Group Actions
Engagement
  • Author Author: Robert Peter Oakes
  • Date Created: Date Created
  • Views 3284 views
  • Likes 3 likes
  • Comments 20 comments
Related
Recommended

The Modular Bench Power Supply ++ - Power Output Stages

Robert Peter Oakes

In previous videos we have looked at the different modules that make up a typical bench power supply, in this post we will continue that trend and look at the Power Stage of typical power supply, evaluating different components (NPN, PNP, N Chan and P Chan MOSFETs), how there controlled and a bit of practical how-to

 

This is a link back to the main PSU blog : http://www.element14.com/community/groups/test-and-measurement/blog/2014/09/15/the-modular-bench-power-system-the-essential-diy-build-for-every-ee-student-and-old-timer-alike

 

Sorry its long (I did shorten it... honest image ) and is the first of two, this one covers a basic intro to the devices and detail of the NPN and PNP BJTs

 

and here is the next video covering the MOSFETS

  • in reply to jw0752

    I was experimenting with various OP amps and I had what seems to be a similar experience to your situation

     

    using a TL082, when the voltage into the op amp went below 0.7V the output shot back up, anything above the 0.7 worked just fine, the datasheet seems to indicate the inputs not going closer than 700mv of - supply rail without risking phase reversal

     

    I replaced the op amp with a TL072 and it work perfectly all the way down to 0v but at the cost of having to restrict the supply to 15V for the op amp so it will not be able to directly drive the FET or BJT for 30V supplies

     

    I have just ordered some TLE2142CP (DUAL) http://www.ti.com/lit/ds/symlink/tle2142.pdf    and TLE2144CP (QUAD)  http://www.ti.com/lit/ds/symlink/tle2142.pdf

    these have an output swing to 100mV of - rail and 1V of + rail and  Inputs can operate between VCC− − 0.3 to VCC+ − 1.8 V without inducing phase reversal, and I think it may have been phase reversal we were seeing as for me the output started to rise again as the input went lower, this device seems to be able to handle past 0v without issues

    a bit of experimenting will be in order once they arrive

     

    worst case the use of a slight negative rail will fix the problem and also allow true 0V output rather than + a few mV, i can get the MOSFET to within 10mV of zero but not better without more circuitry.

     

    The better solution is not to go for 40+ volt op amps and run the control system from perhaps 12 to 15V, this opens up a whole pile of alternate op amps and only will require the addition of a low cost transistor and a couple of resistors to provide the high side control

     

    Have you had ant more testing performed in your lab for this ?

     

    Peter


  • in reply to jw0752

    Please do, happy to help where I can

     

    also, I better get my next video done, if I'm note careful you will get ahead of me image

  • I was using a linear 270 degree 10 K  pot, so you are correct that my resolution at the ends were not good. My voltage readings were being supplied by the oscilloscope so here again accuracy was not being taken very seriously. My goal was to show proof of concept and while I logged data including multiple load settings for several data points I was less than consiencious over all.  Tonight I am going to make the slight redesign of the circuit to accomodate a MOSFET and do a similiar experiment. I also want to see how it works in a constant current mode. I will let you know if I run into any anomolies.

    John

  • in reply to jw0752

    That's interesting, I will have to get myself some of those to experiment with, in the mean time I will build up a similar circuit over the weekend to compare notes, I need to do the next video anyway and this will help cover off two things in one

     

    This should not go none linear unless the inputs are not rail to rail even though the outputs is, I will have to re-consult the data sheets to see what is shows

     

    Are you using a ten turn pot or a single turn, also is it Log or Linear scale, a single turn would loose the resolution (Control) at the low end due to massive changes in the ratio from high side to low side in the last few degrees of turn

     

    try a coarse and a fine pot to see if this helps. Also if you plot the + input volts and the emitter volts across the range you should find the gain remaining constant even though the lower setting are hard to get precise, if not them something weird is happening and we will have to investigate

     

    try increasing the load, or keep it constant at say 300mA and record the result across the range ?

  • Hi Peter, Yes I was able to bring it below 4.6 volts output but my gain became non linear below the 0.7 volt reference voltage input to the
    ISL 28218. It would be easy to set up the voltage divider off the emitter so that there was a different ratio and outputs could be brought to less than a volt linearly. The control voltage was being measured at the non inverting input of the ISL 28218.

    Thanks John