Muff's Modules & More

[SOFTWIRE]

So i have a 0-5v clock signal that is variable from 1khz to 2.5mhz. On breadboard i am gating the signal with a 4066 switch. When the switch is off there is a slight offset around 20khz on the output of the switch. and at 800khz the output of the switch has an offset of 5v. The chip is being powered from 12v. on the TI datasheet it says" Frequency Response, Switch On = 40 MHz Typical" Does this mean the switch can toggle at a rate of 40mhz or it should be able to pass signals up to 40mhz. Should this work or should i be looking for a different chip/transistor. Any suggestions is appreciated. Im sure you video synth guys would know of a better component. Thanks

[daverj]

That's talking about the signal, not the control, under very specific conditions.

If you scroll down a bit more you'll see the maximum control repetition rate is between 8 and 9.5Mhz (under other specific conditions)

Those specs only work if you meet the specific conditions. Otherwise both frequencies are much lower.

[SOFTWIRE]

Any idea why im getting a frequency dependent DC voltage bleedthrough?

[daverj]

What kind of circuitry is driving the input, and receiving the output? Remember that these are unbuffered devices. Basically they act like a resistor that changes from a few hundred ohms to a few megohms as the control toggles. The output can't just float. It needs a load. It works best if both the input is very low impedance and the output is between about 1k and 10k load.

Is the control signal solidly and quickly going between the two voltage rails?

Have you checked that the input doesn't have the same offset that you're seeing at the output?

[SOFTWIRE]

Im using an LTC1799 as a vco which is oscillating at 0-5v amplitude. The direct output of the 1799 goes to the input of the cd4066 which is powered from ground and 12v. on the output of the 4066 it goes through a 10k resistor to the non-inverting input of a lm319 high speed comparator. which is level shifting the clock to 0-12v(via-pullup 500r). So if the 4066 is off(12v control signal from a cd4070-not a high freq signal) the output of the 4066 varying a dc output of 0-5v depending on the oscillating signal present on its input. The output of the 1799 doesnt have any offsets from varying the freq

[daverj]

The output of the 4066 is going through a 10k series resistor to the input of a 319? nothing else on the output of the 4066 or input of the 319?

That's probably the problem.

Try putting a 1k or 2k to ground on the output of the 4066 and change the series resistor to 1k.

I'm assuming the negative input of the 319 is at a fixed voltage around 2.5v?

Instead of using a 4066, have you considered just using a logic gate? Like a 74HCxx AND or OR (depending on what you want the output to do when the control goes off). You could also just use a 7406 for level shifting after the logic gate. (the 4070 would have to be attenuated to 0-5v to use logic gates)

Or am I misunderstanding what you expect to happen when the 4066 is in the off state?

[SOFTWIRE]

OK so the resistor values you suggest worked and it got rid of the dc on the 4066 output. But i ended up switching the circuit around. I took the output of the 1799 through a 1k resistor with a 10k to ground( since 2k was bringing the clock closer to 3v peak) directly to the lm319 comparator. Then took its output and ran it through an cd4081 AND gate which works perfect for gating the clock. Didnt know about the 7406 but i think id have to have a 2nd inverter stage to follow the logic. also i can use both comps in the lm319 but i probably wouldnt use all 6 inverters in the 7606. I initially tried the CD4504 level shifter where i use the lm319 now but it didnt seem like it had the bandwidth i was after. Although could have been my error somewhere.

And as always you are a great help Dave. Thanks a million

[daverj]

This sounds like a better approach than using the 4066. Good job.