Must-haves for EASY DIY controller?

[mammut89]

either way , we have progress !

 

[DustinB]

DWZM: Can you take a look at the schematic and see if you see anything different with this: http://www.arduino.cc/cgi-bin/yabb2/YaBB.pl?num=1280420867/all

"China Red's" board halfway down the page. Maybe you can spot something? I don't have eagle here at work.

 

[der_wille_zur_macht]

That's actually my thread. I've checked vs. the schematics posted there, but will do so again.

 

[der_wille_zur_macht]

And I'm guessing we can rule out software, if you're using the same sketches you got working with the commercial shield?

 

[DustinB]

Yeah I'm sure it's not software. I did use the same sketch with the shield connected to the second AVR. The only other thing I can think to check if you don't see anything wrong with the magjack connections is to start looking at the gate on the shield.

 

[der_wille_zur_macht]

Level shifter checks out pin for pin against China Red's board and nuelectronics' schematic. Looking at the magjack now but it's a little trickier to check because of all the passives and the fact that the silkscreen labels are obscured in China Red's image. . .

 

[DustinB]

I grabbed eagle real quick to check some things out and something popped out at me. Can you verify this, Pin 8 on the magjack is not connected to the ground plane?

 

[der_wille_zur_macht]

5, 7, and 8 are not connected to anything. It seems that's the case on other designs as well. Is 8 connected to something in your shield?

 

[DustinB]

I need check my shield, but I was looking at the HR911105A datasheet and it calls for pin 8 to be grounded. http://www.datasheet4u.com/download.php?id=675627

 

[der_wille_zur_macht]

That datasheet labels pin 8 as "CHS GND" which I am assuming is "chassis ground?" If so, I think we're all set since pins 13 and 14 (the actual metal chassis tabs extending down below the board, ironically not in the datasheet) are grounded.

But you never know. Might be worth a try.

 

[DustinB]

Actually this appears to be quite different from what I can see in the pics my g/f sent me. I'll have to grab the meter tonight and verify everything. On this pin 8 seems to be connected for sure through a resistor, possibly ground?

 

[der_wille_zur_macht]

Definitely, that is different than both the Hydra implementation and the nuelectronics schematic. Are you sure the pinouts on the magjacks are identical?

 

[DustinB]

Same thing, I pulled the HR911105A off this board to put on my hydra just in case the sparkfun jack was the problem.

 

[der_wille_zur_macht]

Hydra and nuelectronics have connections on 1, 2, 4, 6. That board appears to have connections on 1, 2, 3, 7, and 8.

 

[DustinB]

Certainly looks that way. I am hoping these are just simple pwr/gnd connection differences. I was wondering about pin 8 as well after you mention chassis ground, but it does appear that the chassis is grounded as well. Either way it's pretty weird considering the "China Red" guy built a board on the exact nuelectronics circuit, unless he neglected to mention a ground on pin 8 or something else.

Not much more to do til I get home later I guess. I'll figure out exactly what is connected to what across the board.

 

[mammut89]

im feeling rev 2.0 is comming up real soon getting ready to send off gerber files ! heh.

I have a hunch that its the ethernet connector pins that are causing any problems if there are any at this point. not that I can see any problems with the schematics, then again I didnt see the last problem either until it was pointed out

 

[DarthAWM]

im feeling rev 2.0 is comming up real soon getting ready to send off gerber files ! heh.

I have a hunch that its the ethernet connector pins that are causing any problems if there are any at this point. not that I can see any problems with the schematics, then again I didnt see the last problem either until it was pointed out

I hope so too, I am planning on building when 2.0 is up

 

[DustinB]

Definitely don't send any files anywhere yet. Even after being fixed we still have to decide if we want any jumpers for the ENC AVR selection or anything else. Aside from that though, I believe everything else is working without any problems.

Even after this is complete we have to decide how the software part is going to work. If the main goal of this project was something fairly simple, are we going to have a basic code block for all integrated features? Not everyone knows how to or has the capacity to program. I suppose when Rev2 comes out we could use our fixed proto boards to develop a fully functional base sketch.

Anyway, I'm home now. Getting back to it.

 

[DustinB]

Ouch, bad news. That's it for me for about 3-6 weeks. I just found out the other magjack I had is in fact not the HR911105A, it's the HR911102A.Don't know how I missed that. Not only that but the sparkfun magjack schematic is different too, even after the guy assured me they were the same.

The circuit may very well be working right now, but I'll have to wait for the part to come in from china. :hmm5:

EDIT: Then again I just found this on mouser: http://www.mouser.com/ProductDetail/Pulse/J0011D21E/?qs=sGAEpiMZZMvQhAhQbXdbBpokefP7V/dtVIJe0ry9M94=
What do you think DWZM? The schematic looks the same. Page 5, "J0011".

 

[terahz]

Ouch, bad news. That's it for me for about 3-6 weeks. I just found out the other magjack I had is in fact not the HR911105A, it's the HR911102A.

Pm me your address, I'll put one in the mail for you.

Just a summary:
Ethernet fixes so far:

ENC AVR-slave
CS (p9) -> d10
reset (p10) -> reset
p22 -> gnd
p24 -> +3.3V


Will post soon with the results.