sags & swells

Status
Not open for further replies.
I noticed that your post #55 is titled "Record 1". The only way I get that title to appear is if I download to the laptop after the recording is done. I then have three values of Y for any X cursor position.

A title of "Vrms(ac+dc)" is only optained while logging simultaneously with the laptop. I then have one value of Y for any X cursor position.
 
080818-0751 EST

tom:

If you log while collecting data is the data quantized to 1 second or something finer, like 1/60 second?

If 1 second is the sample rate, then if you look at the data file instead of the graphical data are there 3 values for each 1 second sample or just 1 value?

.
 
ptonsparky said:
I noticed that your post #55 is titled "Record 1". The only way I get that title to appear is if I download to the laptop after the recording is done. I then have three values of Y for any X cursor position.

A title of "Vrms(ac+dc)" is only optained while logging simultaneously with the laptop. I then have one value of Y for any X cursor position.
Click to expand...
Your correct, I must have saved the wrong capture. So it has nothing to do with the minimum being zero. The log to PC function just shows the average and not min/max.
 
gar said:
080818-0751 EST

tom:

If you log while collecting data is the data quantized to 1 second or something finer, like 1/60 second?

If 1 second is the sample rate, then if you look at the data file instead of the graphical data are there 3 values for each 1 second sample or just 1 value?

.
Click to expand...
gar:

The copied data from a PC capture only has one value but the copied data from the download from the 43B, it has three values.
 
080818-0911 EST

bill:

What is the difference between "capture" and "download"?

I assume that "capture" means data is continuously acquired at the PC while the 43B is measuring the input. What is the time increment from one captured sample to the next? Is this controllable from the PC or definable at the 43B?

I assume that "download" is the transfer of the current 43B screen display to the PC. In this mode I would certainly expect the PC display to be identical to the 43B display.

.
 
080818-1017 EST

Tom:

Based upon my earlier analysis from reading the 43B manual the 1 second sample time would appear to to agree with with I would expect.

For each 1 second sample period I believe there are three resultant values created by the 43B. I will repeat --- Every 1/60 second the RMS value of the input is calculated. The first sample of a 1 second period is stored in both the min and max registers. In other words these two registers are reinitialized at the beginning of each 1 second period to the input value at the start of the period. Also the averaging register is loaded with this first value. Following this until the end of the 1 second period the min and max registers are updated as needed with a lower or higher value respectively. The average register simply has the new 1/60 second value added to the average register.

At the end of the 1 second period the average register is divided by 60 to get the average value, and the values in the min and max registers are the min and max values of the last 1 second period.

These values at the end of the 1 second period are plotted on the 43B screen so long as we are less than the first 240 seconds.

I would expect that these three values for each 1 second period would also be available for transfer to the PC. Even at 1200 baud (about 120 characters per second) it should be possible to transafer all the data for the three values.

Thus, my question of whether the raw data from the 43B to the PC in capture mode may contain all this data. If it does not. then does a higher baud rate change anything or are their some parameters that might define whether all 3 data values are transferred or just the average value?

.
 
Last edited:
The Fluke View for Scopemeters software looks just like the 43B PQA software.

On the 43B where you set the Vrms or Arms, a Fluke 123 Scopemeter has the option of min/max/avg also. It gives you three seperate plots with different colors on the PC and you can also plot a RMS reading at the same time. The cursor by default on avg. You can select the active waveform from a menu. It's greyed out on the 43B!:rolleyes:

The Fluke 123 does this better than a 199B which doesn't even support TrendPlot at all while logging from a PC!

Here's what that looks like from the Fluke 123.

Plots.jpg
 
gar said:
080818-1252 EST

tom and bill:

I got some information, but I do not have time until later today. Then it will be by PM.

.
Click to expand...


Well please, go ahead and shoot Gar. I am lost on your questioning of sampling rates. I just have the 41PQA, and you have me interested. I'd like to start the question over and go from there. If you were not satisfied from the answer you received from Fluke, than I shall try to provide you with one or direct you in the right direction.

wptski has the 43B, he may answer your question before I. It is late, and I am not into re-reading the whole thread. Please repost.

