Electronics > Projects, Designs, and Technical Stuff
EMF pickup from amplifier in I2C line causing glitches. (Now with scope trace!)
<< < (26/31) > >>
Starlord:

--- Quote from: engineer_in_shorts on July 14, 2016, 01:11:25 pm ---Never said anything about EMC - don't care.  I was curious about you asking for help on a forum and a number of top posters trying to point your project in a different direction, only for you to ignore them.

--- End quote ---

I haven't ignored anything anyone has told me.  Weighing advice against one's budget and deciding you can't afford to implement it doesn't mean you're ignoring the advice, have learned nothing, and won't take said advice into account when designing future devices.


--- Quote ---Nobody is initiating any witchhunts.

--- End quote ---


--- Quote ---And yes i did read the post about your business issues. In fact it is nicely archived here:

--- End quote ---

Nobody's going on a witch hunt!  He's just trying to find out who you are so he can publicly expose you, and digging up archives to old posts you specifically wanted to avoid having people read!



--- Quote ---http://web.archive.org/web/20141220033446/https://www.eevblog.com/forum/projects/customer-w-unrealistic-expectations-what-would-you-have-charged-for-this-job/?

--- End quote ---

After I posted that I was surprised and overjoyed that I got such positive and supportive and helpful responses from everyone.  It made me gad I came to this forum, and think highly of the people here.  What's happened in this thread... not so much.


--- Quote ---If you are struggling to put food on the table STOP tinkering with these projects AND GO AND FIND A SALARIED JOB --- IMMEDIATELY. If you have one and it's not enough money/hours then go and find another TODAY.  Trust me, it's for your own health both physically and mentally.

--- End quote ---

And do WHAT?  I don't have a degree in electrical engineering.  I'm all self-taught.  Nobody would hire me.  And my software engineering skills are for the most part 20 years out of date.  Do you know anyone who's looking to hire a C++ programmer who is well versed in Arduino?  Didn't think so.  Right now, it's this, or work retail, and I'm earning more money doing this than I would be working for minimum wage.  And as a side bonus, I don't want to die every morning when I get out of bed.  I'm poor as shit and barely scraping by but I love what I'm doing, and I've learned a ton these last six years.  Having no money all the time is stressful to be sure but I wouldn't be any better off in a dead end job, and I think in another six years I might be able to grow this into a sizable business, if I can just get over this hill I've been pushing this boulder up for so long.
janoc:

--- Quote from: Starlord on July 14, 2016, 01:27:20 pm ---But I would ask why if shielding is so necessary with data transmission over CAT5 cable, why most CAT5 is not shielded?  Does it operate at frequencies that aren't in important bands?  Or does the differential nature of the signal cancel out the noise?  If I used a differential I2C driver chip, would I not need shielded cable?

--- End quote ---

Actually a lot of CAT 5 wiring is shielded. I would say both the differential nature of the signal (thus the fields cancel out) and probably also the shape of the signals and termination of the cable is what makes the difference.

I am not sure what you mean by "differential I2C driver chip" as I2C is defined as single-ended system. However, there are translators/drivers for e.g. RS485 with I2C interface. So you can run RS485, which is differential, over your cable instead of fighting with I2C which was not designed for that. However, doing that will likely require changing the protocol you are using, because RS485 is not I2C, obviously. So depending on what you are communicating with, you may have to add/reprogram the MCUs as well.
janoc:

--- Quote from: Starlord on July 14, 2016, 01:44:58 pm ---And do WHAT?  I don't have a degree in electrical engineering.  I'm all self-taught.  Nobody would hire me.  And my software engineering skills are for the most part 20 years out of date.  Do you know anyone who's looking to hire a C++ programmer who is well versed in Arduino?  Didn't think so.  Right now, it's this, or work retail, and I'm earning more money doing this than I would be working for minimum wage.  And as a side bonus, I don't want to die every morning when I get out of bed.  I'm poor as shit and barely scraping by but I love what I'm doing, and I've learned a ton these last six years.  Having no money all the time is stressful to be sure but I wouldn't be any better off in a dead end job, and I think in another six years I might be able to grow this into a sizable business, if I can just get over this hill I've been pushing this boulder up for so long.

--- End quote ---

Um, well, so you solution is to pour more money down the drain, because you don't have marketable skills?  :palm:

It is great that you are doing what you love to do, but I think you would have been better off trying to update the skills you actually have first, solve your financial problems and only then try to start a business. Or at least recognize that you are way out of your skill level and get help early. It is going to get only much more expensive later when you have burned through a ton of cash already.

Unfortunately engineering and business don't reward effort, only results, so that you have spent years "pushing a boulder up the hill" sadly doesn't mean it will pay off.

I wish you luck, you will really need it. You have more balls than I would have in such situation.


Starlord:

--- Quote from: janoc on July 14, 2016, 01:27:09 pm ---The argument about Lady Ada and Sparkfun is not really relevant neither, because they are selling kits and components, not complete products. These don't have to be EMC tested, it is the responsibility of whoever integrates them into a final product. It doesn't make sense to test a component without the enclosure or other parts of the final system because they all can affect the performance - e.g. a metal grounded case acting as a shield vs. a plastic one which does not. And even then both Sparkfun and Adafruit are actually applying for FCC testing these days, even though they don't really have to.

