sags & swells

Status
Not open for further replies.
080808-0834 EST

I have found little additional information in the on-line manuals.

In a software manual it seems to imply that the maximum baud rate is 19.2 kbaud, and that 8 data bits and 1 stop bit are mandatory. Apparently Xon-Xoff is optional, and there is no option for hardware handshake. Where I found this information no mention was made of parity, but elsewhere it appears parity must be none.

Post #1 presents a problem with a disparity of readings between the 43B and the data collected on a laptop from the 43B. To try to understand the problem it is necessary to have precise knowledge of how the 43B makes measurements and how these are communicated to the laptop.

The manual does not give an accurate unambiguous description of what happens in the sag-swell measurement. In particular there is no information on what is transferred from the 43B to the RS232 port. I believe we have data every 1 second of the min, max and average of the one cycle RMS values. Is the 1 second data always output to RS232 independent of any other settings of the 43B? If so, then on the computer you should be able to look at all data down to the 1 second sample independent of how much data has been collected up to some maximum. One day is 86,400 seconds and if 100 bytes are required for the 1 second data, then about 8 megabytes are required per day. I think 100 bytes per sample point is quite high. Any modern computer will have no problem storing a number of days of this data.

The data and the readings on the 43B will change as the data collection time increases because of compression. This is because of limited data storage capacity on the 43B. Compression takes the most min and most max of the samples of the interval being compressed and these become the new min and max values and the average of the averages is calculated to become the new average. The fine time increment data is lost at the 43B.

I believe the numeric values displayed on the 43B will be the values of the sample at the cursor location. If the screen width is greater than 4 minutes, then the numeric values at a given real time may be different than the 1 second sample value at that same real time point. A real time sample point after compression is a composite of a number of 1 second sample points used to generate the compressed value. Does this have anything to do with the problem of post #1?

From post #1:
Meter display caught a 2.155 volt spike. Laptop displays a max of 1.096 volt. What gives?
Click to expand...
Was the 43B in sag-swell mode? What was the screen width in time? Was 2.155 the maximum reading, or a peak meaning something else? Or was the meter in transient mode? I do not think that transient and sag-swell can run simultaneously.

ptonsparky:

What additional information can you provide on the settings of the 43B and the computer when these conflicting results were obtained?

.
 
I looked for the spike on the laptop corresponding to the time of the spike on the 43. Both had the same time and both were the max value of the 4 minute recording. I was in S/S. You are correct, the 43 will not capture transients at the same time as S/S.

The laptop was set to record continuously but it stops recording whenever the 43 does.
 
080808-1938 EST

ptonsparky:

I believe what you have said is that you recorded for one 4 minute period. On the 43B screen at one time point you had a swell (maximum reading) of 2.155 V, and at the same time point on the computer display of this data the value was 1.096 V. On both displays there should have been an average voltage reading, and also a sag (minimum) at some other time. Hopefully the time of the minimum should be the same on both the 43B and the computer. The average reading applies to the entire 4 minute period, and it should be the same value for both the 43B and the computer.

What are the values of the average readings? What are the minimum readings and do their times agree?

Besides collecting data from the 43B as it is generated you should be able to do a screen dump from the 43B of the screen in question. Does this dump agree with what you saw on the 43B screen?

In a manual there is a section on Recording Harmonics over Time. This might have some bearing on how recording over time of sag-swell is done. This has an interval between updates parameter. This might cause a difference in the resulting data.

Knowing the values of average voltage, the minimum voltages, and the minimum time correlation may be the next step to pinpointing the problem.

.
 
Since I have a Fluke 43B I was interested in this thread but it was fading away.

I setup a four minute S/S measurement on a branch circuit reading N-G voltage. This circuit has my furnace blower motor on it, so I turned it ON/OFF for a second or so. I can't duplicate what the original OP doing but it's the closest I come come to it.

The 43B shows a voltage drop of 1.869V and the realtime plot in the software showed 1.846V. Using the OC4USB cord on a old PC still using USB v1.1 and I "think" it connects at 19,200 baudrate. This actually the first time that I've monitored the 43B while connected to my PC. I normally take a reading and connect later to my PC if I want to save it.

So, I have idea what the OP problem is!
 
