Author Topic: Beginner, needs to record signals from a display backlit  (Read 9174 times)

0 Members and 1 Guest are viewing this topic.

Offline dogmanTopic starter

  • Contributor
  • Posts: 10
  • Country: us
Beginner, needs to record signals from a display backlit
« on: May 23, 2017, 12:02:47 pm »
Hi all and thanks in advance

I'd like to detect and record the backlit frequency of a display (like a pc display). It could be 60HZ and it could be 20KHZ.
I understand that I probably need a photodiode (http://www.vishay.com/docs/83471/tefd4300.pdf) and an oscilloscope
Given that the photodiode is the correct one, what digital oscilloscope do I need? I saw some recommendations on the saleae - but which version ? I need only a single analog channel?

Thanks a lot!!
« Last Edit: May 25, 2017, 09:02:57 pm by dogman »
 

Offline TK

  • Super Contributor
  • ***
  • Posts: 1722
  • Country: us
  • I am a Systems Analyst who plays with Electronics
Re: Beginner, needs to record dignals from a display backlit
« Reply #1 on: May 24, 2017, 01:12:52 am »
Probably you are ok with a single channel pocket scope like the DSO Nano V3.  The Saleae is a logic analyzer with a capture-and-show analog channel (not real time) and very expensive for what you need.

If you are going to measure the backlight directly, be aware it might be dealing with HIGH VOLTAGE, but if you are measuring like the flicker frequency of the backlight using a light sensor, then it is safer.
 
The following users thanked this post: dogman

Offline LaurentR

  • Frequent Contributor
  • **
  • Posts: 536
  • Country: us
Re: Beginner, needs to record dignals from a display backlit
« Reply #2 on: May 24, 2017, 06:11:16 am »
If you are only after just frequency and not details, many DMMs will have frequency readings and would be simpler (and cheaper) options than oscilloscopes.

Or you could buy a generic frequency meter on EBay. Random example (no experience with these items):
http://www.ebay.com/itm/0-56-DIGITAL-Red-LED-Frequency-and-Tachometer-Rotate-Speed-Meter-DC-12-24V-/371957344295?hash=item569a62a827:g:n5AAAOSwgNRV8PgN&vxp=mtr
 
The following users thanked this post: dogman

Offline bktemp

  • Super Contributor
  • ***
  • Posts: 1616
  • Country: de
Re: Beginner, needs to record dignals from a display backlit
« Reply #3 on: May 24, 2017, 06:48:55 am »
What are you trying to measure?
The light emitted by a display backlight can be a very complex waveform consisting of a low frequency component when the backlight is synchronized with the framerate, a higher frequency for PWM and an even higher frequency when a switching mode power supply is used for generating the LED current.
 
The following users thanked this post: Someone, dogman

Offline dogmanTopic starter

  • Contributor
  • Posts: 10
  • Country: us
Re: Beginner, needs to record dignals from a display backlit
« Reply #4 on: May 24, 2017, 08:52:16 am »
Probably you are ok with a single channel pocket scope like the DSO Nano V3.  The Saleae is a logic analyzer with a capture-and-show analog channel (not real time) and very expensive for what you need.

If you are going to measure the backlight directly, be aware it might be dealing with HIGH VOLTAGE, but if you are measuring like the flicker frequency of the backlight using a light sensor, then it is safer.

Thanks TK!
No, not directly, but via a light sensor, yes.
I'd like to measure the flicker and response times (rise and fall), and maybe other parameters that I'm still not aware of ?

If you are only after just frequency and not details, many DMMs will have frequency readings and would be simpler (and cheaper) options than oscilloscopes.

Or you could buy a generic frequency meter on EBay. Random example (no experience with these items):
http://www.ebay.com/itm/0-56-DIGITAL-Red-LED-Frequency-and-Tachometer-Rotate-Speed-Meter-DC-12-24V-/371957344295?hash=item569a62a827:g:n5AAAOSwgNRV8PgN&vxp=mtr

That's actually a nice option I was not aware of!
I'd like to know the details too, especially rise and fall. I'd like to measure (input) lag too, but it requires some other equipment, no? It would be great to know how can I measure it
I know that I can put a CRT display side by side, but I'd like something more convenient


What are you trying to measure?
The light emitted by a display backlight can be a very complex waveform consisting of a low frequency component when the backlight is synchronized with the framerate, a higher frequency for PWM and an even higher frequency when a switching mode power supply is used for generating the LED current.

