Olin, I thank you for taking the time to critique my circuit design. Your points are all well taken, and I appreciate your comments. The design was breadboarded and tested not in my lab where I have access to quite a bit of equipment, but in a small room near my bedroom where I have set up a very small lab where I can experiment late at night. The only zener that I could find in my junkbox when I was playing around with the design was a 5.1 volt unit, so that is what I used. School is out this week for the Easter vacation, and instead of waiting until I could get back to work in the school lab, I simply made do with what parts and test equipment I had at hand. When the same circuit with different resistor values and a 5.1 volt zener is used to create a regulated 9 volt circuit, the specs improve, because then the zener has much more current flow through it. Of course the current drawn by the circuit also increases. As to the .1 ufd capacitor, I agree with you. I used it to suppress a tiny high frequency component that was generated by the driving supply when the current approached 500 ma (which was the practical limit of the driving supply). Because of the high frequency it turned out that it worked best across the 4.3k resistor. In my first rough drawing of the circuit I had specified a .01 ufd where you recommended in your post, but in practice this did not suppress the noise from the driving supply nearly as well as the .1 ufd across the 4.3k. My guess is that because of the high frequency of the noise there was more phase shifting involved when it had to compensate through all three transistors. I will do some more experimenting with the circuit once school starts up on Monday and I have access to some better test equipment, and a wider selection of parts. ----- Original Message ----- From: "Olin Lathrop" To: Sent: Saturday, April 06, 2002 11:01 AM Subject: Re: [EE]: Improved LDO regulator circuit > > 19k GIF attached. > > I certainly believe you that this circuit works, but I have a few comments: > > 1 - It would be helpful if you numbered the components. That would make > it easier to talk about. I'm calling the power transistor Q1, the left NPN > Q2, and the right NPN Q3. > > 2 - As you pointed out, the zener is being run at a lower voltage than its > rating. Assuming the Q3 base voltage is from 600 to 700mV, then the zener > current is from 320uA to 190uA plus the Q3 base current. Assuming a gain of > 100 for Q3, it can require up to 45uA base current at max input voltage. > This makes the max zener current about 370uA. The zener voltage is going to > be rather unpredictable at this current for an unspecified "5.1V zener". > Some of these zeners might might have too high a voltage accross them at > that current for this circuit to work. In any case there will be large part > to part variations and a strong temperature sensitivity. > > This can be easily fixed by using a lower voltage zener and running more > current thru it. Lower voltage zeners have softer knees and more > temperature dependency, but certainly not as bad as running a 5.1V zener at > 4V. > > 3 - The BE drop of Q3 is part of the reference voltage. This is will > increase the temperature dependency of the output voltage, but may not be > much of an issue especially for a hobby regulator. > > 4 - The 100nF cap between the input and the base of Q2 doesn't belong > there. It actually has the opposite effect of what you intended. It will > tend to feed high frequencies in the input voltage to the output voltage. > Think of it this way: A positive spike occurs on the input voltage. That is > fed directly to the base of Q2 via the capacitor, causing Q2 and therefore > Q1 to conduct more during the spike - exactly opposite of what you want. > > 5 - I would put a small capacitor between the output voltage and the base > of Q3. This will feed back output ripples to the base of Q3 without them > being attenuated by the zener and resistors like the DC is. You have to be > careful, though. With too much of a capacitive load on the output, you > could end up with an oscillator. The best way to determine the capacitance, > especially for a hobby supply, is just to experiment. > > > ***************************************************************** > Embed Inc, embedded system specialists in Littleton Massachusetts > (978) 742-9014, http://www.embedinc.com > > -- > http://www.piclist.com hint: To leave the PICList > mailto:piclist-unsubscribe-request@mitvma.mit.edu > > -- http://www.piclist.com hint: To leave the PICList mailto:piclist-unsubscribe-request@mitvma.mit.edu