Okokok, I'll stop impugning your BOB soldering... sounds like it's "something else".
There is apparently no public schematic for the IRIS so we can't check what "they" though they were doing hardware wise. The ADC channels should be independent and just enabling one for reading shouldn't affect the others, however if they are open maybe there's some bleed through. Once you get your connector, try just grounding the new ADC input and see what you read. What version of TOS are you using? It's possible, but somewhat hard to believe, that the ADC code isn't switching channels correctly. Or that the Voltage sensing is not correct. On the mica's one needs to turn on an external voltage ref in order to get a good reading (on mica2 at least, I forget about the micaz). If the VoltageC code isn't doing that, then maybe everything you read is noise. I only know mica2 under TOS1.x. I use 6 of the ADC channels, including battery and RSSI, sampled at TOS_ADCSample240us (the second slowest pre-scale setting). I have not noticed (nor looked seriously) any cross-channel interference. The mica's use a slightly different ATMEGA chip, but it "should" work about the same. I guess you're on your own here. If you can characterize the problem maybe you can get the suppliers, whomever they are now, interested in fixing it... sorry... MS Francesco Ficarola wrote: > Michael Schippling ha scritto: >> The previous photo from the top of the board, looked like the >> red and blue wires (which I thought were the power) were >> coming from the middle of the board and there were two black >> wires coming from one end. That's what I meant about having >> to actually be there to see what was what.... > > Red wire is VCC -pin51- (connected to left or right pin of the pot), > blue wire is ADC1 -pin42- (connected to the central pin of the pot) and > black wires are GND (connected to the left or right pin of the pot): > there are two black wires because in the strain gauge circuit (an > another experiment) I need of two GNDs. > >> Anyway did you try the single wire on a bare connector thing? > > No, at this moment I don't have another Hirose DF9 connector, I ordered > it yesterday. > >> The Photo/Temp sensor is on the micasb sensor board -- the >> MTS300 or 310...I keep loosing track of the part numbers. >> I think the MDA300 also has them, but can't find any actual >> schematics. > > In the MDA300CA ADC1 to ADC6 are single dedicated channels to connect > external devices as a potentiometer or a strain gauge circuit. > >> The only ADC connections on the mica2,z boards themselves >> are Voltage at ADC7 and RSSI at ADC0. I presume that the >> IRIS board is the same, and also that it has the same >> 52 pin expansion layout. > > I tried other channels on the iris. but nothing changes. The principal > problem is that the Voltage readings are dependent on the ADC readings > (any channels). In fact, without sensorboard (or break-out board) > attached, the VoltageC shows values that change as a result of the ADC > readings. > For example, if I enable ADC1 I have these readings: > ADC1 359, VoltageC 357 > ADC1 358, VoltageC 357 > ADC1 358, VoltageC 357 > ADC1 358, VoltageC 356 > ADC1 358, VoltageC 356 > ADC1 358, VoltageC 356 > ADC1 357, VoltageC 356 > ADC1 357, VoltageC 356 > > but if I enable ADC4 (for example) I have these readings: > ADC4 331, VoltageC 335 > ADC4 332, VoltageC 336 > ADC4 332, VoltageC 336 > ADC4 358, VoltageC 335 > ADC4 331, VoltageC 335 > ADC4 330, VoltageC 335 > ADC4 330, VoltageC 335 > ADC4 331, VoltageC 335 > that are different from previous values (with ADC1). This phenomenon > should not happen, right? In fact in telosb node, VoltageC is always the > same and is not depended on the ADC channel. > > So, I meant that the VoltageC component, even without sensor board or > break-out board, mistakes the readings and is depended on the ADC > channel (1, 2, 3, ...). Moreover, with the break-out board and the pot, > if I rotate the pot VoltageC returns different values depended on the > rotation. > > I believe that there is a software problem in iris' driver or a hardware > problem in the mote. > > Greetings, _______________________________________________ Tinyos-help mailing list [email protected] https://www.millennium.berkeley.edu/cgi-bin/mailman/listinfo/tinyos-help
