EMRFD Message Archive 3716

Message Date From Subject
3716 2009-11-02 11:22:16 w1kilofoxtrot LTSpice IV and filter design for R2/T2 Pro project
This is a follow up to my problems developing a VFO for my R2/T2 Pro project from a while back. I have a j310 Colpitts VFO that has a fairly clean output (2d harmonic and higher at least -30dBc.) I was having trouble making the buffer circuit work to my satisfaction, and I thought I'd try a few simulations to guide my experiments and measurements.

Well, I think I may have found a way to improve my Colpitts design by adding a choke in the source lead and decreasing the source resistance, but when I looked at putting an LC band pass filter on the output to further reduce the higher order harmonics, I ran into a bit of a problem. I don't seem to be able to create an analysis term that correctly shows the AC response that I'd expect.

I went to the book (EMRFD) and entered the triple tuned circuit of Fig 3.21, but still no luck. I've tried a number of different terms with little effect.

I thought I might ask the group here for hints, while I start building the modified circuit to prove out what I've seen in the sims so far. Any thoughts????

73,
Steve
W1KF
3717 2009-11-02 11:52:13 Niels A. Moseley Re: LTSpice IV and filter design for R2/T2 Pro project
w1kilofoxtrot wrote:
> This is a follow up to my problems developing a VFO for my R2/T2 Pro project from a while back. I have a j310 Colpitts VFO that has a fairly clean output (2d harmonic and higher at least -30dBc.) I was having trouble making the buffer circuit work to my satisfaction, and I thought I'd try a few simulations to guide my experiments and measurements.
>
> Well, I think I may have found a way to improve my Colpitts design by adding a choke in the source lead and decreasing the source resistance, but when I looked at putting an LC band pass filter on the output to further reduce the higher order harmonics, I ran into a bit of a problem. I don't seem to be able to create an analysis term that correctly shows the AC response that I'd expect.
>
> I went to the book (EMRFD) and entered the triple tuned circuit of Fig 3.21, but still no luck. I've tried a number of different terms with little effect.
>
> I thought I might ask the group here for hints, while I start building the modified circuit to prove out what I've seen in the sims so far. Any thoughts????
>
> 73,
> Steve
> W1KF
>

Hi Steve,

What kind of a response do you expect and what is the response you see?

73,
Niels PA1DSP.
3718 2009-11-02 14:04:19 w1kilofoxtrot Re: LTSpice IV and filter design for R2/T2 Pro project
Well, doing an AC sweep; I would expect to see a classic filter, so the level on the terminated output would be much lower 'out of band' than it would be near the designed resonances 'in band'

What I usually see is a steady level, decreasing at 10X the design frequency. I have tried more points per sweep, linear, octave and decade. I have tried narrowing the range analyzed, and expanding it.

I even tried a sweep of a parallel LC combinati
3719 2009-11-02 14:42:55 w1kilofoxtrot Re: LTSpice IV and filter design for R2/T2 Pro project
Perhaps I should not follow up my own posts, but I figured I'd save you all some thinking. Turns out that I cannot abbreviate 10 000 000 Hz as 10M. Once I spelled out the frequencies of interest, everything is working as I expect...

73,
Steve
W1KF

3721 2009-11-02 14:46:18 Niels A. Moseley Re: LTSpice IV and filter design for R2/T2 Pro project
Hi Steve,

Yes, I consider that a flaw in LTSpice; 'M' is interpreted as milli, not
as mega. For mega you should use 'Meg'.

73,
Niels.

> Perhaps I should not follow up my own posts, but I figured I'd save you all some thinking. Turns out that I cannot abbreviate 10 000 000 Hz as 10M. Once I spelled out the frequencies of interest, everything is working as I expect...
>
> 73,
> Steve
> W1KF
>
>
3722 2009-11-02 17:25:41 chuck adams Re: LTSpice IV and filter design for R2/T2 Pro project
On Mon, 2009-11-02 at 23:45 +0100, Niels A. Moseley wrote:
>
> Hi Steve,
>
> Yes, I consider that a flaw in LTSpice; 'M' is interpreted as milli,
> not
> as mega. For mega you should use 'Meg'.
>
> 73,
> Niels.
>

I wouldn't think of it as a flaw. It was in the original
SPICE and should be planted in the back of the mind as
something to watch for.

Wait until some one uses F for Farad in a cap. 0.000001F
is not what you think it is. ;-)

chuck, k7qo
3723 2009-11-02 17:55:29 w1kilofoxtrot Re: LTSpice IV and filter design for R2/T2 Pro project
0.00001 femtofarad is a pretty small cap...

I didn't recall the 'm' limitation. It has been a long time since I used Spice. I was well aware not to trust the result, and that I had done something outside the envelope, but... now I know how to get a decent result from my VFO. I now expect the filter will get me harmonics better than -60 dBc

Thanks!

73,
Steve
W1KF