Electronics > Projects, Designs, and Technical Stuff

EMF pickup from amplifier in I2C line causing glitches. (Now with scope trace!)

<< < (16/31) > >>

dmills:

--- Quote from: Starlord on July 06, 2016, 11:58:49 am ---That's cool, and clever it may be, but it's still not perfect. At my sister's place, whenever they use their microwave, the WiFi goes down. It's a pretty new microwave too, and there was no visible damage to the door that might have caused leakage.

--- End quote ---
And thats fine, the wifi product "Must accept interference", there is not (and never has been) a guarantee that a user of an ISM band will not suffer interference from another user, or that a product may not radiate (within defined limits) power in an ISM band (They are after all unlicensed).

If those door seals are good for 40 - 50dB then a 1KW oven might be radiating 10 - 100mW, not a problem because it is doing it in the ISM band, if it was the ovens power inverter radiating a load of hash via the power cord then that would likely be a problem because it would NOT be in the ISM band. 

Regards, Dan.

Starlord:
Well it seems we're at am impasse, CJay.  I'm not going to give up on designing electronics, and you're not interested in helping me learn how to design them in such a way that they don't cause interference.  I'm literally sitting here asking how to design filters to reduce the interference from my amplifier, and you're giving me shit because, I guess, you think I'm too stupid to be allowed to design any consumer product.

Btw, all this arguing is eating into what little time I DO have left to redesign this board so it doesn't interfere with everything around it.  Just putting that out there.

Starlord:

--- Quote from: dmills on July 06, 2016, 11:52:02 am ---You should really be plotting the voltage across R1, not that it really changes anything, and yes radians in that datasheet means radians per second.

--- End quote ---

I tried dropping a multimeter in the schematic to measure across R1, but I can't get the voltage reading to appear in the AC analysis, and the transient analysis isn't plotting the data in a way which is useful.

You said it doesn't really change anything.  Does that mean that -6db drop isn't really there?  And then why is the graph showing a 96db drop at 100Mhz when you suggested that a 1uH indicator + 100nF cap would lead to a much lower drop?  I think you said it would be around 26db?  I can't trust the graph and tweak my component values because I'm getting different results than I expect.

CJay:

--- Quote from: Starlord on July 06, 2016, 12:35:03 pm ---Well it seems we're at am impasse, CJay.  I'm not going to give up on designing electronics, and you're not interested in helping me learn how to design them in such a way that they don't cause interference.  I'm literally sitting here asking how to design filters to reduce the interference from my amplifier, and you're giving me shit because, I guess, you think I'm too stupid to be allowed to design any consumer product.

Btw, all this arguing is eating into what little time I DO have left to redesign this board so it doesn't interfere with everything around it.  Just putting that out there.

--- End quote ---

I'm being a b**tard to you because you don't seem to want to realise just how bad this could get for you if you don't do it right.

Don't give up, I honestly don't want that.

 I have a lot of admiration for people who design and get things to market, I've done it in the dim distant past before the stringent EMC laws and know your pain, but you are trying to run before you can walk and are leaving yourself open to immense legal problems.

Nothing you can do or say relieves you of the obligation to comply with the laws of your land but as others have hinted, there are *legal* ways to pass the responsibility on to the end user.

That's not to say you don't have to make every effort to ensure the product is as good as it can be with regard to those laws.

Starlord:
Well, I installed a free version of Multisim to try it out, but I'm getting nowhere fast trying to simulate an LC circuit with a 4 ohm load.  I think I'm supposed to use the bode plotter, but after creating the LC circuit with an AC voltage source (since I couldn't find something that should produce noise) and attaching the inputs to the bode plotter across the load, and running the simulation, it complained my circuit wasn't grounded, which makes no sense since it I connected both the + and - sides of the AC source to make a complete circuit.  I'm sure this is wrong, but it's really frustrating because it seems like the right thing to do.  I added a ground on the negative side just to get the error to go away, but I'm getting nothing on my bode plot when I run the simulation.  I set F and I to 100MHz and 100Hz and the AC source is 1Hz. 

I tried to find some Youtube tutorials on how to make this all work and was met with even more frustration because almost none of the damn things have any audio.  There's no one explaining why they're doing what they're doing!

I don't know why this is so difficult. It's a digital amplifier. It's an LC low pass filter / crossover. It's a common thing used everywhere. Why the hell isn't there some audio site or software out there with a graphing calculator specifically for this purpose?

I mean I could just use the 1uH inductor and .1uF cap that was suggested here and hope it works, and maybe replace the caps once I have the boards made to see how different values change things, but I'd really like to see for myself what the graph looks like, and I'd like to try out a 2.2uH inductor with maybe a .22uF cap?  And see how much better that is, and make sure there's no peak.  And I'd like to see how large the peak is and see how things change when I reduce the load from 4 ohms to 2 ohms, because both will potentially be used.

Navigation

[0] Message Index

[#] Next page

[*] Previous page

There was an error while thanking
Thanking...
Go to full version
Powered by SMFPacks Advanced Attachments Uploader Mod