High amount of noise in thermocouple reading

[Peaches0204]

Cameron, Sorry about that confusion. So at first when we introduce the electrode to the water (not activated) there is no problem with out LabVIEW output, though the scope has the reading jumping from around 56Hz rate (expected) to around 1kHz. As soon as the electrode is activated, the scope shows a reading of 366kHz and remains stable there while the electrode is running. As soon as the electrode is deactivated, the scope shows a reading 1kHz again. As far as our LabVIEW readout, which we are planning to use as an .exe file, when multiple TCs are used, one is pushed to 0 while another oscillates at a very low negative number. I unfortunately have to run to class so I cannot go more in depth at this moment but will be more than glad to later. I will also try to attach any files we have as to what the LabVIEW readout looks like when the electrode is activated within the saline later

[camerond]

Are you using a real oscilloscope? What do the traces look like? What do you mean "gives a reading of xKHz"? I'm not talking about the output of some software analyzing your noise (even the software contained in some digital scopes), I mean looking at the raw scope trace. All your data which has been massaged through LabVIEW or some other software analysis package tells you is that things are not going right. Take a look at what you're putting into the analysis program to see if it's giving you gibberish or something meaningful. Is there a DC shift when the heater is turned on? What is the magnitude of the high-frequency noise, volts, mV, uV, nV, kV?

Cameron

[Peaches0204]

---

@camerond wrote:

Are you using a real oscilloscope? What do the traces look like? What do you mean "gives a reading of xKHz"? I'm not talking about the output of some software analyzing your noise (even the software contained in some digital scopes), I mean looking at the raw scope trace. All your data which has been massaged through LabVIEW or some other software analysis package tells you is that things are not going right. Take a look at what you're putting into the analysis program to see if it's giving you gibberish or something meaningful. Is there a DC shift when the heater is turned on? What is the magnitude of the high-frequency noise, volts, mV, uV, nV, kV?

Cameron

---

Yes the test was performed using a real oscilloscope, my group partner is going to email me the images we captured (unfortunately only via phone as we didnt have the correct cables in order to capture images directly to the computer). and when i say it gives  reading i mean that the oscilloscope is showing that the frequency sampled jumps to those given values. the magnitude when the electrode is just introduced to the saline was on the order of volts probably around 5-10 if i remember correctly. It was vastly higher than that when the electrode was activated. Sorry if I am short/incomplete with some of the responses. I've either been in or between classes or running around other places. Also I apologize if it sounds like I have no idea what I'm talking about, its part of the reason I am here. I am a biomedical engineering senior with a focus in biomaterials (I've only had limited exposure to circuitry and other related elements).

[camerond]

I don't mean to sound short, but I do think your real problem is not with LabVIEW, but the electronic setup you're using. As Lynn said, you're really exceeding your capabilities with both things in the water. So, since breaking the circuit by holding the TC in air fixes your problem, I think the best move is my first suggestion, to isolate the TC with a nonconductive coating. If this works, you can get on to the interesting stuff.

I would like to see those screenshots, though.

Cameron

[Peaches0204]

Don't worry about sounding short. I'm very aware it is more of a hardware issue than the software (LabVIEW). Thats why we are having difficulty fixing the problem. We are proficient in LabVIEW itself but it's the hardware aspect that is getting to us. I don't mind the criticism or insightful comments at all. Images are still coming and we are going to work on insulating the TC/ attaining a shielded one for testing purposes.

[johnsold]

You indicated that you are a student.  If you were an engineer in a for-profit company, I would suggest hiring a consultant. An electrical engineer with experience in EMI (electromagnetic interference) or instrumentation could probably walk in, look around, make a few measurements, and identify the problem. But it is likely that even an electrical engineering student would lack the experience to be able to do that. Finding a fix might be straightforward or very complicated.

I doubt that you can afford a consultant, but perhaps you can ask someone in the electrical engineering department for some help. The best person might be a staff engineer who maintains the lab equipment, or a graduate student who had 5-10 years experience in industry before returning to school.  Faculty members often lack the practical experience to zero in on the problem quickly.  (No offense meant to any faculty, but I have seen this at several universities including one major research university).

Lynn

[camerond]

---

@johnsold wrote:

Faculty members often lack the practical experience to zero in on the problem quickly.  (No offense meant to any faculty, but I have seen this at several universities including one major research university).

Lynn

---

I couldn't agree with you more, Lynn. That's why I finally threw in the towel and left academia for industry (several years later than I should have, I'm afraid). I like to solve problems, and as an analytical chemist, my orientation towards using whatever technique I could to do so was not generally appreciated (When I would suggest a non-standard solution to another faculty member, either at my university or another, the response I almost always got was "What???"). In my experience, the faculty who do well in academia are ones who drill deeply into one subset of one area and never come up for air. Oh, well, live and learn (and move).

As far as Peaches0204, s/he might do best talking with a person who works in the department's (or college's) electrical shop, if there is one. They are often much better than faculty at fixing pesky problems, since the faculty member often quickly becomes more of a personnel manager and grant writer than a hands-on engineer or scientist.

Cameron

[jab10015]

Hey all,

I'm one of peaches group partners and an idea came from our advisor today. His idea was that a simple attenuator connected to our system could potentially reduce the power and energy reaching the thermocouples. Do you think this is a viable idea? Or will it not reduce the noise effectively? Thanks guys!

[camerond]

What kind of attenuator? Connected to what?

Cameron

[jab10015]

That's where I am confused. There are tons of attenuators out there it's just a matter of finding one that eliminates high frequencies. From what I found in a short time, a VSWR resistance network was mentioned and could reduce these high frequencies (>50kHz). As for where to put it, I am stumped on that as well as I am not very familiar with these circuit elements. Would it connect to the NI 9213 unit we have or the RFE probe?