Sign up Calendar Latest Topics Donate
 
 
 


Reply
  Author   Comment  
Danny Harvey

Sr. Member / Moderator / Supporter
Registered:
Posts: 1,201
Reply with quote  #1 
I posted this in the equipment forum but figured I'd put it here too because it might not get seen there.

SKF uses 4 filters for its gE demod readings: 300-6 kcpm, 3k-60k cpm, 30k-600k cpm and 300k-2.4 mcpm.


Do these filter out everything but what is between the two numbers or do they include on what is between them?
John from PA

Sr. Member
Registered:
Posts: 803
Reply with quote  #2 
One way to remember this is to recall the terminology "enveloping".  In other words the content demodulated is between the limits of the filter.  So a 30k to 600k examines that range to the exclusion of others.  Having said that I do not know the corner frequencies of the SKF filters.

See the article at https://www.engr.colostate.edu/~dga/mech307/handouts/bearing_signature_analysis.pdf for more information.
MarkL

Avatar / Picture

Sr. Member
Registered:
Posts: 902
Reply with quote  #3 
Danny as a general rule of thumb I use Env3 on 85% of the stuff we monitor, stuff under 300rpm, we use  env2 and env 4 as it seems to work better. 
Danny Harvey

Sr. Member / Moderator / Supporter
Registered:
Posts: 1,201
Reply with quote  #4 
John,

That's the logical way but I have found that is not always the way things are named.  Like in Windows 95 when we began going to "start" to stop our computers. This is the 5th time I have learned new software and hardware. It's not as easy to learn new stuff as it used to be but experience has taught me to question things before I assume logic.

In this case, I'm pretty sure it's right though.  It just doesn't make sense to do it the other way. I'll find out for sure today and confirm it.
Danny Harvey

Sr. Member / Moderator / Supporter
Registered:
Posts: 1,201
Reply with quote  #5 
Mark,

Are they named in numerical sequence from lowest freq to highest?

3 would certainly be the most useful in a plant-wide setting. I'll use the others where appropriate but the bulk will probably use 3.

Thanks for the help
MarkL

Avatar / Picture

Sr. Member
Registered:
Posts: 902
Reply with quote  #6 
Hi Danny here is the filter range and as far as I know the env# corresponds. (Page 2 right hand table)

Enveloping Settings Microlog Frequency Speed Analyzing Filters Band Range Range



http://prmeh.ru/pub/catalogues/skf/data/examples_of_application/pulp_and_paper_industry/056729_acceleration_in_paper_machines.pdf
Edwin

Avatar / Picture

Sr. Member
Registered:
Posts: 129
Reply with quote  #7 
Env1 is the lowest range. Env4 the highest.
__________________
= Cost Effective Asset Management =
RRS_Dave

Sr. Member / Supporter
Registered:
Posts: 206
Reply with quote  #8 
I'm having a hard time adjusting to the gE filters compared to the Spike Energy. We get what looks like a lot of noise with gE. Hard to look at it quickly and say go-no go.

D
Danny Harvey

Sr. Member / Moderator / Supporter
Registered:
Posts: 1,201
Reply with quote  #9 

I listened to a good presentation put on by one of SKF's best guys that may be of some help. I'll try to find it. They have lots of info on their site but it can be a little difficult to find.  I'm planning on more exploration but don't know if I'll have time soon.  
skfboiler

Member
Registered:
Posts: 3
Reply with quote  #10 
I posted this in response to your post, Danny in the SKF section of the Equipment forum.  I will repeat my response here.

Using the the 3rd filter, 30k-600k for example, signals below 30k will be filtered out and signals above 600k will be filtered out.  This 3rd filter is probably the most useful filter using the SKF version of this technology.  This filter will even work on slow speed equipment.  However, the other three filters should still be experimented with to determine which gives the better results.  Keep in mind if using a 500 mv/g sensor pay special attention the specs.  The roll off of this sensor may not be in the range of the 3rd band pass filter and therefore poor results will occur.  Therefore on slow speed machines a typical 100mv/g sensor will still be the preferred sensor to use.

I attached a document on how to use the SKF Enveloping technique.

pdf SKF Acceleration Enveloping and Other Bearing Defect Detection Methods.pdf     



stace1g

Sr. Member
Registered:
Posts: 100
Reply with quote  #11 
I find it useful to take a high frequency spectrum out to say 10kHz and look at where the 'hump' or energy is in this to decide where to set my envelope filters too. I must say though that I have experimented with the env filters included with Emonitor and it appears to make little difference a lot of the time which is used.  I tend to stick to 2.5kHz-5KHz for most stuff. With the emonitor online stuff the process is Gse for which I have found 5kHz LP to be the most useful.

Gary
Jorge Luis

Member
Registered:
Posts: 2
Reply with quote  #12 
Quote:
Originally Posted by skfboiler
I posted this in response to your post, Danny in the SKF section of the Equipment forum.  I will repeat my response here.

Using the the 3rd filter, 30k-600k for example, signals below 30k will be filtered out and signals above 600k will be filtered out.  This 3rd filter is probably the most useful filter using the SKF version of this technology.  This filter will even work on slow speed equipment.  However, the other three filters should still be experimented with to determine which gives the better results.  Keep in mind if using a 500 mv/g sensor pay special attention the specs.  The roll off of this sensor may not be in the range of the 3rd band pass filter and therefore poor results will occur.  Therefore on slow speed machines a typical 100mv/g sensor will still be the preferred sensor to use.

I attached a document on how to use the SKF Enveloping technique.

pdf SKF Acceleration Enveloping and Other Bearing Defect Detection Methods.pdf     


Previous Topic | Next Topic
Print
Reply

Quick Navigation:

Easily create a Forum Website with Website Toolbox.