Puma board for FreeEMS

[jharvey]

That chip was also discussed here viewtopic.php?f=9&t=740&p=8954&hilit=max6675#p8954

I seem to recall it has a slow refresh rate (4Hz ish), but that's OK for typical thermo couple info. The other chip was faster, which may offer some additional BS diagnostics. I'd tend to prefer the the original chip as it has more packaging options, but eh, not a big deal one way or the other really. The SPI makes it easier to use multiple chips and allow a per cyl option.

Another con is that Maxim has a tendency to obsolete chips. So who know's if it will be around in a couple years.

[nitrousnrg]

That chip was also discussed here viewtopic.php?f=9&t=740&p=8954&hilit=max6675#p8954

I seem to recall it has a slow refresh rate (4Hz ish), but that's OK for typical thermo couple info. The other chip was faster, which may offer some additional BS diagnostics. I'd tend to prefer the the original chip as it has more packaging options, but eh, not a big deal one way or the other really. The SPI makes it easier to use multiple chips and allow a per cyl option.

Another con is that Maxim has a tendency to obsolete chips. So who know's if it will be around in a couple years.

I'm wondering if I never read that thread or it was simply flushed from my memory... thanks for make me notice it.

[KW1252]

I disagreed on the issue of measure rate importance in that thread, the measure rates expected to achieve with the faster chip require unsheathed element with wires thinner than human hair. It may be useful in lab conditions for very quick measurement stabs, but in daily use it'd burn up before leaving the front yard.

The measurement speed of a thermocouple is defined in time constant. Time constant is the time it takes the output to reach 63.2% of the final output, and it'll take 5 time constants for the signal to approach 100% of the measured signal. Of course the true temperature can be guestimated by extrapolating the partial output, but greater the difference with sample time and time constant, greater the errors.

Here's a chart, borrowed from Omega engineering about thermocouple time constants:

[jharvey]

Is that one of the parts they put at the end of a datasheet? If so can you share the datasheet it was attached to? I'd like to learn more.

[Fred]

Baldur is using some fast couples, I'll point him here.

[jharvey]

Ah this looks familial.

http://www.omega.com/temperature/Z/Ther ... eTime.html

So, bare wire of a very small diameter, can respond in .1 second, while a larger ball 3/8 in dia probe will be around 80 seconds. Quite a range, and the .1 second is slower that I would have expected.

22 awg appears to be a commonly available size, which according to the chart has a 1 second response time.

A handy ref http://www.omega.com/techref/

[jharvey]

June is rapidly approaching which is noted as the deadline in the first post of this thread. Marco, is there going to be another spin in the near future? Hows the switching supply working out?

[nitrousnrg]

Hi Jared,
sorry for loss of contact, I'm being barely able to survive to the uni, and I have an interesting amount of unreplied mails. To make things worse, I'm the professor of a short course (organized back in 2010, when I had no time issues!), that took all the fun of the last two weeks, along with the exams.

A couple of days ago I tested the SMPS, it needs some work. Basically, it has 10mV of ripple at low loads, and 50mV at 0.6A. I would be ok with that, as long as the regulators fix that up, but guess what? After the MCU regulator I have a bigger (60mV ripple). Didn't try the shutdownable regulator, that must happen on tuesday, when my course ends.

Also, capacitors around the tested regulator were rather small (22uF+0.1uF), so there is some place to improvement. If the easy extra capacitance trick doesn't work, then another filter network would be added. Maybe a regulator swap may fix this too. It just needs some test. There are pics of the board, and of the oscilloscope showing the waveform, posting them on tuesday too.

I still have expectations of manufacturing it on june, it will take 2 weeks of layout/details work, approx. Working the next 2 weeks in the SMPS and the P&H must be plenty of time.

This time issue was expected... I guess now you guys get why I had to send Spin1 to be manufactured on december, so I could work on it while the university wasn't in the middle. I'm not kidding about the uni thing. Where I study, only 4% of the students get the degree, and the *average* duration of it is 10 years. I'm not the best in my class, but the one who is closer to finish the damn thing, and nearly 3 years faster than the average :-)

I just saw the post about the map/aap, thanks for cover me up (that sentence probably doensn't mean what I meant), I used the bosch datasheet that worked mint for me back in the 2007.

See you on tuesday, on a super reloaded mega post!

[Fred]

HOLY SHIT, he's alive! :-o

[jharvey]

Good reply, I'm happy to hear you put school first. I'm not trying to pressure you to get it done. I was just curious.

Sounds like the June deadline may have some flexibility in it. So even if you miss the deadline, you'll likely still keep on the project.

I've been a Professor for a day before, for a machine shop class. I did an independent study the semester before I took the machine shop class. In the study I researched and obtained some really cool surface machining CAM software for the school. I knew the software better than anyone one there. So they asked me to teach the class, while the real professor took notes, such that they could teach it the next year. Was good fun, and certainly a time consuming adventure.

I'll look forward to your Tuesday post. Sounds like interesting results from the SMP circuit.

No problem covering you in the MAP thread. I know English isn't your native tongue, and you're spread thin on time. Glad I could help shorthand your needs to chat it up on the forums.

Good luck on the exams. Catch you later.