I understand your frustration, but please, re-itterate.
 
gar said:
080818-1252 EST

tom and bill:

I got some information, but I do not have time until later today. Then it will be by PM.

.
Click to expand...
I passed your PM on to 76nemo because I knew that he'd be interested in Fluke not wanting to answer questions about their products. Not sure if you just talked to a Tech Support or beyond? It may be a case of you knowing more than them. In that case, I'd say that you didn't really talk to a person in the PQ department.
 
76nemo said:
Well please, go ahead and shoot Gar. I am lost on your questioning of sampling rates. I just have the 41PQA, and you have me interested. I'd like to start the question over and go from there. If you were not satisfied from the answer you received from Fluke, than I shall try to provide you with one or direct you in the right direction.

wptski has the 43B, he may answer your question before I. It is late, and I am not into re-reading the whole thread. Please repost.

I understand your frustration, but please, re-itterate.
Click to expand...
76nemo:

I've been trying to help the OP of this thread but understanding how the 43B really works is beyond me. Since this thread started is the first time I ever logged from my PC.

It appears that only the 43B only logs average readings but also the 190 Series Scopemeters too. The 120 Series seems to out shine the 190 Series when it comes logging from a PC!
 
080818-2107 EST

76nemo:

Do you work at Sterling also?

To start over. What I think I understand about the 43B in the sag-swell mode is:

1. At a high sampling rate, maybe about 16/256 MS, the input signal is sampled and from these samples the RMS value for for one full cycle at 60 Hz is calculated. Whether it is 256 samples or 512 in one cycle is not important or whether the RMS value is calculated for 1/2 cycle or a full cycle again may not be important.

2. Important is that we get an RMS measurement every 1/60 of a second.

3. Next important point is that at the end of a 1 second collection period that we get 3 measurements from the input signal. These are used for the internal display. Next everything is reset and we start a new 1 second collection period from another set of 60 (maybe 120) measurements.

None of these 60 measurements are available for display internally or externally. Only the summary values min, max, and average are available.

4. In sag-swell mode the only available data values are the results of each 1 second period. These values are the min. max, and average of the 60 measurements of the respective 1 second period. Also collected is a time stamp.

5. Every 1 second period I would expect the 43B to be able to send these 3 values plus the real time to a collection computer. If 1200 baud is too slow for the way this data is formatted to transfer all this data, then certainly at the maximum baud rate of 19.2 kbaud the transfer can be performed. But I could certainly send that amount of data at 1200 baud. 1200 baud is about 120 characters per second.

6. For internal data there is compression after the first 240 seconds, 4 minutes, unless the buffer is cleared and the initial 4 minute period is restarted.

Fluke claims there is too much data to send min, max, and average every 1 second and therefore only sends the average value for continuous monitoring..

The too much data argument I do not buy.

.
 
gar:

TrendPlot in Fluke's Scopemeters is similiar to the 43B's S/S. As you can see above somewhere that a 123 at 19.2 has no problem suppling min/max/avg in TrendPlot.

The 199B doesn't even support PC logging in TrendPlot only in normal scope function. It only supoorts average readings and that at 38.4!:rolleyes:
 
gar said:
080818-2107 EST

76nemo:

Do you work at Sterling also?

To start over. What I think I understand about the 43B in the sag-swell mode is:

1. At a high sampling rate, maybe about 16/256 MS, the input signal is sampled and from these samples the RMS value for for one full cycle at 60 Hz is calculated. Whether it is 256 samples or 512 in one cycle is not important or whether the RMS value is calculated for 1/2 cycle or a full cycle again may not be important.

2. Important is that we get an RMS measurement every 1/60 of a second.

3. Next important point is that at the end of a 1 second collection period that we get 3 measurements from the input signal. These are used for the internal display. Next everything is reset and we start a new 1 second collection period from another set of 60 (maybe 120) measurements.

None of these 60 measurements are available for display internally or externally. Only the summary values min, max, and average are available.

