Thread: Installing/configuring

Originally Posted by inh

No , i didnt read it anywhere, i was just trying different things to see if i could get the board to work. I'd definatly take the time to label at least pins 1-2 and 7-8 on the next revision, so mistakes like that dont happen again. It looks like you have pin 2 running through a resistor for an external led, so that should have prevented a complete short. I'd guess its around 100 ohms, for current limiting the external led. I'll definatly make sure NOT to do that next time though =]

The brain now works pretty well, just having issues with the skin designer saving my changes as soon as i can narrow down whats goin on ill let you know

yes a complete short didnt happen, but USB is extremely picky, and something like that can actually disturb it.

As for the skin designer, if you are more specific about it I might be able to help out.

The skin designer doesnt like to overwrite existing skins... it just crashes..

also, is there a way already, or could it be implemented, so that you dont have to name the skin you want to use fusionformdynamic.skin, maybe just a simple config.ini with a skinname setting? this wont be such a big deal when the skin editor will overwrite existing skins, but until then its been annoying having to rename a skin each time i make a change..

edit: by the way, i resoldered that joint, and theres been no real problems since. i've noticed that if the brain sits there for awhile (few hours usally, but almost always at least one hour) then when i start the control centre app, the temp sensor readings start off as like, -5, then -456 then -something until it ends up like -2895984979348679.9 usually incrementing a few numbers each time the software refreshes the reading.. sometimes unplugging the usb and plugging back in works, other times i have to power cycle. It also seems as tho if the USB is plugged in and then i plug in power, the device will 'fail to start' according to device manager. maybe implement a simple software delay to keep from initializing the usb interface until everythign on the device is absolutly ready to go? i think like 3 seconds is all it needs.

ugh... so, of course, after i type that all is goin pretty good, my brain decideds to not play nice anymore.. it woudl do the same thing, where it would either come up as unrecognized, or fail to start, and rarely, it woould load up fine. like, once in 30 tries...

i think i narrowed it down.. i tried different power supplies, to no avail. i went over all the solder joints on the board, noticed the crystal wasnt soldered perfectly, fixed that, and i think i fixed a join on a digital output. also went over the repair on the regulator, and touched up the other two pins as well. no difference.

then it hit me. i think the problem is with the power input.. it feels kinda rough when i plug in the power adaptor.. so i'm thinking that the voltage to the regulator fluctates a bit while the connector is sliding in.. in turn causing the PIC to go a lil crazy. If i stop pushing the connector in as soon as the led comes on, or i use a switch placed before the ac adaptor it works perfect, like 99% of the time...

i think it just needs a small buffer circuit between the power connector and the regulator. a lil cap should do the trick nicely. if i wasn't stuck over here i'd order one and solder it up to test.. Heres a site that shows what I'm talking about (Also, you may have done this already, and i just havent looked hard enough at the board) just something to think about for the next revision (hopefully they aren't too far away.) Also, I would love to be able to ground the digital inputs and have them trigger, as this would be much simpler wiring wise. lastly, why not ditch the DIP sockets, and have the chips soldered directly to the board? you could use a small header to program the PIC.. Oh, and if it was cost reasons you had the sockets installed, you could switch to SMD chips and it should be much cheaper..

Just tryin to help =] Sorry if im annoying.. just want to help improve this awesome product for everyone.

Originally Posted by inh

ugh... so, of course, after i type that all is goin pretty good, my brain decideds to not play nice anymore.. it woudl do the same thing, where it would either come up as unrecognized, or fail to start, and rarely, it woould load up fine. like, once in 30 tries...

i think i narrowed it down.. i tried different power supplies, to no avail. i went over all the solder joints on the board, noticed the crystal wasnt soldered perfectly, fixed that, and i think i fixed a join on a digital output. also went over the repair on the regulator, and touched up the other two pins as well. no difference.

then it hit me. i think the problem is with the power input.. it feels kinda rough when i plug in the power adaptor.. so i'm thinking that the voltage to the regulator fluctates a bit while the connector is sliding in.. in turn causing the PIC to go a lil crazy. If i stop pushing the connector in as soon as the led comes on, or i use a switch placed before the ac adaptor it works perfect, like 99% of the time...

i think it just needs a small buffer circuit between the power connector and the regulator. a lil cap should do the trick nicely. if i wasn't stuck over here i'd order one and solder it up to test.. Heres a site that shows what I'm talking about (Also, you may have done this already, and i just havent looked hard enough at the board) just something to think about for the next revision (hopefully they aren't too far away.)