Thanks bktemp!

Would like to start from the backlit frequency (probably the PWM, when exists) and rise/fall
I'm talking about the backlit itself, not the part that actually shows the image (and sometimes is synced with the FPS)

Can you explain how the switching mode effects the image?

---

Thanks all! I'm amazed from all that help
 

Offline TK

  • Super Contributor
  • ***
  • Posts: 1722
  • Country: us
  • I am a Systems Analyst who plays with Electronics
Re: Beginner, needs to record dignals from a display backlit
« Reply #5 on: May 24, 2017, 10:45:23 am »
If you use a light sensor, you will measure the rise and fall time of the sensor circuitry, not the backlight.

About using a DMM,  some can measure frequency and duty cycle

If you explain what you want to accomplish, maybe the approach is a totally different from the one you proposed originally.
 

Offline dogmanTopic starter

  • Contributor
  • Posts: 10
  • Country: us
Re: Beginner, needs to record dignals from a display backlit
« Reply #6 on: May 24, 2017, 10:59:59 am »
If you use a light sensor, you will measure the rise and fall time of the sensor circuitry, not the backlight.

About using a DMM,  some can measure frequency and duty cycle

If you explain what you want to accomplish, maybe the approach is a totally different from the one you proposed originally.

Thanks TK. I'm not knowledgeable

I do basic tests to laptop displays. I want to know their response (rise/fall) and backlit frequency (the one that could be described as flickering)
Not sure I know how to describe better than that

What I figured is that I need a photodiode and some oscilloscope
 

Offline Muxr

  • Super Contributor
  • ***
  • Posts: 1369
  • Country: us
Re: Beginner, needs to record dignals from a display backlit
« Reply #7 on: May 24, 2017, 12:12:02 pm »
Just having a photodiode isn't going to tell you much.. you will have to design a transimpedance amplifier around it to amplify exactly the stuff you want to see.

Maybe you just need a camera with a high shutter speed say 8000/s, manual controls and some math foo. (not sure what you're after)

edit: as someone above mentioned you will not get accurate rise and fall times.. these will be your photodiode circuit's raise and fall times. You might be able to get the frequency.. duty cycle would probably be the other thing you're interested in, but this will require additional considerations in your amplifier and it's a no trivial task.
« Last Edit: May 24, 2017, 12:18:28 pm by Muxr »
 
The following users thanked this post: dogman

Offline bktemp

  • Super Contributor
  • ***
  • Posts: 1616
  • Country: de
Re: Beginner, needs to record dignals from a display backlit
« Reply #8 on: May 24, 2017, 12:24:29 pm »
If you have enough money, there are many devices designed specifically for measuring these parameters. But they can be very expensive.

Using a scope will be much cheaper and you can use it for other measurements as well.
For this type of measurment most entry level DSOs should do the job, but for identifying flicker frequencies having a good FFT function is probably useful.
Maybe a PC based scope?

For the photdiode you need an amplifier with enough bandwidth (at least 10kHz, better around 100kHz).
If you want a cheap, easy to use solution, look for sensors with integrated amplifiers, something like this:
http://www.ti.com/product/OPT101
Otherwise you need to add a similar circuit around your photodiode for converting the current into a voltage.

 
The following users thanked this post: dogman

Offline serggio

  • Regular Contributor
  • *
  • Posts: 149
  • Country: ru
Re: Beginner, needs to record dignals from a display backlit
« Reply #9 on: May 24, 2017, 01:44:55 pm »
http://www.vishay.com/docs/80085/measurem.pdf
To converting photocurrent to voltage will be enough 9V battery and resistor :)
Main problem that for fast rise/fall time measured in this circuit need have active probe with very small input capacitance. 
Right measurement can be done only on backlight driver circuit.
« Last Edit: May 24, 2017, 08:02:58 pm by serggio »
 
The following users thanked this post: dogman

Offline dogmanTopic starter

  • Contributor
  • Posts: 10
  • Country: us
Re: Beginner, needs to record dignals from a display backlit
« Reply #10 on: May 25, 2017, 03:26:05 pm »
If you have enough money, there are many devices designed specifically for measuring these parameters. But they can be very expensive.

