0-30 Vdc Stabilized Power Supply

[Tinny]

Hi,

I've just finished making this PSU, and even tho it seems to run ok. a bit hot atm but ill pick up some bigger heatsinks soon enough. I want to learn a bit more about how it works.

eg

How to work out how much heat things might generate and what heatsink is going to work best

D8 is a 5v6 zener but seems to have around 5.8V. will it change much if i find a way to make it the correct voltage? and why, D7 isn't quite right either.

Is ther a way to work out the efficency of a circut ?

This might be a bit off topic but if somone can point me to some info about this kind of theroy.

I have checked out the theroy article forum. well i still am

Thanks

 

[audioguru2]

How to work out how much heat things might generate and what heatsink is going to work best

Heat is caused by power dissipation. Power= volts x amps. You have a 24VAC transformer that has a peak voltage of 40.0V. The rectifiers will drop it to an unfiltered positive supply of 38.0VDC at full output current. If your transformer is rated at 120VA then it will be able to supply the 40V peak voltage to the rectifiers at an average current of 3A.

If the voltage is set low and the current is set at 3A then the output transistor will have 38V across it and 3A through it. Then its power dissipation is 38V x 3A= 114W.

The datasheet for a 2N3055 power transistor rates it at 115W when its chip is at its max allowed temperature and its case is held at only 25 degrees C somehow. If its case gets warm then its chip will have over-temperature and it will fail. You cannot buy a heatsink that holds the case of a transistor at 25 degrees C. The transistor will get warm and cannot dissipate its full rated power.

Heatsinks are rated in "degrees rise per watt" where a very big one will have a rating of 1 degree temperature rise per watt. A 2N3055 transistor mounted on one without an insulator will be able to dissipate a max power of 56.7W when its chip is at its max allowed temperature.
A fan would allow more power to be dissipated.

D8 is a 5v6 zener but seems to have around 5.8V. will it change much if i find a way to make it the correct voltage? and why, D7 isn't quite right either.

Most zener diodes have a tolerance of 5%. Then the voltage for 5.6V ones could be from 5.32V to 5.88V. It doesn't matter in this project because the pots have a tolerance of 20%.

Is ther a way to work out the efficency of a circut ?

Efficiency is defined as the ratio of the power output divided by the power input.
If you short the output when it is set to 3A then its efficiency is zero.

 

[halen]

It is too bad that you can't get stuff in your country.

I just order parts from a neighbor country (USA) online and they are delivered to me the next morning.

I can buy: LF356N, OP37, OP07, NE5534. It is possible to use one ot those?

EDIT: I can buy OPA604 - 48V supply voltage

 

[audioguru2]

You can look up the spec's for those old opamps to see if they will fail with a supply voltage from this project that is higher than the max voltage allowed for them.
The OP37 and NE5534 are de-compensated and therefore would oscillate in this circuit.

Since you cannot get modern parts then use a transformer with a lower voltage like 20VAC which would give a unregulated positive supply to the opamps of 29VDC without a load, plus the negative supply of 5.6VDC so they will have a total supply of 34.6V which is less than the 36V max allowed for ordinary opamps.
The project will then have a max output voltage at 3A of current of only about 20VDC.

 

[x_dadu]

I just find out about multimmeter with Atmega8 and I'll do soon. After that i find out about this 0-30vcc psu and I see that is perfect for mai multimmeter  but is something that i dont like :
-with 24Vca transformer the Vcc result will be 34Vcc
-D5-D6-C3-R3-D7 make a negative -5,6V power unit
U3 and U2 are bouth powered between +34 and -5.6. Is'nt that 39.6 ? Is'nt that dangerous for TL081 who has 36 (+/-18in datasheet) up limit ? Or maybe something I dont get'it . Thk's

 

[audioguru2]

i find out about this 0-30vcc psu and I see that is perfect for mai multimmeter  but is something that i dont like :
-with 24Vca transformer the Vcc result will be 34Vcc
-D5-D6-C3-R3-D7 make a negative -5,6V power unit
U3 and U2 are bouth powered between +34 and -5.6. Is'nt that 39.6 ? Is'nt that dangerous for TL081 who has 36 (+/-18in datasheet) up limit ?

Yes, the opamps have a supply voltage that is higher than their max allowed voltage. They will fail.
Some of the parts get too hot. They will also fail.

That is why I recommend using OPA445AP high voltage opamps, then the transfomer's voltage can be increased to 30VAC so the project's output voltage can reach 30VDC at a current of 3A.

 

[kcc]

Let's try something different, let's "back into" the circuit to find DC gain (the circuit has AC gain as well).

Hello Again!

Took me a little while to digest the latest post but I think now that I am set!  The equations made it all clear and helped tremedously!

One of the most confusing aspects to my understand went away when I (finally) realized that the Q2/Q4 transistors were configured as a Darlington pair and therefore had a Vgain near unity!

Thanks so much for the patience and guidance offered!