We actually have 5 smoothing caps from the regulator to the PIC on there, as well as 1 bigger one from the power jack to the regulator and they seem to work except in your case... Seems a bit strange. If you look at the back (USB and DC jack facing you), you should see a can-capacitor directly to the left of the USB jack, and to the right of the large orange fuse. Do you have that capacitor? It seems like your board has some missing things so I might as well ask. And thanks for cooperating with us, not just ranting.

Originally Posted by inh

Also, I would love to be able to ground the digital inputs and have them trigger, as this would be much simpler wiring wise.

This is how they work. I probably goofed something in the software, so I will take a look. The only thing I didnt test thoroughly each time, were the digital inputs.

Originally Posted by inh

lastly, why not ditch the DIP sockets, and have the chips soldered directly to the board? you could use a small header to program the PIC.. Oh, and if it was cost reasons you had the sockets installed, you could switch to SMD chips and it should be much cheaper..

Actually we had smd drivers on the beta boards. They worked fine and all, until they needed to be replaced. If you accidentally short a digital output's high current pin to ground, the fuse cannot react quick enough for that, and in turn it blows out an entire circuit in the driver chip. So that output will never work again. Then shortly after that (and by shortly we are talking micro/nano seconds sort of timing), the PTC kicks in, and all is saved for then. But then that output is fried permanently. So with the smd, that is just the way it is, but with the DIPs in sockets, if Joe Schmo does this, he can replace it himself for like 40 cents rather than a whole new board at $50. Then the reasoning behing the PIC being DIP is just for ease of programming. Tim programs. and I ship. So he mails me all the programmed chips, I stick them in and test. Also if the analogue inputs receive more than 5v or less than 0v, the PIC will blow out that input. So again, that entire port is useless until it is fixed. So $8 compared to $50 again. The cost thing is not for us. It costs us more to go wth the DIP sized chips and the sockets, but it costs less for you guys if you do something accidentally, like try to measure the output with huge multimeter probes and blow up your chip (cough Tim cough ).

Originally Posted by inh

Just tryin to help =] Sorry if im annoying.. just want to help improve this awesome product for everyone.

We appreciate all the feedback, and you are most definately not annoying. It is good to have people like you who can diagnose, fix, and suggest improvements with the current design. You seem to be pretty well versed in electronics too!

Originally Posted by 2k1Toaster

You have pins 1 and 2 jumpered? Why did you do that?! Take that off immediately. I made sure that they all shipped over pins 7-8 because those are the ONLY pins that should be jumpered.

Pins 3,4,5 & 6 are mystery jumpers, not to be used by the end user.

now i'm curious. Do these jumpers call our fusion brain alien overlords? I for one welcome any sentient being that bring me food as I'm starving.

Originally Posted by 2k1Toaster

If you look at the back (USB and DC jack facing you), you should see a can-capacitor directly to the left of the USB jack, and to the right of the large orange fuse. Do you have that capacitor? It seems like your board has some missing things so I might as well ask. And thanks for cooperating with us, not just ranting.

Yea, i have the cap there.. Like i said, hadnt really spent too much time tracing the traces out. If you guys want, since the software is open source and all, maybe you could throw up the files for the PCB? I dunno what program you use, but at the least a .jpg of the top and bottom traces with vias and such would be awesome for debugging (at least for me) =]

i dont even have a multimeter here (yet) but im going to order one in a few mins, because we've been having crazy issues with our generator and electrical wiring, and doing ANYTHING without a multimeter sucks =[

This is how they work. I probably goofed something in the software, so I will take a look. The only thing I didnt test thoroughly each time, were the digital inputs.

i think i read that they need + voltage in the user manual pdf..

Actually we had smd drivers on the beta boards. They worked fine and all, until they needed to be replaced. If you accidentally short a digital output's high current pin to ground, the fuse cannot react quick enough for that, and in turn it blows out an entire circuit in the driver chip. So that output will never work again. Then shortly after that (and by shortly we are talking micro/nano seconds sort of timing), the PTC kicks in, and all is saved for then. But then that output is fried permanently. So with the smd, that is just the way it is, but with the DIPs in sockets, if Joe Schmo does this, he can replace it himself for like 40 cents rather than a whole new board at $50. Then the reasoning behing the PIC being DIP is just for ease of programming. Tim programs. and I ship. So he mails me all the programmed chips, I stick them in and test. Also if the analogue inputs receive more than 5v or less than 0v, the PIC will blow out that input. So again, that entire port is useless until it is fixed. So $8 compared to $50 again. The cost thing is not for us. It costs us more to go wth the DIP sized chips and the sockets, but it costs less for you guys if you do something accidentally, like try to measure the output with huge multimeter probes and blow up your chip (cough Tim cough ).

