RC filter component selection

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
ok so.. yes.. im a noob.

i need to create a pair of simple lowpass RC filters for line-level signals. lowpass around 2-3khz.

there are many calculators that let you choose values for R, C and frequency obviously, no problems there.

however, it seems to me you would want the lowest value for R as possible, so you dont attenuate the signal more than necessary.

could somebody explain why im not using some 1 ohm resistors?
 
You can, the formula is F-3db = f=1/2piRC, you can shuffle things around to make the equation more suitable for your calculation, the resistor doesn't make the signal smaller in anyway, it's the result of R and capacitor Reactance Xc interaction.

with R 1k ohm F-3db 2khz, C will be 80nF.
with R 1 ohm F-3db 2khz, C will be 80uF.

1. Capacitor will be different types, with smaller C you can choose high quality Film caps with low leakage, better linearity, 80uF, wish you good luck Sir.
2. with 1ohm, the cable you use and pcb board isn't IDEAL conductor, it has resistance as well, also capacitor is far from IDEAL, it has ESR ESL will play bigger role in this situation
3. with small resistor, it need more current to charge capacitor. so you have to worry about your source drive ability

That's just on top of my head, not sure if that's right. Once thing you have to pay attention is, the F-3db is not the starting point of attenuation. Starting point will be for low pass filter, one decade(right word for it?) before F-3db point, so 200hz the signal will start drooping. Also keep in mind, it's easier to find suitable value resistor than caps, and cheaper.
 
EDIT: further reply has helped a lot.! arrived while i was typing. so i should not consider the resistor in isolation..

----------------------------------------------------------------------------------------------------

thanks for the reply.

would you mind educating me a touch further?


the amp is the tpa3255EVM here:

TPA3255EVM TPA3255 315W Stereo / 600W Mono Ultra-HD Analog-In Audio Class-D Module | TI.com

but i cannot find the input impedance listed anywhere, and methods to measure it seem complex.

then, assuming i have this information, what to do with it? how does it affect the resistor value i choose.

i cant help thinking a 4.7k resistor will cut down my signal rather a lot? this means one amp channel will see a noticeably quieter signal overall?
 
Last edited:
i cant help thinking a 4.7k resistor will cut down my signal rather a lot? this means one amp channel will see a noticeably quieter signal overall?

The amount of loss will depend on the input impedance of the amp. The 4.7K resistor and the amplifier input will form a voltage divider. If your amp had a 4.7K input imp. you'd end up with half the voltage that you'd have had without the resistor (below cutoff frequency, that is).

The formula for a divider is Vout = Vin * R2/(R1+R2), so Vin * 4.7K/(4.7K+4.7K) factors down to Vin * 0.5, you lose 50%.

With a higher input imp. you'd lose less voltage, for instance with 100K you'd get Vin * 100K/(4.7K+100K) which gives you Vin * 0.955, so you've only lost 5%.

I'm not sure why you say one channel would be quieter, wouldn't you put an identical filter in each channel?
 
Moderator
Joined 2011
so i should not consider the resistor in isolation..

It's all relative. The series resistor should be much larger than the source impedance,
but also much smaller than the load impedance. In practice, often you can use
Rfilter = sqrt(Rsource x Rload) as the best compromise.

The input impedance appears to be 10k. If your source is 1k,
then use Rseries = sqrt (10k x 1k) = 3.16k.
Then choose the capacitor that sets your desired frequency corner.

There's a more precise way to calculated C if you know both Rsource and R load.
 
Last edited:

PRR

Member
Joined 2003
Paid Member
....i cannot find the input impedance listed anywhere

And the schematic is a mess.

It is actually VERY well defined, once you sort it out. There's two inputs, for plain unbalanced or for balanced, both on RCA jacks. Each one is 9.09K.

Yes, 4.7K series resistor will be a 3.6dB loss, very significant.

The other end is: what can your preamp/DAC drive? Some strain on anything under a few K, others will drive 600 Ohms or even 100 Ohms. If you can drive say 470 Ohms, then loss is a tolerable 1dB.
 

Attachments

  • tpa3255EVM.gif
    tpa3255EVM.gif
    19.3 KB · Views: 132
ok so im fighting fires on multiple fronts here!

i started a project which is balooning uncontrollably beyond my skillset, which extends to reading schematics, soldering and general design/construction.

i had assumed that buying pre-made amp boards and all other major components, it would be fairly plug-and-play, but im rapidly being proved wrong.

ive purchased a vast quantity of hardware now, so there is no backing out!


to extend the remit of this thread, seek help, and stop hijacking the 3255 evm discout thread, i shall explain the whole issue here and hope somebody can help overall!


so, source is a minidsp 2x4 HD, which apparently outputs 0.9v rms at clipping.

amp is the 3255 evm, which apparently has a 9.09k input impedance, and needs a 2v p2p input.

i have two of them.

the amps are set up in 2.1 mode, with (on each side) one channel of the minidsp going to the sub, and one being split in two, with one half having a lowpass on. these two are sent to seperate channels on amp to drive 1.5 way fullrange speakers. lowpass is to ensure treble is from one driver only.

so four channels from minidisp, and six channels of amplification in total.

clear as mud?


so.. multiple issues:

i need to adjust the gain on the amp so it will run well with the lower voltage provided by the minidsp. my bugging on the evm discount thread has narrowed it down to adjusting the four 10k resistors (Rg) on the board. these are microscopic surface mount jobbies.. yay!

however it has been suggested this will have knock-on effects i dont really understand. (rfi rectification)

now the lowpass filters will attenuate the signal further, meaning gain needs increasing even further. trouble is, increasing gain changes input impedance of amp, meaning lowpass filter needs tweaking ( if i understand correctly)



so im basically a bit lost. !
 
Moderator
Joined 2011
source is a minidsp 2x4 HD, which apparently outputs 0.9v rms at clipping.
amp is the 3255 evm, which apparently has a 9.09k input impedance, and needs a 2v rms input.

To prevent gain loss, you can use an op amp buffer after the RC filter,
or else an active low pass filter. Both can have the input impedance of your choice.
http://www.electronics-tutorials.ws/filter/fil20.gif
http://www.electronics-tutorials.ws/filter/fil28.gif

If you do need more gain, either of these circuits can provide as much as you want,
up to about six times (+15dB) before the filter amplifier clips from the 0.9Vrms input
(if the filter amplifer uses +15/-15V supplies).
 
Last edited:
Well that is good news! Really good news. I was advised of the 0.9v thing in another thread, but possibly i forgot to specify the HD version. The old one was 0.9v.

So now i just need to specify an rc filter that works with the minidsp and the amp with the lowest attenuation possible. Hopefully i wont need to go active.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.