Queries regarding Digital power supply design

Go To Last Post
11 posts / 0 new
Author
Message
#1
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I was designing a 0-35v/6A power supply with keypad entry with ATMEGA32.The main idea is to output calculated output voltage by a 12 bit DAC and feed it to an op-amp configured as non inverting amplifier so the op-amp can take care of the control loop. The targeted design was working in real life with an op-amp powered at 40v as I was using NPN darlington(BD139+TIP35) as the output driver transistors.But To power the op-amp,a separate 40v regulator was needed.

Now i modified the design as in the schematic attached. I am using LT1014(for my previous design as it can handle as much as 44v).I have some queries listed below regarding the new output stage:-

1)The gain of Sziklai pair(for this circuit) is βQ4βQ1+ βQ4βQ2+βQ4βQ3. So depending on that,the current needed to drive the base of Q4 in order to get a max of 6A is 2.6ma as (6000/(30x25)x3).[typical gain of TIP35 is 25 and TIP42 is 30].Is my calculation right??

2)What is the purpose of R4?Is it simply to Pull up the base of Q4??How to calculate the value of it??

3)Is R7 used for current limiting?How it's value can be calculated??

I am attaching my circuit.Please feel free to suggest if anything is required to make the circuit more stable/functional.

Thanks in advance......

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Greetings -

 

Loop stability is going to be a BIG issue. Some capacitance between op-amp output and negative input may fix it. Hard to say how much is needed. Capacitance other places may also work or may also be needed in addition to the suggested one.

 

R4 is a pull-up. You cannot rely on the poorly defined input resistance of Q4 (and its leakage current).

 

Not quite sure about R7. It certainly does reduce the gain and improve stability (both AC & DC) of the Q5 stage. It may also serve a current limiting function.

 

You will have a HUGE amount of heat to deal with. 40V in, and near zero out at 6A gives you 360 Watts! The large fraction of that will be generated by the pass transistors. Better have a very good heat sink!

 

Jim

Jim Wagner Oregon Research Electronics, Consulting Div. Tangent, OR, USA http://www.orelectronics.net

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Thanks a lot for your prompt reply.I used NPN darlington(BD139+TIP35) at output stage before and using that with LT1014 was pretty stable(Not tested with more than 1A load though).But to run that setup,another 40v power supply was needed for the LT1014.

The main reason to put a PNP sziklai pair at output stage (although I could easily used TIP36,but i have several TIP35s laying around)is to eliminate the need for the separate 40v supply for the LT1014(As LM317 regulators were getting fried almost on a regular basis.No HV version found).

Anyway,thanks again for the nice explanation.will let you know what's happening......

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

By stability, I meant AC stability,j mostly. That is, too much phase shift around the loop such that it oscillates. It often happens at some medium current, where device beta is the highest. 

 

I have never heard that NPN/PNP pair referred to as sziklai pair. It has been around for ages; quite sure I saw it used at Tektronix in the 1970s, and that name was not recognized then, though Darlington certainly was. 

 

I don't understand your comments about needing another 40V for the LT1014. You will certainly need a regulator to step down from the input voltage to power that amplifier. But, it will not need to supply much current. Any regulator you use for  that ought to be rated for at least 60V.

 

Jim

 

 

Jim Wagner Oregon Research Electronics, Consulting Div. Tangent, OR, USA http://www.orelectronics.net

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Yes....40v was required to supply the Amplifier.

Although the amplifier takes only few mA,but it was hard to find a linear regulator that supports input voltage as high as 50-53v.I tried with LM317 regulator.But it was failing frequently.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

SHARANYADAS wrote:

I tried with LM317 regulator.But it was failing frequently.

 

The LM317 is cheap, so why not just use a lot of them ? :)

If you need 0-35V out, 6A, then something like 6+6 LM317 in series parallel will spread that voltage, and power (you have LOTS of power..)

If you want to go to 0v, which is certainly useful, the most common  LM317 approach is to make a small negative rail.

 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I'd suggest a switching pre-reg then follow with a linear reg.

I think the LM336 is good for 10A or you could put the pass transistors on a LM317.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

The LM317 i used  only to power up the LT1014.Not to get the full 6A output from it.My input voltage is nearly 50-53v so that's why the LM317 was failing(As the max input rated for LM317 is nearly 40v).I should use the HV version which is unavailable.
 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Kartman wrote:

I think the LM336 is good for 10A or you could put the pass transistors on a LM317.

 

The LM336 is a 2.5v shunt voltage reference...........

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I was close..... Lm338

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Kartman wrote:
I was close..... Lm338

seems the last digit ruined it all.............wink