Variable bench power supply problems

[an3s]

Hello all

I used to be a electronic technician 15 years ago. Now I am in IT and want to make

electronics a hobby again. So please forgive my ignorance as I have forgotten

quite a lot. I built a variable bench power supply which works quite well but when

I draw more than 2 Amps the 2N2222 gets very hot and the 2N3055 also failed once.

The components reflect the age I used to build electronic stuff! The supply voltage is 22V. Please have a

look and suggestions/criticisms are very welcome.

[madires]

Please replace the 2N2222 with a BD135 or a similar BJT. The 2N3055 got a low hFE, i.e. you have to drive it with a high current which is too much for the little 2N2222.

[yashrk]

Try using TIP31

[an3s]

Thanks for the reply Madires. I see the BD135 has a 1.5 Amp collector current value where the 2N2222 only has 800mA. I will obtain some and try them. They will need a heatsink as well. What other transistor can I replace the 2N3055 with ? What about a MOSFET ?

[madires]

Yep, a heatsink wouldn't be a bad idea. The 2N3055 is fine.

[an3s]

Thanks Madires
I built this circuit to replace the 2N3055 with a MOSFET IRFP250. But to use it with 24V the gate source voltage needs to be below 20V which is the maximum. How do I clamp the gate voltage to 20V or do I clamp the op amp supply voltage ?

[an3s]

Apologies - I attached the wrong diagram.

[gaijin]

Use a zener diode to clamp the gate under 20v?

[madires]

For protecting the MOSFET's gate you can put a zener between gate and source. I also would add a 100k resistor between gate and source to make sure that the MOSFET is switched off by default.

[an3s]

Hi its me again.
I replaced the 2N3055 with a TIP162 and the 2N2222 with a BD135 and everything is fine. But now when the current drawn is over two amps the LM324 fails. What can be done to protect the LM324 ? Also the supply drops a lot when high current is drawn. The transformer is rated at 17V 4A 70VA. The cap across the 10 Amp bridge is a 10000 MFD 35V electrolytic.

[madires]

Could you please post the current circuit including the power supply of the OPamp, rectifier and so on, i.e. the complete circuit.

[an3s]

madires
The updated drawing as requested. The decoupling cap on the PCB is mounted near the LM324.

[madires]

I'm not sure but the CC LED has no current limiting resistor and might overload the OPamp.

[an3s]

 madires
Apologies - There is a 1K resistor on the PCB in series with the LED. Attached is the updated circuit. Quick question: For protection of the LM324 I powered it with a separate transformer 12VDC power supply but the output voltage does not go past 10VDC. Even a optocoupler to the output transistors did not help. Is there a way to fix this ?

[madires]

Apologies - There is a 1K resistor on the PCB in series with the LED. Attached is the updated circuit. Quick question: For protection of the LM324 I powered it with a separate transformer 12VDC power supply but the output voltage does not go past 10VDC. Even a optocoupler to the output transistors did not help. Is there a way to fix this ?

First the explanation for the low voltage output. The OPamp driving the pass transistors provides the base current, i.e. the output voltage of the OPamp needs to be a little bit higher then the emitter of the pass transistor (= output of the PSU). That voltage difference is caused by V_BE. You got two transistors in darlington configuration and the TIP162 is already a darlington. That means that you got 3 times V_BE. The typical V_BE for a silicon BJT is about 0.6V. 3 *  0.6V = 1.8V, which matches your measurement (12V - 1.8V = 10.2V, without considering the OPamp's max. output voltage).

The solution would be to redesign the control loop to sink the base current of the pass transistor. A simple method is to add a resistor from collector to base to create enough base current to drive the pass transistor for the maximum output current. The OPamp would sink as much base current as needed to regulate the output voltage to the set value. It might be a good idea to let the OPamp drive a small transistor to sink the base current if you run into overvoltage problems for the OPamp.

[an3s]

madires
Thanks for the reply. Do you mean a resistor between the collector and base of the TIP162 or the BD135 ?

[madires]

madires
Thanks for the reply. Do you mean a resistor between the collector and base of the TIP162 or the BD135 ?

BD135, because it's the first transistor of the darlington stage, i.e. BD135's base is the darlington's base.

[an3s]

madires
I did try it, but what I actually after is to get the full 25VDC output and not 10VDC. Will this be possible with the supply of the LM324 at 12VDC ? Will the voltage control loop still stabilize at the supply voltage of 12VDC ?

[madires]

I did try it, but what I actually after is to get the full 25VDC output and not 10VDC. Will this be possible with the supply of the LM324 at 12VDC ? Will the voltage control loop still stabilize at the supply voltage of 12VDC ?

If you change the design of the voltage control loop to sink the base current of the pass transistor instead of providing it, the answer to both questions is yes. Anyway, you won't get the full input voltage (at the filter cap after the rectifier) at the output because of the pass transistor. Please search for some simple lab PSU circuits and try to understand how they work, especially the voltage control loop.

[an3s]

Thanks for all your help madires. I will breadboard the new design and post the results. I still have a lot to relearn especially opamp feedback circuits.

[an3s]

madires
Attached is the modified circuit. I take it this is what you meant by sinking the base current. In all the designs I checked on the Internet I have not seen one like this. Again many thanks. One last question: What is the best configuration for the sense resistor - Connected to negative (Voltage drop referenced to negative as in my circuit) or in series with the load (Needs extra opamp to reference to negative) ? But in any case I personally consider this post to be successfully answered.

[madires]

A low side sense resistor simplifies the current control loop design and allows you to run the OPamp at a low voltage, like 12V in your circuit. A high side sense resistor complicates the design but offers some benefits, e.g. if you're paralleling PSUs. But you can get high side current sense ICs to mitigate the extra work.

PS: The base resistor of the BD135 might be too high.