Uaf42

Have anyone used UAF42 to design filters? I'm trying to develop a 300Hz LPF using that. But simulation fails. I dont no the reason.

"TexasInstument"'s Application is confusing. Any way I've attached that one two.

I have attached how my capture schematic looks like. I applied variable frequencies but it dint work. The simulation I have attached is values calculated for 2kHz LPF. Any major error as you can see? Help pls

R3 should be between 7 and 14, not 7 and 4.

Thank you so much!

????? I wanted a low pass filter but this thing shows me a totally different results!!!!!!!!!! ?????? Any mistakes that u can see ??

The only UAF42 I've heard of is an old valve, a single diode and pentode, with 100mA series heater, and a glass base who's name I've long forgotten?.

Bavananth said:

????? I wanted a low pass filter but this thing shows me a totally different results!!!!!!!!!! ?????? Any mistakes that u can see ??

Click to expand...

I simulated your circuit and got a very high peak in the response (the actual circuit would probably oscillate). That may be what you're seeing in your simulation. I removed the input resistors at pin 3 and applied the generator input at pin 2. This geatly reduced the peaking and gave me the response shown in the attachment.

The rolloff is at about 30Hz with a slight peak.. (Note the Log-Log display to give a better view of the results.)

You used 4.4 megohm resistors for R2 and R3 in your circuit. Is that the value you intended?

If you need to use the pin 3 input, the value of the input resistors need to be significantly increased to minimize the peaking.

Nigel Goodwin said:

The only UAF42 I've heard of is an old valve, a single diode and pentode, with 100mA series heater, and a glass base who's name I've long forgotten?.

Click to expand...

A picture of one at http://members.shaw.ca/pacifictv/uaf422.jpg

crutschow said:

A picture of one at http://members.shaw.ca/pacifictv/uaf422.jpg

Click to expand...

Still doesn't tell me what base it is, it's not B9A, because that's 8 not 4, 3 is octal - it's been so many years (and 4 series were fairly uncommon) that I can't remember?.

Couldn't take it any more! - been and looked it up, 4 is B8A.

Thanks crutschow! anyway I wanted a 300Hz LPF filter. According to the equations in the data sheet given by TI , I calculated those values. But It doesn't even look like a filter. They provide a software for design calculation.
**broken link removed**
But that is really difficult and contradicting with the data sheet.They have mentioned some sub cct in application bulletin which is not mentioned in the data sheet. So confusing.

I have to modify Crutschow's design suiting to my requirements. Hmmm...
Thank you for your help.

Bavananth said:

anyway I wanted a 300Hz LPF filter.

Click to expand...

I simulated the circuit with some various values and came up with a -3dB rolloff at 300Hz with the values shown in the attachment. The circuit is similar to Figure 2 in the LF42 data sheet. The value of the resistor from Pin 3 (RQ in Figure 2) to ground affects the peaking of the circuit at rolloff. 40K ohms gives a very small peak. A larger resistor reduces the peak and a smaller resistor increases the peak. Try simulating different values for RQ to see the peaking change. That's one thing great about simulations, you can easily change component values and see the results immediately.

I haven't used the software so can't help you there. It looks like it calculates values to allow cascading of several of the filter circuits to build various types and higher order filters (up to 10th order) for faster rolloff. (See Pg. 3 of the App Bulletin for a summary of the various filter types). To do this requires a good understand of the different types and order of filters you can build, and what your filter requirements are. I believe with one chip you can build up to a 3rd order filter (see Figure 16B in the App Bul).

Thanks a lot crutschow. You are really unbeleivable. Thanks a lot again. I'll modify it and check. I really appreciate your help. Have a nice time.