Using a scope will be much cheaper and you can use it for other measurements as well.
For this type of measurment most entry level DSOs should do the job, but for identifying flicker frequencies having a good FFT function is probably useful.
Maybe a PC based scope?

For the photdiode you need an amplifier with enough bandwidth (at least 10kHz, better around 100kHz).
If you want a cheap, easy to use solution, look for sensors with integrated amplifiers, something like this:
http://www.ti.com/product/OPT101
Otherwise you need to add a similar circuit around your photodiode for converting the current into a voltage.



Thanks

I'm a bit flooded here. I don't have real electronics knowledge

0. I can spend some money, not trying to keep it $1 cheap
1. A PC scope like the saleae that I saw recommended here? Don't oscilloscopes have an integrated amplifier?
2. Is there something I just plug in the photodiode?
3. Are there any considerations I need to thinks about regarding rise/fall times?

Thanks again
 

Offline Muxr

  • Super Contributor
  • ***
  • Posts: 1369
  • Country: us
Re: Beginner, needs to record signals from a display backlit
« Reply #11 on: May 25, 2017, 11:57:09 pm »
As bktemp above mentioned instead of using a photodiode, get a full solution of the photodiode and transimpedance amplifier combined. This will at least remove that non trivial (for a beginner) task of making a amplifier out of the equation.

What is your end goal here? Are you just trying to review some monitors and you want to be able to tell which ones are prone to visible flicker, perhaps as you adjust the backlight brightness level? Because I am having a hard time understanding why fall and rise time are a useful metric here. Today's backlights are LED backlit and I could be wrong but from what I remember LED rise times are super quick, way beyond human's perceptibility.

Instead for instance a far more valuable metric would be the duty cycle.


Because when you know the frequency and the duty cycle you can determine the duration gap when the LED is turned off. And hence can probably best judge the amount of flicker a panel has.
1. A PC scope like the saleae that I saw recommended here?
Saleae is a logic analyser, not technically a scope. I would personally get a real scope, a logic analyzer is great at decoding protocols but it's not best suited for this task. It does not have the ability to actually look at the signal at a low level.

1. Don't oscilloscopes have an integrated amplifier?
They do, however just a photodiode will produce really weak signal, and a scope is a wideband device, it will amplify all sorts of signals that the diode and the probe might pickup. That OPT101 a breadboard, and a power supply should be all you need to at least have a chance at capturing something useful.
 
The following users thanked this post: Richard Crowley, dogman

Offline Muxr

  • Super Contributor
  • ***
  • Posts: 1369
  • Country: us
Re: Beginner, needs to record signals from a display backlit
« Reply #12 on: May 26, 2017, 12:18:35 am »
Additional reading: https://electronics.stackexchange.com/a/86720/77469
Quote
Phosphor type LEDs have turn-on and turn-off times in the tens to hundreds of nanoseconds, appreciably slower than direct emission LEDs.

Even in worse case scenario (phosphor LEDs) 100s of nanoseconds is 0.00000001 in seconds, that a really quick raise time, and as mentioned before your photodiode and amp will be the limiting factor. 
 

Offline TK

  • Super Contributor
  • ***
  • Posts: 1722
  • Country: us
  • I am a Systems Analyst who plays with Electronics
Re: Beginner, needs to record signals from a display backlit
« Reply #13 on: May 26, 2017, 01:13:08 am »
Saleae is a logic analyser, not technically a scope. I would personally get a real scope, a logic analyzer is great at decoding protocols but it's not best suited for this task. It does not have the ability to actually look at the signal at a low level.
Saleae has newer models that has analog channels, but they are capture-then-display static images on the PC application.  And they are VERY limited at 50MSa/s, 5MHz bandwidth.
 
