Categories: Electronics

20MHz VFC with take-back-half cost pump

Manner again in 1986, famed analog innovator Jim Williams, in “Designs for Excessive Efficiency Voltage-to-Frequency Converters” printed his “King Kong” 100 MHz VFC. I’ve by no means seen its equal. Definitely Determine 1’s little circuit, topping out round 20 MHz, is nowhere shut.

Determine 1 Take-back half (TBH) cost pump provides easy VFC affordable efficiency at 20 MHz.

Wow the engineering world along with your distinctive design: Design Concepts Submission Information

Nonetheless, though left in Kong’s mud with its doorways blown off, Determine 1’s VFC is nonetheless a number of occasions sooner than commercially out there VFCs (e.g., the 4-MHz VFC110) whereas conveniently operating on lower than 10 mA from a single +5-V provide/reference.

What makes it work at such a excessive output frequency (with out Ok. Kong’s complexity) is (primarily) the self-compensating TBH diode cost pump described in an earlier design thought: “Take-back-half precision diode cost pump”. We’ll get to that shortly.

In the meantime, right here’s an outline.

A 0-to-1 mA full-scale enter metered by R1 is built-in on C1, inflicting the enter amplifier’s output to ramp up, turning on present sink Q1. The sink present ramps down the voltage at Schmidt-trigger U1 pin 1 till its destructive set off degree (~1.5 V) is crossed. This begins a cascade of transitions by means of the three-inverter daisy chain delay line. Pin 2 snaps excessive, making pin 4 go low, flipping pin 6 excessive. Propagation by means of the chain takes about 20 ns. Arrival of the ramp-reset pulse at pin 6 is fed again by means of D5 to pin 1, pushing it by means of U1’s optimistic set off degree. This initiates a complementary wave by means of the daisies, ultimately finishing the cycle in ~40 ns.

Oscillator frequency is thus (very roughly) proportional to R1 enter present. It’s the job of the pump and op-amp to make it precisely so. The trick for doing this depends on the TBH pump with its two humorous trying anti-parallel diode pairs: D1 D2 and D3 D4.

D3 and D4 couple input-balancing destructive suggestions present to C1 that’s theoretically equal to -100 µA/MHz however in apply is diminished by sundry error phrases attributable to varied diode non-idealities. These embrace ahead voltage drop, reverse restoration time, stray and shunt capacitances, and so on.

In the meantime opposite-polarity D1 and D2 couple optimistic suggestions present to C1 that’s (once more theoretically) equal to +50 µA/MHz however is virtually diminished by precisely the identical troublesome checklist of nonidealities listed for D3 and D4.

Consequently, when the 2 opposing currents are summed on C1, the errors phrases neatly cancel, leaving solely the specified -(100 – error) + (50 – error) = -50 µA/MHz of correct destructive suggestions, making:

Fout = 20MHz Vin (1000 / R1)

Please see “Take-back-half precision diode cost pump” for a considerably much less abbreviated derivation.

Just a few choosy design particulars embrace this stuff.

Q1’s base drive resistor was chosen in accordance with the 2N3904 datasheet min/max beta vary to be low sufficient to permit adequate collector present for a full 20 MHz, however excessive sufficient to forestall dragging down D5 and U1 pin 6 excessively and killing oscillation as a result of pin 1’s optimistic set off degree can’t be reached. This latter situation would doubtlessly trigger the converter to latch up

Leakage-killer R4 prevents U1, D5, and Q1 summed leakage currents from producing zero offset oscillation even when the op-amp has turned Q1 off.

In case you can’t discover a use for the remaining components of U1 which can be unused, be sure you floor their floating inputs or tie them to +5.

Banana, anybody?

Stephen Woodward’s relationship with EDN’s DI column goes again fairly a great distance. Over 100 submissions have been accepted since his first contribution again in 1974.

Associated Content material