In a newsletter I got today (Full Spectrum Diagnostics, http://www.fullspec.net/information.html) the author, Dan Ambre, suggested setting up a special "diagnostic route" to load in addition to the regular PdM route for taking extra data when a problem is suspected.

Using a 2120 I always take "extra data" anyway, but it's a pain to specify the setup each time I do (I don't find the "expert" setups very helpful). But CSI sets the collection up at the "measurement point" level wherein you specify the parameter set for each point, and you can't change that in the field (can you?). Any ideas about how we could simplify the collection of extra data? I'd prefer to have it stored in the database along with the regular route data.

Rusty,

Do you use the "Acquire mode" for saving your extra data? I find it easy to modify the setup for the point of interest and you're allowed to save the data directly to the Measurement point as "Acquired data". I particularly like this method because I can customize the measurement for better resolution or a different frequency range without going back to my computer. Then the data is saved in the database without affecting your trend data.

Rusty, I'm not sure if I follow your question. I've always use the aquire feature to store information to the measurement point. The data will be tag either "R" for route, or "A" for Aquired on screen BB01 under manual plotting. See "Tag Field Descriptions" in the help index.

PS: as a default the aquired data does not include trends, which could be changed by aquiring different information than was in the original AP settings.

This message has been edited. Last edited by: Waylon M,

Waylon,

If I understand Rusty right, he would like to make a route of follow-up readings that he decided to take as he was doing the report.

Say you are looking at the data on a fan. When you collected the data you saw maybe .2 ips at running speed axailly on the fan bearing. You may or may not take notice of it during the route, but when you are doing the report, you think, well that could be imbalance or it could be misalignment, I will check that next time.

I think those are the things Rusty is looking to be able to set up to take next time. That way he wouldn't be reliant on his memory or have to refer to his report when preparing for the next survey. Just remember to load the follow-up route and you would be ready.

(At least that is what I'd like to do.)

Are you back at work yet?

Danny

Danny Harvey,
I once made little yellow stickers with acronyms such as brg. misl. This worked great until the production superintendent deciphered my code. Why don’t production people like to know how many issues they really have?

Unlike you and Rusty being nested has the advantage that you are never far from the equipment. However, I could see where being far away from the equipment has advantages as well.

This message has been edited. Last edited by: Waylon M,

Rusty,
I don't believe you can set up parameters when collecting extra data not already defined in a route. I know you can't in Acquire Spectrum nor can you trend it and I just tried off route mode and could not find a place there for parameters. If you acquire a spectrum nothing shows up above F3 where you access your parameters. I still use acquire spectrum for my extra data because it is attached to the measurement point and you can add a note or just hit the mark key and it sets a field alert note just to remind you where you took the extra data. I don't know of anything easier than acquire spectrum, at least for me, and if I did set up the extra route, I don't think it would cover all the possible scenarios of extra data I might want on a given problem.

Ronnie

We used to have a dummy machine configured into and residing at the end of each route. This was 4 bearings (each V, H, A) set up to obtain hi-res spectra and waveforms e.g. velocity spectra to 1000Hz at 3200lines and acceleration spectra to 5KHz at 6400 lines, along with any other useful parameters that the system could gather.
If required for an ongoing investigation, the dummy was converted to a real machine by renaming it with an S at the end of its tag e.g. P105 would become P105S: this allowed collection of routine data and the collection of additional hi-res stuff which remained in the database, just in case it was required again at some future date.

Rusty I too add a selection of high resolution diagnostic pre- built machine to every route.
going off route with my wquipment is very easy but there is nothing better than picking a pre built machine.
if I have to use it I just up load it and report on the findings

Buzz, mcdm... that's what I was looking for. I too use "acquire" mode heavily, but I find I go through about the same routine for extra diagnostic data.... 5kHz acceleration spectra/waveform, hi-rez velocity spectrum, low-rez velocity spectrum (to get a shorter, higer resolution velocity waveform)..... transferring the data from a "dummy" machine to the actual machine would be pretty easy if they are configured exactly alike.... will have to give it more thought. Oh, CSI has really useful parameter setups for phase data as well which would be good to have available as a dummy machine.

On routes where different people collect data from month to month, I define a machine called 'WHO' with a single point (keypad input) and you put in the number of the person who is collecting the data. The 'key' goes in the machine point description field.... 1=RC 2=DR 3=JR ... that way you know who collected the data and makes for a nice trend plot.

I have a point defined for many machines as 'Status' and I use a numeric code to note the apparent condition of each machine 1=OK 2=Watch 3=Problem, etc... this code shows up the next month and reminds me what is going on with that machine.

There are lots of 'special' things you can do with data collection to make things easier if you are handling a large number of machines/routes/plants.

Rusty,

I have a dummy machine setup that collects everything that I think I might want. It takes several different TWF that vary in length so it will get a number of different mchine speed. It also takes different high res spectrums and spike energy.

I have it set up to take data on any point of a machine train. When this is downloaded into my database I just add a note as to what it is.

Note: These points take a lot of data and the acquisition time is loooong. It usally takes more data than I need but it is ready anytime I run into a problem in the field.

I'm thinking that a 'dummy' machine could contain ALL of the standard measurement points that you might use -- motor, pump, fan, gearbox - and you simply select the ones that "fit" the machine you are looking at. That way, if you want to transfer the detailed data to the actual machine they came from, you don't have to rename any points... just select the points used when you do the merge with the existing machine.

Rusty,

The other advantage of a Dummy Machine is that the extra/special Parameter Sets become available in analyzer if you use Off-Route to define extra measurement points.

Walt

Very good point, Walt... I hadn't thought of that.

I am even more convinced that given a little time, most all of us could do a much better job of handling data collection and database issues... if we just had the time (or maybe if we didn't have a life outside of work )