Spice Oddity

Jan 19, 2016 14 Replies

Hi



I have a line driver circuit in LTspice simulation. It's a bridge circuit with op-amps and bipolars with a neat well-defined quiescent current. Sadly, I can't show the design, but I'm sure it's not original. The frequency range is 10kHz to 80kHz.



So, I do a transient sim with a sinusoidal source stepping from 10kHz to



80kHz and all is good. Not flat, but within a dB. Then I do an AC analysis and find that 80kHz is apparently >16db down on 10kHz, which is not good. I check for clipping etc on the transient sim, but none found and the output waveforms look nice and clean.

I'll build this soon, and I'm inclined to believe the transient sim, (probably because I want to) but any insight as to why there should be such a disparity? is there a well-known gotcha?



As the man with the steering wheel on his penis said, it's driving me nuts.



Cheers


Syd

What do you see when the transient analysis is run with really small signal levels?

The AC analysis assumes the bipolars are biased at their quiescent current, which (?) might be nearly off.

John Larkin Highland Technology, Inc picosecond timing precision measurement jlarkin att highlandtechnology dott com http://www.highlandtechnology.com

Simulators often fail on .AC analyses when part or all of the signal path had non-linearities.

LTspice particularly fails with some of my TANH stuff, linearizing absurdly.

Choosing "Alternate Solver" fixes some of it.... if you're lucky.

In general .TRAN trumps .AC when there's a non-linear signal path.

"Why should we hang, Dick did the shooting" >:-} ...Jim Thompson

| James E.Thompson | mens | | Analog Innovations | et | | Analog/Mixed-Signal ASIC's and Discrete Systems | manus | | San Tan Valley, AZ 85142 Skype: skypeanalog | | | Voice:(480)460-2350 Fax: Available upon request | Brass Rat | | E-mail Icon at http://www.analog-innovations.com | 1962 | I love to cook with wine. Sometimes I even put it in the food.

Using 14 times smaller input signal, I see the same frequency response, and the transient is 14 times smaller.

Cheers

Syd

Alternate Solver don't make no nevermind.

That's encouraging, thanks.

Cheers

Syd

Strine?

Spice Ordditaey!

(b'thlight Davabaowea)

Graoncon trolla Myjoh Tome...

On Tue, 19 Jan 2016 19:24:04 -0800 (PST), Bill Beaty Gave us:

Yeah it's a bunch of monkeys (Monkees) standing around a big bright white monolith dancing to Saturday Night Fever in their birthday suits.

Or is that a stage act for a bunch of male strippers?

So this isn't an Antipodean tribute to the late David Bowie?

--sp

-- Best regards, Spehro Pefhany Amazon link for AoE 3rd Edition:

formatting link
Microchip link for 2015 Masters in Phoenix:
formatting link

Not really, though LTspice does have a ground control on the toolbar, and it's often the case that my circuit's dead, there's something wrong.

Cheers

Syd

Consider transfer functions:

____ ____/

and / / __/ / / /

...Jim Thompson

| James E.Thompson | mens | | Analog Innovations | et | | Analog/Mixed-Signal ASIC's and Discrete Systems | manus | | San Tan Valley, AZ 85142 Skype: skypeanalog | | | Voice:(480)460-2350 Fax: Available upon request | Brass Rat | | E-mail Icon at http://www.analog-innovations.com | 1962 | I love to cook with wine. Sometimes I even put it in the food.

On Wed, 20 Jan 2016 15:21:35 +0000, Syd Rumpo Gave us:

And of course everything is made from Ziggy's stardust.

On Wed, 20 Jan 2016 08:25:58 -0700, Jim Thompson Gave us:

Hit the hole, not the bowl.

It would be interesting to poke around various nodes and see where the difference originates.

John Larkin Highland Technology, Inc picosecond timing precision measurement jlarkin att highlandtechnology dott com http://www.highlandtechnology.com

On Wed, 20 Jan 2016 10:43:19 -0800, John Larkin Gave us:

The difference is between Bowie and Kubrik.

His circuit's dead! There's something wrong!

Join the Discussion

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

Didn't find your answer?

Ask the community — no account required