Author Topic: Easy & Accurate way to Measure Flickering Frequency of a light source  (Read 6453 times)

hgg

  • Newbie
  • *
  • Posts: 5
Hello to all the forum members!   :)

My name is George and I am a new member.  I accidentally discovered this forum while
searching for reviews in order to buy a new multimeter.  Luckily I found Martin and his
excellent YouTube reviews!

I decided to buy the UNI-T UT61E multimeter and I think that I took the right decision.
For its price range I don't think that there is a better multimeter.  The major negative
is the omission of the already supported (!) backlight.

Anyway, I would like to ask your opinion on something.
Back in the days of CRT's the refresh rate of the monitors was quite important for me because
flickering is affecting me and makes me dizzy.  I don't like flickering...   :)  All the CRT monitors
I bought had more than 100Hz refresh rate which made it more comfortable than the 60, 75Hz etc.

Now we have the new LCDs of course and nobody mentions monitor frequency any more.
Indeed, I can work more relaxed on LCD's but still there was something "annoying" about them.
I swear that I could detect very high frequencies that made me feel uncomfortable after some time
working on them, despite the fact that everybody said that there is no flickering.

So, I decided to check it.  I bought a multimeter, and together with a solar panel I had from a
broken car battery charger I made an optical frequency meter.

To my surprise, when I measured the monitor I am typing right now, I got 157 Hz !
I immediately liked that meter I just made!  :-)  I started measuring everything.  My SONY laptop
that was very pleasing to my eyes had a 210Hz refresh rate despite being 10 years old! 
Everybody that worked on that laptop said that the image was like printed paper.
Some led lighting I've installed and did not like, had an 100Hz flickering, which is exactly double
the mains frequency and its probable a result of bad rectification.  I also measured successfully
a UV light source.

So despite the fact that LED's do not actually flicker, their PWM implementation destroys an
excellent light source!  In the case of the older LCD panels that are using CCFL for backlight
the LCD flickering comes from the actual lighting bulbs.

I have also made a YouTube video that demonstrates the measurements.
http://www.youtube.com/watch?v=iu_44hP9mQc


So, my questions are:

1) Does anybody know if they are selling any measuring devices like the one I've just made?
    I could not find one, and I think its very useful and easy to make.  Actually I've already ordered
    the tiny UNI-T UT120T multimeter and I am searching for a small solar panel in order to combine
    them in a single portable measuring device.  I can even replace the multimeter battery with a
    high capacity rechargeable one that will be charged from the solar panel.

2) Does anybody know a way to find out the frequency range that a solar panel can measure?
    Specifically the upper limit.  Do they have an upper limit?

3) The UT61E has a (hz) mark on Volts & Amps measurement.  It also has a dedicated (hz) switch
     position!    The frequency measurement works only on the (hz) switch position. 
     Do you know when to use each one?
     My guess is that the (hz) on the Volts & Amps works only when measuring an AC source and
     the (hz) dial position works only when measuring DC ripple voltages.  Am I correct?

4) Is there any better way to do the same thing?  Maybe the BPW34 photo-diode instead of the
    solar panel?

Sorry for my the lengthy first post but I wanted to share something that I thought was very
interesting.

Thank you.
George.

« Last Edit: April 18, 2013, 02:00:17 PM by hgg »

ProBang

  • Newbie
  • *
  • Posts: 8
Re: Easy & Accurate way to Measure Flickering Frequency of a light source
« Reply #1 on: April 18, 2013, 05:17:19 AM »
Hello

[...]

Anyway, I would like to ask your opinion on something.
Back in the days of CRT's the refresh rate of the monitors was quite important for me because
flickering is affecting me and makes me dizzy.  I don't like flickering...   :)  All the CRT monitors
I bought had more than 100Hz refresh rate which made it more comfortable than the 60, 75Hz etc.

At CRTīs is the refresh-rate equal to the light-emitting-rate. With CCFL/LCDīs it is a little different.

Quote
Now we have the new LCDs of course and nobody mentions monitor frequency any more.

Nope. You can buy  LED-TVīs with 100, 200... 400 Hz refresh-rate of the display without any problem.
But there is a big difference to the CRTīs: The light-emitting-factor ist the PWM-Frequency of the Backlight,
especially the Duty-Cycle,  not the display-refresh! (Of course: Except, thatīs a higher refresh-rate need a higher PWM-rate.)

Quote
Indeed, I can work more relaxed on LCD's but still there was something "annoying" about them.
I swear that I could detect very high frequencies that made me feel uncomfortable after some time
working on them, despite the fact that everybody said that there is no flickering.

They are wrong. Itīs a well known problem with some LED-Backlights.

Quote
So, I decided to check it.  I bought a multimeter, and together with a solar panel I had from a
broken car battery charger I made an optical frequency meter.

Cool idea!!!

Quote
To my surprise, when I measured the monitor I am typing right now, I got 157 Hz !
I immediately liked that meter I just made!  :-)  I started measuring everything.  My SONY laptop
that was very pleasing to my eyes had a 210Hz refresh rate despite being 10 years old! 
Everybody that worked on that laptop said that the image was like printed paper.
Some led lighting I've installed and did not like, had an 100Hz flickering, which is exactly double
the mains frequency and its probable a result of bad rectification.  I also measured successfully
a UV light source.

