Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

new here psu design question

Status
Not open for further replies.

adamjclark

New Member
hello all, This is my first post here. I am pretty new to electronics and I have a very basic understanding. I am building my own psu. Heres what I want. I would like to have a 1.5-30vdc 5a dual supply with adjustable voltage and current. I want to use as much stuff for it as I can that I already have. I made a schematic that uses 4 lm317's with pass transistors and an lm308 for the voltage regulator circuit. My transformer has 3 sec. coils. the HV sec. with center tap is 44.5vac, the other 2 doesn't have center taps and one is 38.5vac and 8vac. Im using a mb354-w bridge rectifier that is 40-100v at 35a. after the rectifier the HV sec is 62vdc. this is where the problem starts. the max input for the 317's is 30vdc and the 308 is 18vdc. I need to step down the voltage from 62 to 30 without limiting the current. I have a few other issues but this is the first and most important one I need to get past before I can move on. Thanks
 
hi and welcome! first a warning i am very young and pretty new to electronics myself so dont take what i say as gospel, but from a quick read of the datasheet i cant see a problem. from the datasheet it says
because its floating then it can regulate supplies of several hundred volts as long as the differential voltage isnt exceeded, it also says in the sheet that the differential voltage is 40V max, so to me that says you can stick 60 volts in one and drop it to say 30V then into another to drop the 30V to whatever you want, but i might be wrong also look at the LM138 wich can handle 5A on its own,
anyway wait for one of the clever people as i might have it wrong
 
I was thinking the same thing. the reason for the lm317's and not a lm138 is that I want to be able to get down to 1.5vdc. the 138 cutoff is 2.85vdc
 
ahh ok i didnt check the 138 datasheet :D, i tried making a variable supply with them a while back but alot lower voltage, was ok my main problem was the pot was great, i now have some precision multi turn pots (i got from some old equipment i found in tip), i dont actually need a psu now as i have use of two HP ones that are fab, but i might try again and make one with these pots! also at some point i want to try tweaking a ol ATX psu, i want to try altering the voltage by altering the pwm output on the chip inside. but thats for another day as i have loads of projects on the go and i want to get some finished first :D
good luck with the build as its alot of fun building your own stuff
 
I do not know why you have 62VDC. You need about 35VDC for the positive supply and another 35VDC for the negative supply.
The if the output voltage is set at 1.5V and the load draws 5A the output devices heat with (35V - 1.5V) x 5A= 167.5W which is a lot of heat in output devices. You will need monster heatsinks and fans.

Some opamps work with a supply as high as 44VDC.
 
Hi Adam,
So you want to get started, and, presumably, using your transformer.
Not sure what current ratings apply to each winding, so be patient.
Most electronic stuff will require a power supply voltage of around 5 volt or less for logic circuits, and around 24 volt or less for general purpose stuff.
The 8.5 volt winding suggests a full wave bridge rectifier to produce around 10 volt DC which can then be regulated with a 'regulator' to give 5 volt or so. maybe even a +/- 5 volt supply.
The 44.5 volt centre tapped winding suggests a full wave rectifier to give around 28 volt DC which can be regulated to give up to say 20 volt DC.
The 38.5 volt winding is a puzzle and full wave bridge rectified will produce say 50 volt DC. This could also be +/- 25 volt.
Logic tends to use + 5 volt; Linear op amps tend to use +/- supplies up to +/- 15 volt DC.
So the traffo you have is a good basis for a general purpose bench power supply.
Its up to you now to say what you would like to have as a finished product.
I can say that you could have a power supply with:
0 to 8 volt Dc variable;
0 to 25 volt DC variable;
+/- 25 volt DC variable.
How does this sound?.
In the short term, forget about the iC's you have; just concentrate on the ultimate capability of a power supply based on the transformr.
 
Ok. heres what I was thinking, using only the HV sec. with the center tap. The ac voltage between the center and either leg of the coil is 22.6vac which works out to around 31.7vdc. I have a 40-100v 35a bridge rectifier im using. If I use the center as common and one leg of the coil to an lm317t for the +v and the other leg to an lm337 for the -v for a dual supply. I will be using 4 lm317's parallel with pass transistors to increase the current capabilities to 6a which I plan to make adjustable and limit to 5a. this would give me a +/- adjustable 1.5-30vdc 5a supply. then to further it, Im using 3 icl7107's and 7 segment displays to make digital volts, amps, and temperature meters. Well,, at least that's the plan.
 
An LM317 needs an input at least 2V more than its output. A boost transistor needs an additional 1.5V. The rectifier drops about 1V. Ripple from the rectifier reduces the raw DC about 3V. Then your maximum output is 31.7V - (2V + 1.5V + 1V + 3V)= 24.2V at high current or a little more at low current.
 
well.. I figured there would be some voltage drop, so I guess that would be acceptable. As long as I can get 5a of current between 12-14vdc, Then im golden.. Anything extra is a plus.
 
As long as I can get 5a of current between 12-14vdc, Then im golden.. Anything extra is a plus.
The "extra" voltage simply makes a lot of heat that you need to get rid of with huge heatsinks and maybe a fan.

If your unregulated voltage is 32V and the load uses 12V at 5A then the load heats with 60W and the power supply heats with (32V - 12V) x 5A= 100W.
 
I was looking at the 317 datasheet and I can get 3-4 amps from a pair of them paralleled with pass transistors, I think I could live with that. I am in the mobile audio industry and I have to bench test a lot of amplifiers, so I need a solid 2a with a little room for surge.
 
I was looking at the 317 datasheet and I can get 3-4 amps from a pair of them paralleled with pass transistors, I think I could live with that. I am in the mobile audio industry and I have to bench test a lot of amplifiers, so I need a solid 2a with a little room for surge.
A car battery is about 13.8VDC. A bridged class-AB amplifier with the 13.8V supply driving a 4 ohm speaker and the amplifier is barely clipping produces an output of about 15W with about 13W of heat so the total power is 28W. If the amplifier is stereo then the power supply current must be (28W x 2)/13.8V= 4.06A.
If the amplifier has 4 channels then the power supply current is 8.12A.
 
that seems about right, but I often deal with 4-5k class d monoblock amps that can easily surpass 150 amps at full volume at .7 ohms. but to test an amp it actually requires few amps to turn it on and play a tone from a sine wave generator to test the power supplies and output. the output of the amp plays through a 4 ohm 4 inch speaker at low volume/power. So 2-3 amps would be sufficient for this.
 
How can you test an amplifier at only a few percent of its rated output power?? Through a dinky little speaker!
 
Adam,
Going back to your post #7, I can relate to your wishes.
The problem with a 'Universal' power supply of the type you are intending is that it is difficult to design a 0-30 volt 0- 5 amp power supply because a linear supply generates massive heating in the pass transistor when high current at lower voltages are required. There is another thread about using LM338 as a series pass regulator. I suggest you wade through that thread and then try to do the thermal design for what you want to do. Doing what you want will run you ragged with how to get rid of too much heat. Alternatively, your transformer is a good one because with it you could design a very versatile wide range power supply but it would have different sections; viz, a high current low voltage section, a high voltage medium current section, and a balanced =/- supply for operating operational amplifiers and other stuff needing a balanced supply.
 
Status
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips

Back
Top