awesome! im that case, keep the DIP stuff =] Way to think ahead, or maybe learn from experience

Im gonna play with my brain some more, see if what i suspect is the problem really is the problem, and if so just deal with it for now. When the next revision of them comes out (as it inevitably will) i'll order one to use in my final project. I still got like 10 months to get my car pc setup before i go home and can install it, so im good for now =]

[/quote]
We appreciate all the feedback, and you are most definately not annoying. It is good to have people like you who can diagnose, fix, and suggest improvements with the current design. You seem to be pretty well versed in electronics too![/QUOTE]

Not too long ago, i had an idea for a product. i drew up a schematic, and then the next thing i know im learning how to design PCBs and have them produced overseas.. Set up a storefront using OsCommerce just like you guys, and the next thing I knew I was in business =] I really appreciate and respect all you guys have done. I know its tough as hell sometimes, but you guys are really doin great work!

Originally Posted by Ruffy

now i'm curious. Do these jumpers call our fusion brain alien overlords? I for one welcome any sentient being that bring me food as I'm starving.

They might...

Originally Posted by inh

Yea, i have the cap there.. Like i said, hadnt really spent too much time tracing the traces out. If you guys want, since the software is open source and all, maybe you could throw up the files for the PCB? I dunno what program you use, but at the least a .jpg of the top and bottom traces with vias and such would be awesome for debugging (at least for me) =]

We would rather keep the PCB and firmware code under wraps at least for now. We are happy to answer questions though. And I am sure with enough determination you can reverse engineer it since it is only a 2sided PCB, but we would like to keep that along with the mystery jumpers to ourselves for now.

Originally Posted by inh

i dont even have a multimeter here (yet) but im going to order one in a few mins, because we've been having crazy issues with our generator and electrical wiring, and doing ANYTHING without a multimeter sucks =[

That sucks. I would be lost without my multimeter. Even the $10 Radioshack multimeter would be better than nothing... Well, maybe.

Originally Posted by inh

i think i read that they need + voltage in the user manual pdf..

Actually you are correct. I misread your question. I dont think it would be simpler either way for sending 0v or 5v into the chip on contact. For now though it does send +5 into the chip.

Originally Posted by inh

awesome! im that case, keep the DIP stuff =] Way to think ahead, or maybe learn from experience

Im gonna play with my brain some more, see if what i suspect is the problem really is the problem, and if so just deal with it for now. When the next revision of them comes out (as it inevitably will) i'll order one to use in my final project. I still got like 10 months to get my car pc setup before i go home and can install it, so im good for now =]

Sounds good.

Originally Posted by inh

Not too long ago, i had an idea for a product. i drew up a schematic, and then the next thing i know im learning how to design PCBs and have them produced overseas.. Set up a storefront using OsCommerce just like you guys, and the next thing I knew I was in business =] I really appreciate and respect all you guys have done. I know its tough as hell sometimes, but you guys are really doin great work!

Thanks for the props! OsCommerce is really awesome for being the low low price of free. Of course all the extra tabs I had to code up myself, and it wouldnt accept multi-currency or multi-language like I wanted so all that I re-wrote. Other than that it is definately nice.

I completely understand wanting to keep the pcb and code under wraps. I hope to see it at some point in the future though =]

Man.. speaking of $10 radioshack multimeters, i have one of those somewhere.. along with a sears brand one i swear is just re branded fluke but who knows, it was free and works great =] Hrm, maybe i'll see how much it is for a decent fluke.. *prepares for sticker shock* Amazon.com is great. We can get all the movies/books/microwavable food/electronics we want shipped over here!

Any ETA on the com interface for the software? Have you guys looked in to just tossing in the code from the roadrunner source, in the SDK example. Its ported to like 6 or 8 differnt languages =] I'd do it myself, but i don't have visual studio (anymore.) I do have VB though, and it was very easy to move the com interface portions of the code to my own app, so that it could communicate with roadrunner (or any other app.) All we really need is support for 3 commands; reading analog and digital inputs, and setting digital outputs.. Man i wish i had visual studio so i could do this. I'd have a climate control whipped up for my carpc in no time =]

Hrm, kinda way off topic I'll try and keep things focused and make threads for different topics from now on =]