Creating a clean CDG relay curve with subdivisions 1

[DiscoP]

I am not sure if anyone can help, but I will give it a try.

I help run training courses on protection relays - we have a section where we get the students perform a simple settings study using a CDG11 curve.

The graph we use just shows the operating times when the time multiplier is set to 1, but the graph is an enlarged photocopy of a page from the old GEC PRAG, and is difficult to read.

Alstom/GEC have graphs in their publications, however they show graphs for time multiplier settings from 0.1 to 1, as opposed to just the one curve at time multiplier equal to 1

We used this version of the curve in our most recent course, and from this experience I think that having a curve with just the time multiplier of 1 makes initial understanding of the time multiplier concept easier to follow.

So - I have tried asking Alstom for a nice clean version of the graph (as shown in their old version of PRAG) - but they are unable to assist.

The graph is still shown in their current version (NPAG), but is small and doesn't look that great when it is enlarged (and lacks some subdivisions)

I can also plot the CDG curve on an excel log-log chart, but (as far as I can work out) the limit for subdivisions is 1 between 1 and 10 on the log scale, so it leads to a lot of guess work. It would be nice to include 0.1 subdivisions in this part of the graph, but I cannot work out how to do this.

So I have three options I can think of

1. See if anyone else has exactly what I am after
2. See if someone knows how to show more subdivisions when using log-log graphs in excel
3. See if anyone knows another program I can use to perform a log-log plot of this curve (with more subdivisions)

I understand I am being fussy with this request - and not many people actually use these curves to perform grading studies (including myself), but I think it is a great first step for people learning about this.

Is there anyone who can give me some tips on any of these 3 options.

 

[Hoxton]

From memory,the curve is now defined as a formula in IEC62055 and BS142 (I think BS142 has been superseded)

So if you google "IDMT curve formula" you will find references. You should be able to calculate basic settings to draw a graph

 

[dpc]

EasyPower has CDG11 (GEC) relay curves. But there are options for the curve shape (Normal, Short, Ext Inv, ANSI). If you let me know what you want, I can make a TCC for you and send you a .pdf. (I know nothing about this particular relay, so assume you are explaining this to a reasonably intelligent border collie)

Dave

 

[ScottyUK]

I might have an old original CDG manual at work. If I can find it and it's in reasonable condition I'll scan it and post as a pdf and/or jpeg. I expect a few minutes with GIMP or Photoshop would tidy up any atrefacts from the scan.

I think the GEC curves will be designated SI, VI, EI and LTE or standard inverse, very inverse, extremely inverse, and long term earth, unless they made also relays with ANSI curves for the North American market. In that case, which set of curves do you want?

 

[DiscoP]

Thanks Scotty and dpc it is the 3 second standard inverse curve for a cdg11.

It is a course for new protection technicians. For many it is the first and only time they will ever do a settings study, so I try to remove as many unnecessary complications as possible.

Hoxton, the formula is close but not the same as the CDG curve. My other option is to update the course to use a more modern relay that follows the formula, but the old relays are good to start with as you can see things happen.

 

[ScottyUK]

DiscoP,

Try this link:

http://www.electricalmanuals.net/pdfs/Protective Relays/Areva-Alstom-GEC-Midos/CDG/R4066.PDF

I know it's the VI curve but does it serve the purpose?

The website itself is worth registering on: it's free and it doesn't send you loads of spam and crap.

 

[dpc]

Let me know if this look like the right curve.

 

[DiscoP]

Hi ScottyUK and DPC

Thanks for your help.

Using google I eventually managed to find the same picture we had been using (see link) - it is perfect for what we want, but a low resolution.

http://sayedsaad.com/Protection/files/Over_current/9_Overcurrent_files/image004.jpg

We had changed in the recent course to the same drawing scottyuk has linked (but the SI version), however having all the time dials shown was confusing some people.

I was also able to produce a drawing similar to the one linked by dpc using excel, but these drawings don't have enough graph lines on them.

We are trying to get the students to accurately determine either an operating time for a given time dial setting, or determine an appropriate time dial setting for a given time delay.

So all the minor subdivisions help. Most of the drawings you get from grading software don't have enough graph lines to use them for this purpose as there is too much guess work.

If anyone has a higher quality version of the drawing I have linked, I would be thankful if you could post a copy.

Likewise, if anyone knows how to get the minor line divisions on excel to match the lines shown in the linked drawing, please let me know - or if anyone knows some other plotting software that may be able to reproduce this graph.

I know I am being overly fussy - I think I have been in protection for way too long

 

[bacon4life]

You can add arbitrary gridlines to excel as shown at

http://peltiertech.com/Excel/Charts/ArbitraryGridlines.html

Here is a log-log_graph.jpg from excel with the minor subdivisions. I just put a couple of points off of CDG11 curve into log_scales.xlsx .

http://files.engineering.com/getfil...4bf0-b0cf-7e2ac25f7202&file=log-log_graph.jpg

http://files.engineering.com/getfil...4-4021-9ede-15edd6b36f72&file=log_scales.xlsx

 

[DiscoP]

Thanks bacon4life - this is a great solution.

 

[DiscoP]

Thanks bacon4life - this is a great solution.