Sweep generators

These are obsolete concepts no longer needed. An op amp integrator makes a nearly ideal replacement.

All a Miller Integrator did was use the Miller Effect in a vacuum tube to approximate a constant current charging a capacitor. Another name for the Miller Sweep was the Phantastron.

Hmmm, I did a bootstrap ramp generator recently. 3 volts in 40 ns, linear to about 0.1%.

No, different circuits.

John

A linear ramp voltage is commonly created with a constant- current source feeding a capacitor.

A boostrap current source can be easily made with tubes, taking advantage of the fact that vacuum tubes in linear operation generally have negative grid voltages. So a cathode follower's cathode voltage will be positive compared to its grid, and this provides a convenient higher-voltage node that's bootstrapped to the grid voltage. One connects a resistor back to the grid, creating a constant current into the grid node.

These days one can do the same thing with a JFET or a depletion- mode MOSFET, like my favorite LND150.

Here's an example of an opamp bootstrap current source. This circuit is especially useful for making small currents, and it was the first thing I made for myself after coming to the Rowland Institute for Science about 17 years ago.

.. V+ .. | 100pA current source .. 10k etc .. | .. +------------110k --+-----, 100mV above Iout .. | | | .. LM385 ,----------, 10k Rout .. 1.2V | __ | | 1G-ohm .. | | /- |---+---' | .. '----+--< | | --> Iout .. TLC27L7 \\_+|-- 10k ------o--- 100pA current source .. MOSFET opamp

First we have an opamp connected to follow the CS output node. The opamp has a voltage reference that's bootstrapped to its input. A resistor back to the input will always have a fixed voltage across its terminals, hence it'll have a fixed current out of its terminals. This current can't go into the opamp, so it has to go into whatever is connected to the output node.

The TLC27L7 opamp is chosen for its low input current (0.6pA at room temp), low offest voltage (0.5mV max) and low cost. The NSC LMC6081A is an even better choice. However, both of these opamps only have 16V maximum supply voltage ratings (17 years ago I used an OPA128, which was rated for 36V total supply). Sometimes when using an unusual opamp with special characteristics but limited supply-voltage capability, it's necessary to bootstrap the opamp's power supply. Ahah, if we did that for this circuit, we'd have a bootstrapped bootstrap. That'd be serious bootstrapping!

This current source can be turned into a very slow ramp or sweep generator by adding a capacitor and a mosfet discharge switch.

.. 100mV above .. \\ output .. ,------- R1 110k -------+--- 10k -- V+ .. | R2 | .. +- 10k --+---------, LM385 .. | | __ | Vref = 1.2V .. 1G-ohm '-|- \\ | | .. R | | >---+----+------ sweep out .. o--- 10k --|+_/ .. | TLC27L7 dV I R2 Vref 0.1 .. | -- = - = ----------- = --- .. +--------| dt C (R1+R2) R C R C .. C _|_ ,->|| 2n7000 .. --- +--||_______ discharge _ .. | | pulse _| |_ .. gnd gnd

For example, if the 100pA current source feeds a low-leakage 1.0uF capacitor we'd get a nice slow 0.1 mV/s ramp. Note one nice feature of our bootstrap current source: a low_z buffered output voltage is included at no extra charge (pardon the pun).