Linear Technology's FilterCAD

Works fine for me. Are you giving it parameters that lead to an order greater than 16? Look at the "order" box, assuming you're using the "enhanced design". What _exactly_ is the complaint it registers? Perhaps it's telling you that you are asking to implement an impractical design, and it refuses.

It's Burridge, what do you expect ?:-)

...Jim Thompson

-- | James E.Thompson, P.E. | mens | | Analog Innovations, Inc. | et | | Analog/Mixed-Signal ASIC's and Discrete Systems | manus | | Phoenix, Arizona Voice:(480)460-2350 | | | E-mail Address at Website Fax:(480)460-2142 | Brass Rat | |

| 1962 | I love to cook with wine. Sometimes I even put it in the food.

Thanks, Tom. I was simply (initially) trying to get it to come up with design values for a L/C notch filter centered on 153Mhz with -30dB attenation at the centre and *passband* 3dB corners at 147Mhz and 159Mhz at < 4dB overall insertion loss. Is that really so difficult???

I had it choke on 'quick", but then when I went back to "enhanced" it even carried over the parameters from quick and produced results.

--
Ben Jackson

http://www.ben.com/

Well, yes, Paul, it is very difficult for an active filter circuit. Remember, the Linear Technology software is for designing filters with their active filter parts.

You should go to some LC filter design software such as Elsie, the filter software from AADE, or the filter design stuff in RFSim99. Be sure to include inductor Q in your design. You have to do that "by hand" by adding resistance to the inductances in RFSim99. The AADE software lets you enter coil Q, and I think Elsie does too. And depending on what you REALLY want to do, it may be quite inappropriate to ask for a symmetrical notch and passband flat from DC to close to the notch.

I have found that RFSIMM99 seems to incorporate Q when I click on USE PHYSICAL MODEL.

This works for capacitors and inductors. For resistors, I have to add the parasitic capacitance and inductance as separate components, but that may, or may not, be required depending mainly on frequency and what sort of physical component will be used.

Now, in the circuit Paul wants, I'd use the Monte Carlo tolerance sweep to see if the circuit is too critical to be reasonable to build. To get that notch, and narrow response, is probably not going to be easy.

Jim N6BIU

Yes, you're right that it's more automatic than I suggested, though AFAIK you have to set the Q and SRF parameters for each component individually after doing the design. Plus of course RFSim99 won't do notch filter designs as Paul is looking for.

I suppose Paul is wanting to make a one-off filter, so using tuning caps should let him make the notch be where he wants.

Thanks for pointing out the "use physical model" thing.

Cheers, Tom

Oh! Thanks for that info, Tom. Yeah, I'm only interested in L/C passives and had no idea about that LTFC was built around their ICs. Thanks for the heads up! Thanks to everyone else who chipped in their

2p worth, btw.