0
\$\begingroup\$

I'm making a 0-±12V 20mA supply for breadboard. Circuit simulates fine with ltspice. How to implement tracking (or tracking ratio) WITHOUT opamp?

enter image description here

I found a schematic here https://groenholdt.net/projects/LAB-PSU/schematic-and-pcb.html but it is a bit difficult for me to understand and extract the relevant part.

I'm thinking of using a ganged dual potentiometer for the voltage control, and another potentiometer to adjust the tracking ratio... saves space.

Any help will be appreciated. (I know how to do it with opamp, just trying to see if it can be done with just one or two transistors... maybe smaller pcb without saving any money, since lm358 are already so cheap.)

PS: At 12V 20mA, the MMBT3904 and MMBT3906 seems to dissipate about 220mW. Will this be a problem? Any tips on if PCB layout (large collector pad) will help? Dissipation will likely be <150mW during normal use. Otherwise will have to put in BD139 / BD140 in darlington configuration (will also increase voltage drop).

\$\endgroup\$
1
\$\begingroup\$

schematic

simulate this circuit – Schematic created using CircuitLab

For a tracking circuit without opamps, try something like this. The change is to make Q4 compare the midpoint of the two outputs with ground, so its emitter needs to be biassed one diode drop ABOVE ground. (Eliminate R9 and U1)

I also recommend replacing the pass transistors with something slightly bigger - BD139/140 are suitable.

\$\endgroup\$
  • \$\begingroup\$ Thanks! I simulated it and it works nice above 6V. Too much difference at lower voltages. With two diodes, it is very good at low voltages but too much difference above 5V. I guess it's good enough for no extra transistor tracking! \$\endgroup\$ – Indraneel Nov 30 '18 at 13:51
  • \$\begingroup\$ there's probably some way to use a differential pair here. |I was lazy and used an op-amp \$\endgroup\$ – Jasen Nov 30 '18 at 18:55
  • \$\begingroup\$ @Jasen I don't blame you, an op amp costs less than 3x MMBT3904 (not counting the resistors), and less circuit complexity, better accuracy and probably less board space. Also, probably slightly better heat dissipation (particularly for slightly more expensive PDIP). \$\endgroup\$ – Indraneel Dec 1 '18 at 12:09
  • \$\begingroup\$ The lm358 schematic does show differential pairs in it, as do schematics of all other opamps. But those usually take up 4 transistors. \$\endgroup\$ – Indraneel Dec 1 '18 at 12:16
0
\$\begingroup\$

check the data sheet, 220 mW seems like a lot for a small transistor

for tracking

schematic

simulate this circuit – Schematic created using CircuitLab

\$\endgroup\$
  • \$\begingroup\$ Thanks. Forgot to add NO opamps! (hence the confusion). Sorry about that. Yes 220mW is at the edge. Trying to get away with it if possible. But will be ablut <150mW in normal use. \$\endgroup\$ – Indraneel Nov 30 '18 at 10:29

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.