I tried this again at home but don't have the fluctuations here compared to the shop. If you get a chance to look at the 43 display you will see the darker part of the tracking. That shows up on the laptop. Every once in a while you will see a faint spike that peaks higher. These faint peaks are recorded on the 43 but not transfered to the laptop.

I tried a graphic link

19.2 is all mine will connect at.
 
From your first post: Meter display caught a 2.155 volt spike. Laptop displays a max of 1.096 volt. What gives?

Sounds like you are refering to something different now. Your original post questions a difference of about 1V between the meter and your PC, I don't see that at all.

What exactly are your measuring where you see something on the 43B and not on your PC? Generally, what you see on the PC has more detail.
 
080813-1941 EST

What you want is a very controlled experiment. Even without an elaborate arrangement you could do the following:

Connect a 1000 ohm (1/4 W is fine) nothing critical about this resistor across the 43B input. 100 ohms would be ok and even 10,000 ohm would work. The lower resistance will result in lower ambient noise.

Use a small pushbutton or toggle switch, electronic type, gold contacts preferred, but plain silver, if clean are OK, between the + lead of the battery and the probe, and a wire from the - side of the battery to the 43B common.

This will be a stable two state signal, 0 V or about 1.55 V. You can generate less than 1 second pulses and longer ones to simulate different inputs to the 43B and in turn the laptop.

This way you can expedite your experiments and be dealing with a known signal. It is important to get the min, max and average readings from both the 43B and the laptop.

.
 
I need one of those flashing light bulb avitars. Put it ahead of my...

HS, now I understand!!

I've got my head stuck on Sags & Swells and you want me to do a voltage recording. Duh, two different functions. We can do that. Apprentice needs some experience with cycle timers anyway.
 
080813-2047 EST

ptonsparky:

I do not understand the plots. Are both the top and middle from the laptop? And how in time does the bottom correlate with the other two?

The top plot appears to have the min and max values defined by the grayed area and superimposed on this is the average value per whatever quantized time period.

On the left part of the middle plot I can see the general pattern of the average plot from the top plot. The time scale appears to be different between the top and middle and is not synchronized at the left.

I do not know where the bottom plot fits into the top and middle.

On the right side of the top and middle plots I see more disparity in the curves.

I suggest doing experiments with just the 240 second capture period. This would be the best way to compare the different results.

.
 
The top is a "trend" that the laptop can receive from the 43 after the recording is done.

The screen shot at the bottom is the 43s version of the laptop recording(middle screen). When I "recall" this screen shot on the 43, I can scroll left/right with the cursor to view the min/max values and the time.

The upper & middle screen are very close to same time period. The vertical dashed lines under the 2 are at about the same time. The 2 corresponds to apout the right hand third of the lower screen shot. I forgot to move the cursor before capturing that screen.

Data from the trend and laptop capture can be exported to a spreadsheet for those that really want a headache.
 
080813-2138 EST

ptonsparky:

I do think you want sags and swells mode.

In an earlier post I described what I think Fluke is doing. Every full cycle, 1/60 second, the RMS value of the input is calculated. During every 1 second period the min and max values are saved, and the average of the 60 measurements is calculated. This provides you 3 data values for each 1 second period. I believe that is what is in the top plot. I also think that is what the screen plot should be.

For example: If I have 59 values at 1.5 V and one at 1.0 V, then for 1 second the result is max = 1.5 V, min = 1.0 V, average of the 1 second is 1.4917 V.

If you do not go beyond a 240 second initial period, then the screen shot should be the same as the laptop.

.
 
It just came to me that your using a 43 and not a 43B! I'm wondering if your using Fluke View PQA v3.3? It works with the 43 also and can be downloaded here.

My 43B mirrors what's shown on my PC down to .1V variations.
 
The top graphic includes all the information of the bottom screen shot inside the shaded blue area. I compared the the two by dragging the cursors accross the screen both on the laptop and the 43B. The three points given are equal. 6, 13, 81 mv for the point I examined

I have to go to the feature itself to compare this screen (middle) with the other two. There is a time stamp difference of 8 seconds for each plot because the the laptop does not start recording the same time as the 43B. It must be manually started. I get one value of 17 mv. Don't know how that value is calculated.

May have to examine the headache causing data.
 
Last edited:
Status
Not open for further replies.
Top