157 Hz are for sure much to low! It seemīs, the laptop had in his origin design a CCFL-Backlight.
Why? Letīs look, how the displayīs work...
First: The LCD, which is showing the picture, is with CCFL and LED Backlight the same.
As long there is no changing in the picture to show (reading a text... ), as long the display looks steady.
(Because it switched at every refresh the same pixels on and hold them until the next refresh. There is
no flickering at all.)
If the picture is changing (scrolling the text... ), the display is actualizing at the next refresh. 
Standard at laptops:
The display refreshes 60 times per second ( =60 Hz), regardless there is anything to actualize or not.
Thatīs the refresh-rate.

Now we add a backlight. First the CCFL.
The Brightness of the display is controlled with PWM. The frequency of the PWM-Signal is always the same.
Important is the Duty-Cycle. At full Brightness you have a duty-cycle-ratio of 100:0 (say 100%) on-time.
The Backlight is lighting up all the time.
At half brightness you have a ratio of 50:50 (say 50%). At half time the backlight is full lighting,
in the other half of the time it is darker (not black!), because a CCFL has some after-glowing when it is switched off.
The Contrast between on and off isnīt very hard. That avoids a big part of flickering.

Now we change to a LED-Backlight.
The Brightness is also controlled with a PWM. It works the same way. But there are two differences:
A LED-Backlight is brighter as a CCFL. To produce the same result in absolute brightness as the
CCFL with 50% duty-cycle a LED needs (for example) only 33%.
Much more important: If the Backlight switched off, the LED`s are immediatly black! There is no after-glow.
Thatīs making a very hard contrast.
In the first case (CCFL) you have a backlight, which is half the time full bright and half the time dimmed.
In the second case (LED) you have a backlight, which is at 1/3 of the time at full brightness
and at 2/3 of the time absolute black.
The On/Off-Cycle of the backlight is building the Backlight-Rate.
A CCFL-Backlight work fine with (at least) 150 Hz.
To avoid flickering with a LED-Backlight is a rate of (at least) 250 Hz needed.
Unfortunatly are the manufacturers not giving informations about the backlight-rate.
It seems, your little device should be very useful when you buy your next laptop...

Today it seems to be the only way to make the display a little less flickering, to increase the brightness-setting
(except it is at the max). Perhaps it helps a little...
You can, just for fun, fiddle with this setting and look what the 61E is showing at the value of the duty-cycle (Function "%").

Quote
1) Does anybody know if they are selling any measuring devices like the one I've just made?
    I could not find one, and I think its very useful and easy to make.  Actually I've already ordered
    the tiny UNI-T UT120T multimeter and I am searching for a small solar panel in order to combine
    them in a single portable measuring device.  I can even replace the multimeter battery with a
    high capacity rechargeable one that will be charged from the solar panel.

Iīve never seen. Probably there is no big market and because it is easy to build yourself.
(I have a deja-vų - a DMM, powered from rechargable cells which feeded from solar-panels...
possible that Iīve seen that anywhere before?)

Quote
2) Does anybody know a way to find out the frequency range that a solar panel can measure?
    Specifically the upper limit.  Do they have an upper limit?

It belongs to the frequency do you want to measure.
Probably are solar-panels not the first choice. Because they are optimized for producing current,
not to follow quick changes in lighting.
Perhaps it is interesting to measure their rise-time. But I canīt do it myself. Iīve got no solar-panel at hand.

Quote
3) The UT61E has a (hz) mark on Volts & Amps measurement.  It also has a dedicated (hz) switch
     position!    The frequency measurement works only on the (hz) switch position. 
     Do you know when to use each one?
     My guess is that the (hz) on the Volts & Amps works only when measuring an AC source and
     the (hz) dial position works only when measuring DC ripple voltages.  Am I correct?

Sorry, no. With the function at the dedicated rotary-switch position is the DMM more sensible at 3,3 to 5 Volts (common digital-signal-level), is responding up to 220 MHz and has an upper limit of 30 V. When you using the second function Knob at measuring Amps or Volts, then the sensivity is controlled through their range and it is responding only up to 1 kHz.
For more details: Please RTFM. http://www.uni-trend.com/manual2/UT61English.pdf

Quote
4) Is there any better way to do the same thing?  Maybe the BPW34 photo-diode instead of the
    solar panel?

Perhaps it is a more effective way to build it with a photo-transistor.
You need only the photo-transistor, a voltage-source (DC), a resistor, some wire and something
(e.g. a plastic-tube) as a housing to avoid anbient stray-light.
That gives you a quite nearly digital signal and you can measure up to MHz.

Quote
Sorry for my the lengthy first post but I wanted to share something that I thought was very
interesting.

Sorry for my long answer, but I donīt know, how much you know, neither the level of the other readers.
Iīm still hoping, my english is good enough...

Quote
Thank you.

