Help make a ripple meter

  • Dec 09, 2021

In the Lamptest project, I really miss a simple pulse meter that would show the percentage of light pulsation continuously. I think such a device would be useful to many others.

Help make a ripple meter

Now I measure the ripple coefficient with the Uprtek MK350D spectrometer, but for each measurement you need to press a button on the device or in the program. It would be great if there was an additional device that shows the ripple level constantly, and at the same time it would be nice if it showed the ripple shape on its screen.

For some reason, the Chinese do not make cheap pulse meters, they are not on Aliexpress. In Russia, they make TKA-PKM 09 for 36,000 rubles, eLight02 for 28,900 and Lupine for 7,000 rubles. I have Lupine, but he is not very convenient because the results of his measurements are influenced by electromagnetic interference.

There is nothing difficult in measuring the ripple: it is enough to make several dozen brightness measurements in a short time and calculate the ripple factor using one of two formulas, using the minimum and maximum value obtained illumination.

instagram viewer

There are two formulas for calculating the ripple coefficient - simple and complex. Here is a simple formula (this is how the Uprtek MK350D ripple counts).

Here's the tricky one (Lupine thinks so).

The maximum ripple factor according to a simple formula is 100% (with such a ripple, the light source is completely extinguished 100 times per second), according to the complex formula, the KP can be more than 100% if the light not only completely extinguishes, but the light time is less than the time darkness.

I have made several attempts to make a ripple meter on Arduino, but so far nothing good has come out.

First problem: light sensor.

It is most convenient to use digital sensors with the Arduino, such as the TLS2561 or BH1750, but they are not suitable for measuring ripple, since they are too slow. At best, these sensors are capable of making 10 measurements per second, and we need to make at least 400, and preferably 3000 (to get a nice graph).

I think about digital sensors, you can forget - it must be something that connects to the analog input Arduino, because even in slow mode, you can take about 8000 measurements in give me a sec.

I tried to use the TSL257-LF sensor, but it turned out that it is too highly sensitive - at 8 lux illumination it "rolls over" (surprisingly, there is no word about lux in the datasheet). Here is something that shows only if you cover it with your hand.

I tried to wrap the sensor with a thick layer of white electrical tape, but this, of course, is not serious.

I also have OSRAM SFH5711-2 / 3 sensors, which have a range of 3-80000 lux in the datasheet, but they transmit the illumination with a varying current to output (5-50 μA) at a voltage of 0.5 V and I don't know how to pair this with the Arduino, besides, they turned out to be tiny - I can hardly solder to them wires.

I think we should try the sensor based on the TEMT6000 phototransistor (here it is on Ali).

Cheap and cheerful, although the range is only 10-1000 lux (1000 lux is approximately 40 cm from a 1000 lm lamp), but nothing - you can also make a light filter to reduce the sensitivity.

Second problem: programming. I can do some simple things on the Arduino and probably can even read data into an array, and then display the graph on TFT-screen, but for the correct calculation of CP, you need to filter and average the minimum and maximum levels, and with this I definitely do not I can handle it.

I imagine a ripple meter like this:

Hardware: Arduino, cheap TFT screen 1.44 ", analog light sensor. Hopefully, we can do without an external ADC.

Program:
- by pressing the button, we memorize the level of darkness (in fact, the level of the ambient light);
- we capture 160 brightness values ​​in the buffer in ~ 0.05 seconds (~ 2.5 periods of the 50 Hz network, one measurement every 0.3 ms);
- display two periods in the form of a graph in the lower part of the screen (128 pixels wide);
- we calculate the ripple coefficient using two formulas, display two values ​​on the screen.

I know that a lot of electronics and programming specialists read me. Maybe you can take on the creation of such a device for public benefit? If everything works out, I’m in both hands if the sources are published on github and I’ll write a detailed one myself. instructions for assembling such a device so that everyone can quickly, easily and cheaply make a meter for themselves ripple.

© 2021, Alexey Nadezhin

For ten years I have been writing every day about technology, discounts, places of interest and events. Read my blog on the site ammo1.ru, v LJ, Zen, Mirtesen, Telegram.
My projects:
Lamptest.ru. I test LED lamps and help figure out which ones are good and which are not so good.
Elerus.ru. I collect information about domestic electronic devices for personal use and share it.
You can contact me in Telegram
@ ammo1 and by mail [email protected].