• Author Author: Instructorman
  • Date Created: Date Created
  • Views 739 views
  • Likes 0 likes
  • Comments 2 comments
Related
Recommended

BenchVue Road Test: Lots of learning and first impression of the Function Generator Pro app

Instructorman

I have been working on the background steps necessary to perform the complete Road \test of BenchVue and the pro apps I was provided.

 

To recap I plan to use BenchVue and the pro apps for the  34461A34461A multimeter and the  33622A33622A function generator along with BenchVue Test Flow to automate testing of RGB LED strings I am using a PIC18F4520 microcontroller to generate the WS2811 waveforms that control the Red Green and Blue LEDs in each of 240 devices in a 4 m long string So far I have successfully data logged the end of string voltage by controlling the  34461A34461A using BenchVue That worked well I mentioned an unexpected waveshape for end of string voltage in a previous post Here are the details

 

The microcontroller was programmed to generate a data waveform that would produce a pattern of alternating red and green lights along the length of the string, as shown in the photograph below.

 

image

The pattern changes intensity over time.  As the green light intensity increases, the red light intensity decreases. The green light intensity reaches a maximum when the red light intensity reaches a minimum.  The sequence the repeats in reverse.  That is, green light intensity decreases as red light intensity increases.  The changing pattern is shown in the short video clip below.

 

The Keysight  34461A34461A was set up to measure the DC voltage at the end of the LED string BenchVue was set up to data log this voltage over a one minute duration with NPLC set to 1 Input voltage to the LED string was 6.000 VDC Maximum string current measured on another multimeter was 1.292 A with an average current of 1.216 A The resulting strip chart produced by BenchVue is shown below The blue flags indicate the position of user inserted markers Hovering a cursor over the flag within the application produces a pop out window with user provided text annotation You can also see two orange cursors labelled 1 and 2 These are user placed cursors Data about the sample number time stamp and data value at each cursor position,along with delta time and delta voltage are presented in the box at the bottom of the strip chart

 

image

 

The behavior I don't fully understand is the periodic sharp dips in voltage.  They occur under two conditions:  when the green LEDs are at maximum brightness and the red LEDs are at minimum brightness, and when the green LEDs are at minimum brightness and the red LEDs are at maximum brightness. Notice that in between these events the voltage is fairly flat, ignoring the noise caused by 1 NPLC sampling, with a slight slope up before a spike and a slight slope down following a spike.  Since end of string voltage is affected by current draw, it appears that current draw increases either right at maximum brightness, or just around maximum brightness.  But why this much?  The voltage scale on the left shows that the spikes are only about 100 mV deep, which is not much, but I don't know why they are there at all. I should data log current draw under the same conditions to see if there is a spike in current draw at maximum brightness.

 

There is no visual aberration caused by these spikes, so I'm not really concerned about them, just curious about what causes them.  I would not even know they occurred without the detail revealed by using BenchVue to examine the end of string voltage over time, as shown here.

 

Working toward using the  33622A33622A and its BenchVue Pro app

 

To test the LED strings I have programmed the PIC18F4520 to generate a series of fixed pattern illuminations.  Minimum and maximum brightness for red, green, and blue, as well as minimum and maximum brightness for white (R, G, and B LEDs on).  I then used my oscilloscope to capture the first 24 bits of the resulting data string.  Only 24 bits need to captured because 24 bits fully describe the light pattern for one device; 8 bits for green intensity, 8 bits for red intensity, and 8 bits for blue intensity.  The 24 bit pattern is repeated 240 times to illuminate the whole string.

 

The captured waveforms in CSV format are then put on a USB memory stick and moved to the  33622A33622A where they are loaded into arbitrary waveform memory The  33622A33622A is then programmed to burst the arb waveform 240 times This has worked perfectly so far My next step is to get the BenchVue test flow application to automate all of the steps for me to produce the data logs I'm interested in I'll be working on that over the next day or so and will post my findings here

 

Mark A

  • in reply to Workshopshed

    Good point Andy.  Could be a supply issue.  The supply I am using is capable of providing 6 A and that is where I had the current limit set.  The string was drawing an average 1.216 A with a maximum recorded current draw of 1.29A, so I was working on the assumption the supply could handle the current requirements, but I have not confirmed that independently. I have another brick supply that can provide 10A and I can easily test it to see what happens to the dips.

     

    Or, maybe I should write maximum brightness red and/or green to all LEDs while monitoring the current and see if that causes a 100 mV dip in the supply voltage

     

    Mark A

     

    Um, I notice that many periods are missing from the ends of my sentences in the blog post.  I do use periods at the end of sentences, but something has gone through and deleted some of them.  Minions on the loose?

  • Voltage dip at maximum current suggests a supply issue?