The following users thanked this post: Muxr, dogman

Offline dogmanTopic starter

  • Contributor
  • Posts: 10
  • Country: us
Re: Beginner, needs to record signals from a display backlit
« Reply #14 on: May 26, 2017, 09:29:35 am »
I'm really amazed with all the help and willingness, really

Muxr:
1. Yes, I do laptop reviews and one aspect I scratch is the display stuff. The rise and fall times are related to the change in the image. It's important for gaming and even movies and the resulted "blur" or "fudge" can be seen in many screens. Maybe the reason is completely different? But still these differences are visible

2. Yes, flicker too. True, in some cases it's hard to see the flicker right away, but it does effect you. And yes, I check various brightness levels.
Also, I want more precise measurements, as I can't count on my own eyes and brains

About the duty cycle - you talk about measuring the PWM frequency on the motherboard?

3. Any recommendations for such a scope? Anything with good PC software?

many thanks
« Last Edit: May 26, 2017, 09:32:41 am by dogman »
 

Offline dogmanTopic starter

  • Contributor
  • Posts: 10
  • Country: us
Re: Beginner, needs to record signals from a display backlit
« Reply #15 on: May 26, 2017, 09:34:39 am »
Saleae has newer models that has analog channels, but they are capture-then-display static images on the PC application.  And they are VERY limited at 50MSa/s, 5MHz bandwidth.

Oh, so no graphs and stuff like that? They can record the values?
5MHZ should be enough for me, no?
 

Offline Muxr

  • Super Contributor
  • ***
  • Posts: 1369
  • Country: us
Re: Beginner, needs to record signals from a display backlit
« Reply #16 on: May 26, 2017, 03:19:32 pm »
Muxr:
1. Yes, I do laptop reviews and one aspect I scratch is the display stuff. The rise and fall times are related to the change in the image. It's important for gaming and even movies and the resulted "blur" or "fudge" can be seen in many screens. Maybe the reason is completely different? But still these differences are visible

It's important to know how a display delivers an image, right? You have the white phosphorus LED in the back of the display illuminating the thin-film transistor (TFT), and this produces the image. Say you push an all white image to the screen. You will be able to measure backlight's frequency and duty cycle. You can calculate the amount of flickr from that.

To measure the TFT response time, you might try pushing an all black and then one of the colors Red/Green/Blue, you might get a rise time by measuring a change from all black to all blue.

But remember the backlight is constantly "flickering" in the background so it might not be trivial to extrapolate these two things. Photodiode is technically color blind. All it sees is intensity. So extrapolating TFT's response time from that might be tricky. Might be doable though. But you have to know what's happening in order understand the signal you're measuring. I haven't done this myself so I am purely speculating here. And excuse the crudity of the following diagram (wireless mouse is not the best for drawing).

So TFT's raise time if it's slow will get chopped up by the backlight flicker. Remember backlights are LED, LEDs need to be current limited otherwise they burn up. Display manufacturers use something called PWM to limit current to the backlight which is why the backlight is always alternating from on to off state.

Also TFT's refresh and PWM of the backlight are not necessarily synced (most likely not) so the TFT intensity measurement will be "chopped" up by the backlight randomly.
« Last Edit: May 26, 2017, 03:22:02 pm by Muxr »
 

Offline Muxr

  • Super Contributor
  • ***
  • Posts: 1369
  • Country: us
Re: Beginner, needs to record signals from a display backlit
« Reply #17 on: May 26, 2017, 03:24:30 pm »
As for your other part of the question. Which oscilloscope to get?

I like the DS1054z from Rigol for this. Multiple reasons.

- It's a great beginner scope.

- It has an LXI interface and should you automate some of this testing, there are tons of examples of folks doing just that.

- And it's affordable.

But it's certainly not the only way to go.
 