I'm off to build a power supply!

Kevin

 

[Tinny]

Hi,

I am using HA17741 op amps which are rated at below the almost 40V supplied to them. for a 36V op amp. tho replacing them for the TL081 wouldn't really fix that would it, they are 36V also? 

Which is why they failed. or so i think they did. not sure how to test that. U1 and U3 have voltage but point 3 and 4

and U2 ther is no voltage. maybe i can trace the circut back from U3 untill i find the voltage?, donno heh

OPA445 are $34.10 each. thats around $15 - 20 USD. does anyone know of somthing rated between TL081

and OPA445. thats between 36V and 50V? I did find some US suppliers for cheaper then that but then the postage kinda makes it pointless

If not i guess its the OPA445, in the end i don't want somthing thats going to just melt.

 

[audioguru2]

There are a few opamps rated at 44V:
TLE2141 and MC34071.

Why are your prices nearly double the price in the US?

View attachment 40402

 

[Tinny]

yea im not sure,

im in new zealand, but even things like TL081 which i found in the end. for $1.61NZD. thats less then $1 US and the 2n3055 for $2.20NZD. maybe the OPA445 is just rare

and thx. ill try one of the 44V op amps

 

[x_dadu]

has someone redesign PCB for duble operational amplifier like MC34072 ? or even quad ? I can't find 34071 or TLE2141 in my place and OPA is to much expensive here ... but I have a few 34072 ...

 

[audioguru2]

U2 needs to be a single high voltage opamp because it has input offset voltage adjustment pins for RV1. Then the value of RV1 and R10 must be adjusted to match different opamps. Some opamps need RV1 to connect to the positive supply instead of the negative supply.

 

[halen]

You can look up the spec's for those old opamps to see if they will fail with a supply voltage from this project that is higher than the max voltage allowed for them.
The OP37 and NE5534 are de-compensated and therefore would oscillate in this circuit.

Since you cannot get modern parts then use a transformer with a lower voltage like 20VAC which would give a unregulated positive supply to the opamps of 29VDC without a load, plus the negative supply of 5.6VDC so they will have a total supply of 34.6V which is less than the 36V max allowed for ordinary opamps.
The project will then have a max output voltage at 3A of current of only about 20VDC.

EDIT: I find UA741 OP Amp, it is 44V, its not decompensated....

Ok I try it with some 44V OPAmp.  I can buy OPA445 now, but for $14.11 per unit :-D Its wery expensive, so its better to buy professional lab power supply.

 

[valorous]

Hi
I've got a bridge rectifier module GBPC2502P 25A 200V. Can I use it in this circuit?

Also can I make some measurements on the circuit before placing the op-amps 445AP. It was very difficult to get them so I don't want to burn them. I'm looking for expected voltages at different test points in the circuit.

Thank you for your help.

 

[audioguru2]

I've got a bridge rectifier module GBPC2502P 25A 200V. Can I use it in this circuit?

It is bigger than is needed but it will work fine.

Also can I make some measurements on the circuit before placing the op-amps 445AP. It was very difficult to get them so I don't want to burn them. I'm looking for expected voltages at different test points in the circuit.

The 30VAC transformer will be about 31.5VAC without a load. Then its peak voltage is 44.5VDC and the rectifier bridge will charge C1 to +43.1V which is the positive supply for the opamps.
The negative supply should be close to -5.6VDC.

If anything shorts the output of an opamp then it will get pretty hot. Double check which pin is which on the Q2 and Q4 transistors.

 

[subodhthok]

I'm Kinda confused :-\...... though this seems to be a good project!
Can somebody gimme the Block diagram!
Thnx in advance!

 

[audioguru2]

I'm Kinda confused :-\...... though this seems to be a good project!
Can somebody gimme the Block diagram!

It is just a power supply that has voltage regulation and current regulation.
Here is a simple block diagram. What is confusing you?View attachment 40403

 

[adiankhoo]

hi everyone,
its me again. I have a ques, can the 12,000uF, 63V is replaced with a 18,000uF, 50V. i cant find one with 63V. and the other i can find is 8000uF at 50v and 100V.
u suggest parallel-ing the 8000uF o use a single 18,000uF 50v? but i feel the 50V would do good enough.
i would like to thank for any comment or suggestion. Im cant access the forums the office in PC. so, Thanks again in advance.

 

[audioguru2]

Hi Adian,
The peak voltage from the 30VAC transformer is only 44VDC when there is no load. A 50V capacitor should be fine. I selected a 63V capacitor because it was available on Farnell's site.

You can use any value over 10,000uF. You can parallel capacitors for a total over 10,000uF if you want.

 

[Aan]

Also read this: http://www.electronics-lab.com/forum/attachments/Digital_Panel_Meter.pdf

It's true that you need to connect a load to the output. Read this to understund how to current limiter works.

I can't access to this file http://www.electronics-lab.com/forum/attachments/Digital_Panel_Meter.pdf

Regards,