| alitaido |
| So, This is a "design" I'm thinking about trying. Not much unusual, but if anyone sees any glaring problems, let me know! |
|
|
| SpittinLLama |
What is the purpose of R1 and R2? C1 and C2 at 680pF seems a little high to me but I don't think it will be a problem. Just seems smaller values are more common (I don't have any data comparing values to tell you an optimum value). Do you find that the L + R on the output are really necessary? For my amps I have never used those components, only the snubber on the ouput. Don't seen any problems. Surely your power supply will have much more capacitance than the 01.+ 10 + 1000 on the chips.
-SL |
|
|
| alitaido |
Yes, the PSU will have about 10000uF per rail, those rail caps you see in the schem are just "buffers".
R1, R2, the 680pF caps up through the "standard" 1uF DC bypass are all part of an input filter. I once saw someone use a similar design and I was intrigued, so I put it into PSPICE to see how it behaved. I decided on the 680 since it more agressively attenuates any high frequencies (above 100Khz). Of course, that shouldn't be coming in through the audio anyhow, but since I keep hearing about the instability of these chips in the HF range, I thought I'd try a little insurance. |
|
|
| jaycee |
All looks good to me, I've just built a similar circuit myself :)
http://www.darkmatter.myby.co.uk/pr...p-schematic.png
As you can see I didn't bother with the capacitor on the MUTE pin. I found it unneccesary as I get no clicks at switch on without it, as the LM3886 includes supply monitoring that doesn't unmute the amp until the rails are >12V.
R7 on the diagram really has an inductor wound around it.
The capacitor between the + and - inputs is unneccesary if you have an ultrasonic filter on the inputs, provided your PCB is well laid out. You can also get away with no Zobel network, but if you really must have one, you can put it on the speaker output terminals instead of on your PCB. |
|
|
| alitaido |
Thanks for the info , Jaycee.
How do you handle the LR on the out? I mean, do you actually buy an inductor, or just wrap some wire around the output resistor? My DMM can do just about everything except inductance, so I worry a little bit about trying to wing it with a wire wrap. Who knows what L value I'd come up with. |
|
|
| jaycee |
I just wound the wire around the resistor. It's not actually that critical. If you buy, say, a 3W ceramic resistor and wind some enamelled wire around it, that's about right.
You can see the end result here:
http://www.darkmatter.myby.co.uk/pr...es/S2020016.JPG |
|
|
| darkfenriz |
alitaido
C5 and C6 are intended for better stability?
Are you sure it works like this? |
|
|
| alitaido |
"C5 and C6 are intended for better stability?
Are you sure it works like this?"
in the case of these 220pF caps, I am taking the idea directly from the NS datasheet of the 3886. |
|
|
|
