Best way to remove DC offset ?

Sep 04, 2017 9 Replies

Could some electronics guru please help ? The SPICE simulation for a 50 MHz Colpitts oscillator has a 1.5 V DC offset. What would be the best way to filter it out ? AC coupling capacitors are not useful because of capacitive reactance. Any hints amd/or suggestions would be helpful. Thanks in advance.


Why is the reactance an issue?

Sylvia.

At 50MHz, the capacitor is the answer. The reactance of a 3.3nF capacitor is only 1 ohm, and it that's a problem (then what the hell are you driving?) you can easily go bigger.

Well, a transformer is accurate. Really, though, it depends on what 'best' means.

I am fully aware of AC coupling and how it works. Adding a AC coupling capa citor at the oscillator output(right before a 50 Ohm load -- thereby creati ng a RC high pass filter) of a value close to that suggested by you, gives strange spikes whose value is in the range of 10^-7 Volts -- useless for pr actical purposes. I would be happy to share the SPICE netlist.

Please do. If you don't know where 100nV spikes are coming from, changing the output coupling will not help you. Because it's not the capacitor.

Also, you almost certainly have left out parasitics in your sim that will have much bigger effects.

Clifford Heath.

Probably the 50r load is killing the oscillator. Post your sim and we'll see.

John Larkin Highland Technology, Inc lunatic fringe electronics

Ya how is anyone supposed to know anything without a schematic? Trying to drive 50 ohms right out of a 50 MHz Colpitts is NOT-AH SO GOOD

ote:

capacitor at the oscillator output(right before a 50 Ohm load -- thereby cr eating a RC high pass filter) of a value close to that suggested by you, gi ves strange spikes whose value is in the range of 10^-7 Volts -- useless fo r practical purposes. I would be happy to share the SPICE netlist.

I bave changed the LC tank completely, from a simple parallel LC to a PI with a single inductor and 2 capacitors as the legs. The output is AC coupled through a 0,1 uF capacitor and the load is a 5K resistor. The DC offset is gome, the oscillation frequency is now

1 MHz, which I will boost up, by tweaking the LC tank. Thanks for all your helpful hints.

e:

g capacitor at the oscillator output(right before a 50 Ohm load -- thereby creating a RC high pass filter) of a value close to that suggested by you, gives strange spikes whose value is in the range of 10^-7 Volts -- useless for practical purposes. I would be happy to share the SPICE netlist.

Increasing the frequency by a factor of x100 is gonna be more than a "tweak ."

Join the Discussion

Have something to add? Share your thoughts — no account required.

Didn't find your answer?

Ask the community — no account required