Offline serggio

  • Regular Contributor
  • *
  • Posts: 149
  • Country: ru
Re: Beginner, needs to record signals from a display backlit
« Reply #18 on: May 26, 2017, 06:08:43 pm »
Muxr, just let me ask you:
- How much you did measured LCD monitor backlight update frequency thru TN, IPS, MVA or other panel type with photodiode and amplifier without monitor disassembling? What you did with frame rate?
- Have you ever did any measuring just for rise/fall time for silicon PIN photodiodes? Your best measuring?
Regards
« Last Edit: May 26, 2017, 06:14:05 pm by serggio »
 

Offline TK

  • Super Contributor
  • ***
  • Posts: 1722
  • Country: us
  • I am a Systems Analyst who plays with Electronics
Re: Beginner, needs to record signals from a display backlit
« Reply #19 on: May 26, 2017, 07:23:17 pm »
Saleae has newer models that has analog channels, but they are capture-then-display static images on the PC application.  And they are VERY limited at 50MSa/s, 5MHz bandwidth.

Oh, so no graphs and stuff like that? They can record the values?
5MHZ should be enough for me, no?
5MHz is on their high end logic pro 8 or logic pro 16.  The logic 4 ($109) has a bandwidth of just 600kHz.  There are pocket scopes for around $50 that has bandwidth around 1-5MHz.  If you don't need the logic analyzer on the saleae, don't go for it, it is too expensive (Logic Pro 8 is $479, you can buy a REAL scope with digital channels for that kind of money).

The saleae is not a real oscilloscope, it can capture analog signal when you ask it to start sampling and shows the waveform when the memory is filled up.
« Last Edit: May 26, 2017, 07:35:03 pm by TK »
 

Offline Muxr

  • Super Contributor
  • ***
  • Posts: 1369
  • Country: us
Re: Beginner, needs to record signals from a display backlit
« Reply #20 on: May 26, 2017, 07:30:05 pm »
Muxr, just let me ask you:
- How much you did measured LCD monitor backlight update frequency thru TN, IPS, MVA or other panel type with photodiode and amplifier without monitor disassembling? What you did with frame rate?
I thought I covered this part in my post:
Quote from: Muxr
I haven't done this myself so I am purely speculating here.
If you have, feel free to correct any misconceptions I may have.

- Have you ever did any measuring just for rise/fall time for silicon PIN photodiodes? Your best measuring?
My experience with photodiodes is limited to IR applications though.

So from the OPT101 datasheet:


Those rise fall times are too slow to capture backlight rise times.. but perhaps not too slow to capture TFT rise time. My understanding is TFT at least is slow:


I could be wrong, as I said I never tried this. Your input is welcome.

« Last Edit: May 26, 2017, 07:48:26 pm by Muxr »
 

Offline PointyOintment

  • Frequent Contributor
  • **
  • Posts: 327
  • Country: ca
  • ↑ I scanned my face
Re: Beginner, needs to record signals from a display backlit
« Reply #21 on: May 27, 2017, 04:55:13 am »
You have the white phosphorus LED

White phosphor-based LED. I sure hope there is no white phosphorus in there.

Remember backlights are LED, LEDs need to be current limited otherwise they burn up. Display manufacturers use something called PWM to limit current to the backlight which is why the backlight is always alternating from on to off state.

PWM is used to regulate the time-averaged current, and therefore the brightness, in response to the user's brightness setting. But the instantaneous current while the PWM is in its on state still needs regulation to avoid burning out the LED (though the allowable pulsed current through an LED is often higher than the allowable continuous current).
I refuse to use AD's LTspice or any other "free" software whose license agreement prohibits benchmarking it (which implies it's really bad) or publicly disclosing the existence of the agreement. Fortunately, I haven't agreed to that one, and those terms are public already.
 
The following users thanked this post: dogman

Offline bktemp

  • Super Contributor
  • ***
  • Posts: 1616
  • Country: de
Re: Beginner, needs to record signals from a display backlit
« Reply #22 on: May 27, 2017, 07:13:14 am »
Who cares about the rise/fall time of the PWM?
Humans can perceive flicker up to a couple of kHz at most. Therfore you don't get any useful information from that.

So being able to measure from DC to 10kHz is enough, maybe up to 100kHz if you interested in the actual PWM frequency.
So even cheap photodiodes with a nearby amplifier should be fast enough for that job. Phototransistors are typically too slow.

As already shown, the signal measured can be difficult to analyse if you don't have full control over the backlight PWM.