--- End quote ---

They are selling products to people with the knowledge that the people buying them don't know what they're doing, and will wire them up in such a way that they're likely to emit all sorts of radiation. You're just giving them a pass because they're helping people learn electronics rather than simply designing a finished product and selling it.  But what if I told you my boards don't come with a battery or a power switch, or a speaker or even an enclosure, and that it's the end user's job to add those things?  Is it my fault if they don't install it in a shielded metal case that is grounded or that their speaker wires won't be less than three inches long?  I know they're not going to do that, but Adafruit and Sparkfun also know their end customers won't be doing that when they buy the convenient electronics "kit" that is a battery pack, microcontroller, and neopixels to solder together and stick on their clothes, sneakers, or backpack.  And I know for a fact that the amplifiers Adafruit sells have no more than a ferrite bead filter on the output and just require you to plug in a 12V supply.  How could they possibly expect that to be used in a manner that won't cause EM interference?   


--- Quote ---Oh puh-leeze. You have been told and explained many times what you have to do, but you shoot down most of the proposed solutions right away. Or you lack the understanding or means to implement them. And that is somehow the fault of everyone else but you.

--- End quote ---

I didn't shoot them down.  I explained that I don't have the funds to implement them, and asked for alternatives.

I can take criticism of my designs just fine. What I can't take are personal attacks because you don't like the fact that I have two hands tied behind my back and I'm doing the best I can with what I've got.


--- Quote ---And the deal with the new account - you somehow think that if you do that people will miraculously start solving your problems for you, because obviously the only thing standing in the way is them knowing it is @Starlord who is asking these questions?

--- End quote ---

I think that now nobody is going to be willing to answer any of my questions because they know I'm trying to make money at this and that I can't afford to do full EMC compliance testing.


--- Quote ---You are asking people to do free consulting for you, refuse to provide information, shoot down the proposed solutions, think that rules don't apply to you for some reason and show incredible arrogance in general and now think that making a new account will somehow change this? Do you think that everyone is an idiot?

--- End quote ---

I don't know what information it is you think I've refused to provide, unless you mean my name and address, and if that's what you mean, too bad!

And incredible arrogance?  The more I learn about electronics, the more I realize how little I know.  And being unwilling to simply take people's money and run, which what you're asking me to do, isn't arrogance, it's honesty.  Also, wanting to keep a roof over one's head?  Also not arrogance, just a will to survive.

You mistake my unwillingness to implement the solutions presented to me because I can't afford to as me thinking I know better than everyone else.


--- Quote ---A guitar pedal with a few opamps or linear power amp is typically not going to spew interference all over the place, unlike your class D power amplifier switching in the megahertz range and connected using long wires. If you don't see or understand the difference between these, I think you really should first educate yourself more before you dig yourself into an even bigger  hole than you are in already.

--- End quote ---

And yet someone relayed a story in this very thread about how I should be so scared of the FCC because they'd heard of some boutique guitar pedal designers getting fined.  So either they're lying, or you're wrong.  And I admit, I have no idea which is which.
Starlord:

--- Quote from: janoc on July 14, 2016, 01:50:40 pm ---
--- Quote from: Starlord on July 14, 2016, 01:27:20 pm ---But I would ask why if shielding is so necessary with data transmission over CAT5 cable, why most CAT5 is not shielded?  Does it operate at frequencies that aren't in important bands?  Or does the differential nature of the signal cancel out the noise?  If I used a differential I2C driver chip, would I not need shielded cable?

--- End quote ---

Actually a lot of CAT 5 wiring is shielded. I would say both the differential nature of the signal (thus the fields cancel out) and probably also the shape of the signals and termination of the cable is what makes the difference.

I am not sure what you mean by "differential I2C driver chip" as I2C is defined as single-ended system. However, there are translators/drivers for e.g. RS485 with I2C interface. So you can run RS485, which is differential, over your cable instead of fighting with I2C which was not designed for that. However, doing that will likely require changing the protocol you are using, because RS485 is not I2C, obviously. So depending on what you are communicating with, you may have to add/reprogram the MCUs as well.

--- End quote ---

NXP sells a buffer chip that takes I2C as input, converts it to a differential signal that can be sent over two twisted pairs, and can then convert it back at the other end:
http://www.nxp.com/documents/data_sheet/PCA9615.pdf

They also sell a version with 3 lines, which I'm guessing is for I2C chips that have an interrupt:
http://www.nxp.com/documents/data_sheet/PCA9616.pdf

I've mentioned this chip before in the thread.  I considered using it at one point early on in the project, but due to costs and my calculations indicating the system would work without it, I decided not to go with it, and I put the spare lines to use for some other stuff so I can't really go back now and use it.  Unless maybe I don't have to worry about voltage droop and my logic levels with this in place, in which case maybe I could get rid of a couple of my power lines to make room for it.  But though that might work it would have to be saved for a future revision.
Navigation
Message Index
Next page
Previous page
There was an error while thanking
Thanking...

Go to full version
Powered by SMFPacks Advanced Attachments Uploader Mod