wrong. I calculate the original design to give a minimum current limit of 10mA (while it says 2mA in the description), and your gives 20 mA.

Edit: I found the change, it was just to get the voltage divider to give the correct minimum voltage. However I'm still confused, but this can probably anyone answer. It should be simple. I don't get the formula for the current limit.
If we have a minimum of 5mV at the non inverting U3 input, that should allow for a 5mV drop over the current sense resistor right? That's 10 mA... Not 2..

Edit2: LTSpice seem to agree with me...

 

I might have more questions but I'll search and think a little more before I speak :-)

 

Best Regards

Björn

Ps. Thanks for the pdf with QA and other useful info. Would have saved me some time if I read that before I started study the circuit. But imI might learn it better by figuring things out on my own.

 

Share this post


Link to post
Share on other sites

Audio amplifiers do not need a voltage regulator or a current regulator. A car radio amplifier is powered directly from the car battery with no voltage regulator but the low current radio parts might use a simple little voltage regulator.

14V at 6A is 84W!  One amplifier produces a maximum output of 16W into a 4 ohm speaker then 4 channels make 64W. If all 4 amplifiers are producing 64W then if they are linear they heat with an additional 36W and the current is 7.14W but it does not need to be regulated.

 

 

Share this post


Link to post
Share on other sites

Yeah but it's a lot cheaper to use a regulator rather than quality caps to get rid of the 100 hz ripple :-)

Plus I could make sure it runs on 14 V regardless of load (volume, usb charger etc) and temperature!

The two amplifiers I had in mind were 45W and 50W into two channels, with losses and other functionality I estimated 6A.

Share this post


Link to post
Share on other sites

I could add that when I was looking for a suitable amplifier I didn't know electronics-lab even existed. So I just browsed for the most suitable class d chip and went on from there :-)

It was when I gave up on the 14V PSU idea because of a stronger need for a lab PSU that I found this design and this place!

Share this post


Link to post
Share on other sites

Hi  Repairman

I am interested in Liquibyte Rev-8 PCB.  Do u have any left?

How to go ahead about ordering from u?  Any idea on the shipping costs from ur location to India (Kerala\Trivandrum )?

I understand u are charging 2.5 AUS$ per PCB. I want to know total costs before commiing.

rgds

Share this post


Link to post
Share on other sites

I vouch for @repairman2be 

I got one from him recently and it was very neatly packed and arrived on time. Check out the pics below. If he still has some, you can get them from him if you only need one board. Otherwise, organize yourselves for a batch order from the fabhouse of your liking.

IMG_1876.JPG

IMG_1878.JPG

Share this post


Link to post
Share on other sites

So, I ended up getting the PCB:s from Repairman (great service there) and have put one together, I do have a weird issue someone might be able to help me with:

The circuit the PCB is based on is the Liquidbyte version from the top of page 88 of this thread (http://electronics-lab.com/community/index.php?/topic/29563-0-30v-stabilized-power-supply/&page=88). The things where I deviate a bit are:

  1. The ballast resistors for the 2N3055's are 0.1 Ohm rather than 0.33 since I didn't have those available at sufficient power rating
  2. The sense resistor is 0.46 Ohm: Rather, it's a small network of two 0.56 Resistors in parallell after which there is a single 0.18 one in series, yielding 0.56/2+0.18 = 0.46 Ohms, again for availability
  3. The single 10 uF Polyester cap was also lacking, so it's replaced by two 4.7 uF ones in parallell, they turn out to measure almost exactly 10 uF
  4. My rectifier bridge is only rated at 12 Amps at 250 V, this should be sufficient though
  5. The Transformer is a 30V/4A, yielding the same 120VA

Now, when I fire the thing up, I can adjust voltage fine from about zero (haven't been able to calibrate it yet) but only up to 17.56V (which may not be that exact given it's from my multimeter). Turning the trim pot adjusts down from this fine, but when adjusting up it goes to those 17.56 and stops. Now, if I disconnect the voltage adjust pot entirely the output shoots up to 40V, connecting it back makes things stop at 17.56 again. Have looked things over and can't find anything obviously wrong, maybe someone has an idea to what may be happening would be greatly appreciated!

Cheers

Share this post


Link to post
Share on other sites

Hi!

I have bought the same board from the same man :-) but not built it yet. Ive done lots of simulations though.

I guess this is unloaded? Repairman told me about the 100k trimpot, you need to short pin 1-2 there, did you do that? 

 

Share this post


Link to post
Share on other sites

Whats the voltage drop over your 0.46 ohm sense resistor?

It is past bedtime here, can think more tomorrow. But I think it could be current limit.

When you remove the voltage trim pot you get really high gain, which means that the input voltage at the rightmost op amp can be low (caused by current limit) and still provide a high output voltage.

If the 100k current limit trimpot is connected wrong it will always be 100k. This will mess up the current limit op amp.

So I hope the solution is to put a bridge/jumper/short between pin 1 and 2 on the 100k trim pot.

Share this post


Link to post
Share on other sites

Unloaded or loaded with a minimal load (a 1k resistor and a led) same thing. Voltage drop is about 180 mV, and I don't think it's the current limiting, the current limiting led doesn't turn on, which it does if I set it low enough to kick in on my small load, so that part of the circuit seems to work fine. And yes, the bridge on the 100k trimmer is in place.

I did notice something while probing the thing though. As expected, the voltage over D6 is 5.6V. However, and thus over R14, the non-inverting input to U2 is half, at 2.8V, making it's output 5.6 and thus the output of U3 roughly three times that at max voltage adjustment, which is those 17.5V I get at the output, and it can of course never go above that. Now, the reason for that U2 input beats me.

Share this post


Link to post
Share on other sites

The middle opamp in Liquibyte's schematic is the 11.2V reference because it uses a 5.6V zener diode and it has a gain of 2 times. The voltage setting pot R27 is simply a voltage divider that feeds 0V to 11.2V to the non-inverting input pin 3 on the opamp on the right side. The gain of this opamp is 1+ (56k/28.8k)= 2.9444 times so that the maximum output of the circuit is 11.2v x 2.9444= 30.0V. Your middle opamp is messed up since its output voltage is too low.

Share this post


Link to post
Share on other sites

Great, we are on to something!

If R10 in liquibytes schematic is shorted, this would happen. Can you measure it's resistance?

R10 is the 1k resistor which sets the 5mA current for the reference zener.

Share this post


Link to post
Share on other sites

Sorry for not getting back, been out of town. Turned out the Op-Amp was defect; put a fresh one in and it worked like a charm. Thanks for the help

Now, another question: As audioguru says above, the voltage pot just feeds 0-11.2 V to the right Op-Amp. I wanted to try to control the voltage from another source, which I can't really get to work as I want: From 11.2~8.2V it works fine and the output adjusts down perfectly. However, no matter how much lower I try to feed it, it won't get lower, actually stopping the input source from going any lower. As soon as I shut the supply off or disconnect it jumps back down. Interestingly, with a multimeter set for Voltage measuerment between pin 3 (Non-inverting in) and pin 4 (ground) it does suddenly work as expected and adjusts all the way down to 0, as soon as I remove it it's back to ~8.2. I'm sure I'm missing something obvious here..

Share this post


Link to post
Share on other sites

The TLE2141 opamp has PNP input transistors so that the outputs work with an input as low as 0V. Then you must pull the input down and not let it float up. Your multimeter pulls the input of the opamp down. The maximum input bias current for a TLE2141 opamp is 2.1uA that tries to pull the input voltage up. 

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now