I have been working with the LM2907 as I have been designing a few circuits around it. This has lead to me wanting to understand more of its inner workings and build up a breadboard model of the internal circuit. The reason for not just using the chip is I need to display the relative key functions.

The Ouestion

Is there possible errors in the below schematic? Specifically:

  • Charge Pump
  • The use of Q2 to charge the 22nF which is in the external area
  • My choice of C1, being electrolytic

Below is the schematic referring to the question:

LM2907 internal How I believe it should work:

The comparator (IC1A) would turn the input frequency into the correct "pulse" to the Charge Pump.

The Charge pump is based off of a single stage charge pump model. I placed Q2 so as to charge C2 as would the LM2907 do.

After this it would go to the OP-AMP, aside LM741 is used but could be any, which will control the output resistor Q1 with feedback to the Inverting input from the emitter.

Reference Schematics

These are images from the datasheet. The first is the circuit to which the above is meant to mimic in operation(less components for visibility).

enter image description here

enter image description here


1 Answer 1


No, your "charge pump" circuit doesn't make any sense and won't work.

If you look at the equivalent circuit in the datasheet (figure 51 on page 21), and then read the functional description on page 8, you'll see that it's much more complicated than that.

The charge pump monitors the output of the input comparator, and on every transition, it charges or discharges the timing capacitor C1 between 0.25×Vcc and 0.75×Vcc using a constant-current source. While C1 is charging or discharging, a fixed current flows out of pin 3 and through R1. This converts each input transition into a fixed-width output pulse. The resulting duty cycle is then averaged by C2 in order to turn it into an analog voltage, for which the output opamp provides buffering.

  • \$\begingroup\$ I guess it did look simple contrasted to the actual thing, do you think its possible to set something up with the second half of the comparator and a PNP, NPN pair and a current source, or is that still far off? \$\endgroup\$
    – RSM
    Mar 4, 2015 at 21:59
  • \$\begingroup\$ No, I don't think it's that simple, either. It take at least two comparators, some logic, and several current mirrors in addition to switching transistors and a current source. \$\endgroup\$
    – Dave Tweed
    Mar 4, 2015 at 23:48
  • \$\begingroup\$ No point in me trying to reinvent the wheel with squares at this point. My endeavors to make a model can wait at this point. +1 too. \$\endgroup\$
    – RSM
    Mar 5, 2015 at 6:07

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.