I did some quick measurements using a OPT101 (14kHz BW) and another photodiode+amplifier (~250kHz BW). The OPT101 is much more sensitive than the other sensor and therefore better suited for the typical brightness from a display (that does not mean you can't get a fast sensor with a high BW, but it gets more difficult if you need both, because at some point you end up using a photomultiplier tube).

OPT101:


250kHz sensor:


As you can see the waveform measured using OPT101 has the same shape as the other signal, but it is missing the high frequency contents (double the operating frequency of the CCFL). The lower frequency is the 180Hz PWM frequency (3x framerate).


This is the light output of the CCFL.


Let's say you want to measure the response time of the display, this would be difficult with this signal because of all the other frequencies and the PWM signal masking the wanted signal. To get rid of the PWM signal, I did set the brightness to 100%, ac coupled the signal and applied some filtering to remove all of the CCFL operating frequency. I have no idea why the waveform looks like this, maybe the monitor can't do 100% PWM but only 99.9% or so, producing a small spike. But you can clearly see every 3rd cycle looks different.



Looking at the frequency spectrum we can clearly the the peak at 91.5kHz and half of that at ~46kHz. The small spike at ~66kHz could be ripple from the main SMPS. But except from that there is not much else interesting to see above a couple of kHz.



If we zoom in, we see spikes at multiples of 180Hz. Hard to see here, but with a very long FFT (>1M samples) we can see peaks at 30Hz and 60Hz showing the actual flicker of the TFT LCD. I can also see a small spike at 15Hz. That could be dithering because the TFT panel itself probably is 6bit and is expanded to 8bit using framerate modulation.

That's why I suggested using a PC based scope or a scope with a good PC interface for downloading the samplebuffer for further processing, because the built in FFT in most low end scopes isn't good enough for high resolution FFT using the full sample memory length.

You can do many different measurements on displays using optical sensor and get lots of information from that, but interpreting the results can be difficult.

Typical photodiodes have a peak resonse in the near IR, but that is ok if you only interested in changes of the brightness.
There are photodiodes with an additional filter to match the response of the human eye, but since they filter out light, the sensitivity is lower and they are more expensive.
If you care about absolute brightness or colours, there are also calibrated colour sensors, but they are even more expensive (you can easily spend >100k on optical measuring equipment for displays).
« Last Edit: May 27, 2017, 07:18:17 am by bktemp »
 
The following users thanked this post: Muxr, dogman

Offline Muxr

  • Super Contributor
  • ***
  • Posts: 1369
  • Country: us
Re: Beginner, needs to record signals from a display backlit
« Reply #23 on: May 27, 2017, 10:13:00 am »
Great info bktemp!

That's why I suggested using a PC based scope or a scope with a good PC interface for downloading the samplebuffer for further processing, because the built in FFT in most low end scopes isn't good enough for high resolution FFT using the full sample memory length.
Yeah, this is why I suggested 1054z because there are a lot of examples on how to fetch samples from it:

http://hackaday.com/2015/09/22/a-better-spectrum-analyzer-for-your-rigol-scope/
 
The following users thanked this post: dogman

Offline dogmanTopic starter

  • Contributor
  • Posts: 10
  • Country: us
Re: Beginner, needs to record signals from a display backlit
« Reply #24 on: May 28, 2017, 12:14:20 am »
Thanks all!

That's a bit of a problem quoting every reply!

1. I really don't know how to deal with the flickering interfering with the rise/fall. Maybe indeed what should be tested is the signal going out to the display from the motherboard?
Here is an example of a review that does include both tests
https://www.notebookcheck.net/Dell-Latitude-5289-i5-256GB-8GB-Convertible-Review.221832.0.html (Search for the "PWM" keyword)

How do they do that?

2. The diode I linked to in the first post, has a rise/fall times of 100ns, which should allow a good enough measurements

3. The FFT graph is indeed hard for me to understand. I mean, I don't understand what I can get out from it (except the constant spikes).
By "actual flicker" you meant the frames? otherwise, what are these 30HZ and 60HZ signals?


Thanks for the Rigol recommendation too!
« Last Edit: May 28, 2017, 11:35:20 am by dogman »
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf