How much current do OLED displays use?
Those little OLED displays are everywhere, and there's a good reason. They're inexpensive, easy to program and they look good. I wanted to explore how much current they draw so that I can plan better for battery powered projects.
Questions I want answered:
Materials:
These are the displays I used for my tests. I like to use the I2C breakout board versions for simplicity. There may be inefficiencies in the linear regulator and other discrete components which come on the breakout boards, but since the active displays use milliamps, losing a few microamps shouldn't be a big deal.
What about color and grayscale OLED displays?
I was asked on twitter to include a quick test of these OLED displays. I could tell that the power usage was high because the controller chip gets hot to the touch when barely anything is displayed. I no longer have my SSD1351 color OLED to test, but I do have an SSD1327 128x128 4-bit grayscale display. With the display set to 'off', I measured 600uA. Here are some values I measured when the contrast (brightness) was set to the maximum value:
Display filled with color 0 - 2.5mA (every pixel is off/black)
Display filled with color 1 - 85mA
Display filled with color 2 - 111mA
Display filled with color 3 - 147mA
Display filled with color 15 (display reset due to my 3.3v power supply sagging)
In general, the display uses a lot more current to get the same pixel brightness compared to the monochrome SSD1306 displays.
Procedure:
I ran 8 different tests on each display. The "dim" setting has the contrast set at a value of 31. This is comfortable to read in indoor light or at night. 127 is the default contrast and is reasonably bright. 255 is the max brightness. The current was measured by inserting the multimeter probes between the Vcc of the display and a 3.3V power source (in this case, the regulated output of the Arduino). There was very little visible difference between contrast 1 and 31.
Results:
* The 72x40 display didn't appear to respond to the contrast setting and had a permanent and somewhat dim contrast.
* The 128x128 display is sold by Pimoroni; I don't have other examples to test
I thought that the 128x32 was an anomaly from a specific vendor, but a second display from a different vendor behaved the same way. Strangely, a second 64x32 display from the same vendor, bought in the same batch, used twice as much power as the one tested above.
I chose the most efficient display to do a final test which displays text (less than 50% pixel coverage). My ammeter read 630uA for a static text display on the 64x32 OLED.
What about active data writing?
Besides extra activity on the display controller, there will be current leaking through the pullup resistors when the SCL or SDA lines go low. Typical pullup resistors for these displays are 4.7KOhm, so while the data and clock lines are active, about 700uA (3.3V / 4700 ohms) will leak through them. I tried a test where I continuously wrote the same data to the display (same display and same text as above). The results were a reading of 1.47mA. Different pullup resistors and different data being written will produce different results.
Conclusions:
To answer my last question about the practicality of OLEDs versus commonly available LCDs for battery powered projects, I think it depends on the project and power source. A decent sized LCD with the backlight off can consume less than 275uA (my measurements from a 2" 128x64 UC1701). If your power budget is small and your MCU is programmed to sleep most of the time (avg current < 1mA), this can make a huge difference compared to a 5-20mA OLED displaying the same info. On the other hand, if you're powering a Raspberry Pi Zero or ESP32 w/WIFI on (80-120mA), another 5-20mA won't make as big a difference to your power budget.
Questions I want answered:
- How much idle current is drawn when the controller is off versus on (blank display)?
- What's the minimum current to see text indoors?
- What's the maximum current when all pixels are on at their brightest?
- How does display size affect current draw?
- How does active data writing affect current draw?
- It's obvious they use more, but how much current do grayscale and color OLEDs use?
- How practical are OLED displays compared to common LCDs for battery powered projects?
Materials:
- Cheap multimeter (manual scale, assumed accuracy +/-10% at worst)
- Various OLED displays (72x40, 96x16, 128x32, 64x32, 128x64-0.96", 128x64-1.3")
- Arduino compatible microcontroller (Adafruit nRF52840 Feather Express)
- Software (my ss_oled library)
These are the displays I used for my tests. I like to use the I2C breakout board versions for simplicity. There may be inefficiencies in the linear regulator and other discrete components which come on the breakout boards, but since the active displays use milliamps, losing a few microamps shouldn't be a big deal.
What about color and grayscale OLED displays?
I was asked on twitter to include a quick test of these OLED displays. I could tell that the power usage was high because the controller chip gets hot to the touch when barely anything is displayed. I no longer have my SSD1351 color OLED to test, but I do have an SSD1327 128x128 4-bit grayscale display. With the display set to 'off', I measured 600uA. Here are some values I measured when the contrast (brightness) was set to the maximum value:
Display filled with color 0 - 2.5mA (every pixel is off/black)
Display filled with color 1 - 85mA
Display filled with color 2 - 111mA
Display filled with color 3 - 147mA
Display filled with color 15 (display reset due to my 3.3v power supply sagging)
In general, the display uses a lot more current to get the same pixel brightness compared to the monochrome SSD1306 displays.
Procedure:
I ran 8 different tests on each display. The "dim" setting has the contrast set at a value of 31. This is comfortable to read in indoor light or at night. 127 is the default contrast and is reasonably bright. 255 is the max brightness. The current was measured by inserting the multimeter probes between the Vcc of the display and a 3.3V power source (in this case, the regulated output of the Arduino). There was very little visible difference between contrast 1 and 31.
Results:
Type
|
Display
Off
|
On,
but Black
|
Contrast 31,
50% pixels lit
|
Contrast 127,
50% pixels lit
|
Contrast 255,
50% pixels lit
|
Contrast 31,
100% pixels lit
|
Contrast 127,
100% pixels lit
|
Contrast 255,
100% pixels lit
|
128x128 1.12”
|
5mA
|
7.1mA
|
49.5mA
|
104.0mA
|
108.0mA
|
89.9mA
|
104.0mA
|
104.0mA
|
128x64 1.3”
|
490uA
|
490uA
|
15.4mA
|
26.1mA
|
31.4mA
|
20.6mA
|
33.7mA
|
39.1mA
|
128x64 0.96”
|
26uA
|
469uA
|
5.6mA
|
8.1mA
|
11.0mA
|
10.3mA
|
15.1mA
|
20.7mA
|
128x32 0.91”
|
6uA
|
6uA
|
12.1mA
|
16.2mA
|
21.1mA
|
21.3mA
|
29.0mA
|
38.7mA
|
128x32 0.91”
|
4uA
|
4uA
|
16.2mA
|
19.7mA
|
21.7mA
|
27.5mA
|
33.8mA
|
36.0mA
|
96x16 0.69”
|
7uA
|
7uA
|
4.45mA
|
5.62mA
|
6.96mA
|
7.23mA
|
9.44mA
|
12.11mA
|
64x32 0.49”
|
6uA
|
6uA
|
1.56mA
|
2.71mA
|
3.11mA
|
2.49mA
|
4.63mA
|
5.17mA
|
72x40 0.42”
|
4uA
|
230uA*
|
630uA*
|
N/A
|
N/A
|
93
|
N/A
|
N/A
|
* The 128x128 display is sold by Pimoroni; I don't have other examples to test
I thought that the 128x32 was an anomaly from a specific vendor, but a second display from a different vendor behaved the same way. Strangely, a second 64x32 display from the same vendor, bought in the same batch, used twice as much power as the one tested above.
I chose the most efficient display to do a final test which displays text (less than 50% pixel coverage). My ammeter read 630uA for a static text display on the 64x32 OLED.
What about active data writing?
Besides extra activity on the display controller, there will be current leaking through the pullup resistors when the SCL or SDA lines go low. Typical pullup resistors for these displays are 4.7KOhm, so while the data and clock lines are active, about 700uA (3.3V / 4700 ohms) will leak through them. I tried a test where I continuously wrote the same data to the display (same display and same text as above). The results were a reading of 1.47mA. Different pullup resistors and different data being written will produce different results.
Conclusions:
- There is a nearly linear relationship between the number of lit pixels and current use
- The display size is directly related to the current use
- There are some oddball displays that use more current than expected. Need to test each one individually
- Actively writing to the display uses more current than a static display - hard to know if it's the I2C activity or the controller chip or both
- There's something odd about the 72x40 displays I have; I will need to get some from another vendor and retest
- The ubiquitous 128x64 0.96" OLEDs can potentially display a screen full of useful information on less than 5mA
- The power usage of grayscale and color OLEDs is much less efficient when compared to the same size and brightness of monochrome displays.
- A small OLED with indoor readable text can run on less than 1mA
To answer my last question about the practicality of OLEDs versus commonly available LCDs for battery powered projects, I think it depends on the project and power source. A decent sized LCD with the backlight off can consume less than 275uA (my measurements from a 2" 128x64 UC1701). If your power budget is small and your MCU is programmed to sleep most of the time (avg current < 1mA), this can make a huge difference compared to a 5-20mA OLED displaying the same info. On the other hand, if you're powering a Raspberry Pi Zero or ESP32 w/WIFI on (80-120mA), another 5-20mA won't make as big a difference to your power budget.
If you appreciate my blog posts and open source work, please consider sponsoring me on Github:
My Github Sponsors Page
My Github Sponsors Page
The reason why the 72x40 is so din is because you have to issue commands "AD 30" which is apparently a non-standard SSD1306 command they call "Internal IREF setting". Found it in the source code examples of the datasheet of the 72x40 module (page 22)
ReplyDeletehttp://www.icbanq.com/data/ICBShop/board/ZJY001_0.42_16P%20OLED.pdf
Great find - I'll give it a try and let you know if it works.
DeleteI also found that they're not SSD1306. They accept commands for the SSD1316 (although it's not that one either). 0xAD is "External or Internal VCOMH Selection / External or internal IREF Selection". Did you know they can fade and zoom? After experimenting I found out most of the Chinese "SSD1306" displays accept those commands and in fact I recommend using the internal IREF for all displays because different resistors on the backs of e.g. the 128x64 and 128x32 give them different brightness with the same initialization sequence. With the internal IREF delected, all of them (even the small one with smaller pixels) look equally bright.
DeleteIt works! I'll add the fix to the library after I test it on the various display I own. Thanks again for finding that info.
DeleteInteresting effect of contrast on the 72x40 OLED is that contrast values 0-32 give the full range of values while 33-255 are the same (full brightness)
DeleteMust be an artifact from your displays. Here the full range has effect. Although 0x80 almost looks as bright as 0xff but that's because brightness works linearly and the eye works almost logarithmically. At 0x20 it definitely looks much dimmer.
DeleteI know how brightness works for the eye. The brightness value behaves differently from other displays and hits its peak around 32 instead of 255.
DeleteThe one at the upper right of your SD1306 'group photo' looks identical to mine. Even has the same silkscreen typo "70*40".
DeleteVery helpful writing. Was useful for selecting the display for my product.
ReplyDeleteThank you very much! Very helpful!
ReplyDeleteVery helpful, thanks a lot!
ReplyDeleteVery usefull! Thanks a lot.
ReplyDeleteAnybody know about similar measurements with the 1.8inch 128x160 SPI Module ST7735S?
I will measure and add info about color LCDs this week. I've been meaning to do it for a while.
DeleteDo you have a guess about the differences in "display off" current? 490uA vs 4uA is... quite a lot.
ReplyDeleteUnfortunately these little test boards have built-in voltage regulators and I didn't strip them off to get a more accurate current test. I would guess that there's leakage or something else related to the LDO.
DeleteThank you very much for your data, I was looking for a solution for a small distance measurement instrument.
ReplyDeleteLovely reference! Thank you! Been meaning to make a little old exercise timer with a Pro Micro and an OLED screen running off of a LiPo. Looks like I don't have to worry about battery life as much!
ReplyDeleteThanks for doing this research!
ReplyDelete