No problem. Youīre welcome.

Hartmut
Increasing the accuracy and the resolution of DMMīs has one big advantage:
Now it is possible to make far more precise mistakes...
and very exact errors.

SeanB

  • Administrator
  • Hero Member
  • *****
  • Posts: 1012
Re: Easy & Accurate way to Measure Flickering Frequency of a light source
« Reply #2 on: April 18, 2013, 11:38:30 AM »
You would probably be better off using a large area silicon photodiode without an IR filter or with a visible light sensitivity ( buy new or buy on eBay) with a small amplifier to give a voltage that is usable with the meter. If you buy the UT81 scopemeter you will also have the ability to see the actual waveform of the backlight illumination, but that will be a lot more. Photocells generally have a very slow response, a small one might be able to do low audio frequencies, but will not have much response over 400Hz or so, and larger ones are worse.

hgg

  • Newbie
  • *
  • Posts: 5
Re: Easy & Accurate way to Measure Flickering Frequency of a light source
« Reply #3 on: April 18, 2013, 01:56:50 PM »
Harmut thank you for your reply.
Some things I knew and some things I didn't.

Measuring the rise time of the solar panel might be the way to measure the maximum
frequency that is can handle!  I will have to test that but I need an oscilloscope...

Concerning my third question, if I put the multimeter to the V/A Hz position it will not
show anything even when actual voltage is more than 12V and the frequency very low. 
It only works on the dedicated Hz position that is why I am wondering if the Hz position
has a completely different function from the others.

SeanB if you've seen the YouTube video I've posted, I actually measured a PWM flashlight
flickering at 4 KHz!  So the panel can go much more than 400Hz.  That is why I want to
find a way to measure the upper limit of any solar cell.

Measuring the rise time might be the way to find out. 
Can anybody with an oscilloscope a solar panel and a flash-light do the test?
Any other suggestions?

Thanks!
« Last Edit: April 18, 2013, 02:27:59 PM by hgg »

hgg

  • Newbie
  • *
  • Posts: 5
Re: Easy & Accurate way to Measure Flickering Frequency of a light source
« Reply #4 on: April 19, 2013, 01:54:37 AM »
Hi,

Maybe the following is a better solution than the solar panel:

http://www.toshiba-components.com/mobile/IlluminationSensor.html

Its so small that it could easily be an integral part of any multimeter!
Its a compact photo IC with an integrated op-amp which means less work...   :)
The output voltage I guess will be supplied directly to the multimeter.

Toshiba gives a typical rise time of 11μs but a fall time of 400μs.  I am not
sure how this is going to affect the total response time.  It looks fast though.

http://www.semicon.toshiba.co.jp/info/docget.jsp?type=datasheet&lang=en&pid=TPS852

But the question remains.  Will it be faster than a solar panel?
How can we compare the two?

SeanB

  • Administrator
  • Hero Member
  • *****
  • Posts: 1012
Re: Easy & Accurate way to Measure Flickering Frequency of a light source
« Reply #5 on: April 19, 2013, 02:51:43 PM »
Solar cells ( small ones) can do low frequency audio, but will be poor at the high end. The photodiodes are available with frequency responses up to the Gigahertz range. Common ones will typically go up to 100kHz at typical light levels, though this depends on the type of loading, they are poor as voltage sources ( high capacitance) but good into a virtual earth as a current source.

hgg

  • Newbie
  • *
  • Posts: 5
Re: Easy & Accurate way to Measure Flickering Frequency of a light source
« Reply #6 on: April 20, 2013, 10:37:25 AM »
It looks like that a photodiode like the BPW34 will be a better candidate than a solar cell. The BPW34
has a rise & fall time of 100nsec...  There are also other photodiodes with picosecond rise times! The
price though gets way to high!  From $135 to more than $500.  But things will also get more complicated. 
I suppose that first of all you will need a transimpedance amplifier to convert the current
into voltage.  In order to take full advantage of the speed of the diode, the driver circuit must be fast
too or much faster, correct?  Then, I am not sure what is the voltage range that the multimeter will measure
frequency reliably, especially if you go up to the Mhz range.

For all these reasons I think that if a solar cell will go up to 1 Mhz, that will be more than enough.
That's why I want to find out a way to measure the actual response of a solar cell. 
Not in theory, but in practice. 

SeanB

  • Administrator
  • Hero Member
  • *****
  • Posts: 1012
Re: Easy & Accurate way to Measure Flickering Frequency of a light source
« Reply #7 on: April 20, 2013, 01:37:43 PM »
Simple way is to connect a LED to a signal generator ( a red or green one will be best, not a white one as this has phosphor that has a time delay) and wind up the frequency when driving the LED at around 2v, so it is visibly bright, with a square wave. Then check with the solar cell. Most red/green LED units will work up to over 1MHz, though it is not really specified in the data sheets.

hgg

  • Newbie
  • *
  • Posts: 5
Re: Easy & Accurate way to Measure Flickering Frequency of a light source
« Reply #8 on: April 20, 2013, 04:14:14 PM »
SeanB thank you for the suggestion!  I will try it out.