4. In sag-swell mode the only available data values are the results of each 1 second period. These values are the min. max, and average of the 60 measurements of the respective 1 second period. Also collected is a time stamp.

5. Every 1 second period I would expect the 43B to be able to send these 3 values plus the real time to a collection computer. If 1200 baud is too slow for the way this data is formatted to transfer all this data, then certainly at the maximum baud rate of 19.2 kbaud the transfer can be performed. But I could certainly send that amount of data at 1200 baud. 1200 baud is about 120 characters per second.

6. For internal data there is compression after the first 240 seconds, 4 minutes, unless the buffer is cleared and the initial 4 minute period is restarted.

Fluke claims there is too much data to send min, max, and average every 1 second and therefore only sends the average value for continuous monitoring..

The too much data argument I do not buy.

.
Click to expand...


After running this by our, (not FLUKE), engineers, I am more confused:confused: I don't know who or where you say the sampling is only corrected on a 1 second interval. I don't have the 43B or a PQA that logs to a PC. I have to agree with you GAR. This doesn't sit right. I don't know about the specific sampling rates in real time using a PC, but I was baffled that my/the 123 wouldn't support real time to a PC? When it gets down to baud rates and logging, it is well over my head. If you or anyone else is that determined to land this one, I will make it a point to get to thee bottom of it. I will have to have it explained to me more than once to understand it:roll:

I would just like to add that you must have been put through to someone who wasn't in the design issue of the 43B, and I hope you are not discouraged as you seem to be.

As far as logging to a PC, it's well over my head, yet I will try to solve this with some help from them.
 
080819-2002 EST

76nemo:

I suspect that per data value to be transmitted that no more than 12 bits are required. 12 bits can be transmitted in 2 8 bit words (equivalent to 2 ASCII characters). If the value was not in binary but in ASCII text characters, then 12 bits plus sign (+/-4096) can be transmitted in 4 or 5 characters. Time representated in ASCII characters could be 080819-2009.45.32 (representing year, month, day, time to 1/100 second), which is very verbose and all of it is not required for each sample point. But if all this was transmitted, then we have 15 characters for data and 17 characters for date and time, or 32 characters.

Thus, I could send all the data I need per 1 second sample, in a verbose mode, in 32/120 or 27% of the available time at 1200 baud. Using a better format for data transfer I might need maybe 10 to 12 characters worth of communication space instead of 32.

The only available data in the sag-swell mode for continuous monitoring is 3 values (min, max, and average) of the last 1 second period plus the time of occurance of the 1 second period. But as I understand the operation from the various comments and discussion only the average value and time are sent to the computer.

Background on baud rate. Baud rate is the number of clock cycles per second used to transmit serial binary (bits) data. In asynchronous communication there is always 1 start bit, followed by X data bits, next a parity bit if parity is being used, and one or more stop bits. If no parity is used, then to send 8 data bits requires at least 10 bits of baud rate time.

Baud really implies rate but I and many others append rate because it flows nicer.

The full ASCII alphabet plus an extended set is contained within 8 bits. Thus, if my baud rate is 1200 bits/second, I can send alpha-numeric characters at the rate of 120 per second. If I go to 19.2 kbaud, then the character rate is 1920 per second. For CNC communication I regularly send at a rate of 115.2 kbaud or 11,520 characters per second.

I am not particularly unhappy with Fluke other than the response I got from the PQ person on the phone. If your customers can have a full understanding of how an instrument works and why certain limitations exist, then their use of that instrument can be more effective.

I have a greater AC bandwith capability on the high end with my Simpson 270 and with several HP meters, compared to your 27 and 87 models. The HP instruments are quite different and do not represent a fair comparison. From memory the 27 and 87 start to give bad data at maybe 40 kHz, while the Simpson goes to about 200 kHz and then simply rolls off. I am not complaining but using this to illustrate that one needs to know and understand the operation of an instrument.

All of the discussion on this thread resulted because the original question from post #1 was relative to a disparity between what was seen on the 43B screen and what was collected on Tom's laptop.

.
 
Status
Not open for further replies.
Top