Author Topic: EEVblog #1242 - Memory LCD+Supercaps+Low Power Design  (Read 2303 times)

0 Members and 1 Guest are viewing this topic.

Offline RobertYates

  • Newbie
  • Posts: 1
  • Country: us
Re: EEVblog #1242 - Memory LCD+Supercaps+Low Power Design
« Reply #25 on: August 31, 2019, 05:20:13 pm »
Just thought I'd chime in here to point out that what you're trying to create seems very similar to one of these modules.

https://www.hardlineproducts.com/product/hourtach-meter/

They have a very simple operation ( ie. they only count hours or show rpm ) and the lcd is always on, very similar to your cycle counter concept. It might be interesting to get one and tear it apart to see what they use to drive the LCD, though they are for use on outdoor equipment so they're probably fully potted.  :-\

Also note that there are other manufacturers of these style of engine hour meter, but I've actually used this one so I know for a fact that the LCD is always on.
 

Online GeorgeOfTheJungle

  • Super Contributor
  • ***
  • Posts: 1980
  • Country: pl
Re: EEVblog #1242 - Memory LCD+Supercaps+Low Power Design
« Reply #26 on: August 31, 2019, 05:45:49 pm »
Specialists in low power ARM: https://www.silabs.com/products/development-tools/mcu/32-bit/efm32-zero-gecko-starter-kit

Quote
6.3 AEM accuracy and performance
The Advanced Energy Monitor is capable of measuring currents in the range of 0.1uA to 50mA. For currents above 250uA, the AEM is accurate within 0.1mA. When measuring currents below 250uA, the accuracy increases to 1uA. Even though the absolute accuracy is 1uA in the sub 250uA range, the AEM is able to detect changes in the current consumption as small as 100nA. The AEM produces 6250 current samples per second.

Quote
1.2 Features
• EFM32ZG222F32 MCU with 32KB Flash and 4K BRAM.
• Advanced Energy Monitoring system for precise current tracking.
• Integrated SEGGER J-Link USB debugger/emulator with debug out functionality.
• 20 pin expansion header.
• Breakout pads for easy access to I/O pins.
• Power sources include USB and CR2032 battery.
• 2 user buttons, 2 user LEDs
• 2 capacitve touch pads
• Ultra low power 128x128 pixel Memory-LCD
• Crystals for LFXO and HFXO:32.768 kHz and 24.000 MHz.
« Last Edit: August 31, 2019, 05:51:50 pm by GeorgeOfTheJungle »
Even when the experts all agree, they may well be mistaken.
 

Offline Howardlong

  • Super Contributor
  • ***
  • Posts: 4755
  • Country: gb
Re: EEVblog #1242 - Memory LCD+Supercaps+Low Power Design
« Reply #27 on: August 31, 2019, 06:03:36 pm »
Specialists in low power ARM: https://www.silabs.com/products/development-tools/mcu/32-bit/efm32-zero-gecko-starter-kit

Quote
6.3 AEM accuracy and performance
The Advanced Energy Monitor is capable of measuring currents in the range of 0.1uA to 50mA. For currents above 250uA, the AEM is accurate within 0.1mA. When measuring currents below 250uA, the accuracy increases to 1uA. Even though the absolute accuracy is 1uA in the sub 250uA range, the AEM is able to detect changes in the current consumption as small as 100nA. The AEM produces 6250 current samples per second.

Quote
1.2 Features
• EFM32ZG222F32 MCU with 32KB Flash and 4K BRAM.
• Advanced Energy Monitoring system for precise current tracking.
• Integrated SEGGER J-Link USB debugger/emulator with debug out functionality.
• 20 pin expansion header.
• Breakout pads for easy access to I/O pins.
• Power sources include USB and CR2032 battery.
• 2 user buttons, 2 user LEDs
• 2 capacitve touch pads
• Ultra low power 128x128 pixel Memory-LCD
• Crystals for LFXO and HFXO:32.768 kHz and 24.000 MHz.

Silabs and their Gecko series have good marketing. In practice, I haven't found them particularly good compared to some other options.

Cortex M0, being 32 bit, is quite a tax in power terms despite what perceptions may be.

The devil is in the detail.
 

Online Fungus

  • Super Contributor
  • ***
  • Posts: 10114
  • Country: 00
Re: EEVblog #1242 - Memory LCD+Supercaps+Low Power Design
« Reply #28 on: August 31, 2019, 09:36:42 pm »
I agree, that has been my experience with them too. While they do indeed draw no current when you’re not changing the display, it’s a completely different story when you do. You also seem to need to do a complete clean and refresh, at least on some units, as small areas that you change relatively frequently become “dirty” and the only way to correct for this is a complete refresh, expensive in terms of power.

I've read a few datasheets and they say things like the display can become permanently dirty if you don't do complete refreshes every few updates.

Also that direct sunlight will destroy them, and lots of other scary things.

If there's an LCD that maintains the display with no external signals and only draws 2uA then that seems the way to go.
 

Offline Kilrah

  • Supporter
  • ****
  • Posts: 1758
  • Country: ch
Re: EEVblog #1242 - Memory LCD+Supercaps+Low Power Design
« Reply #29 on: September 01, 2019, 08:28:55 am »
Also while e-ink will "retain an image forever" without power, the image may degrade over time.
Might be more pronounced with the "3-color" ones, but I bought one of those white/black/red ones and it likely sat for a couple of years on a shelf with the B/W company logo on, and it got all tainted red over time. So it'll likely still be readable, but contrast may reduce.
 

Offline thmjpr

  • Regular Contributor
  • *
  • Posts: 118
  • Country: ca
 

Offline mariush

  • Super Contributor
  • ***
  • Posts: 3818
  • Country: ro
  • .
Re: EEVblog #1242 - Memory LCD+Supercaps+Low Power Design
« Reply #31 on: September 01, 2019, 09:53:13 pm »
could go with a plain solar cell... https://www.digikey.com/product-detail/en/ixys/KXOB25-01X8F/KXOB25-01X8F-ND/9990432
high efficiency ... the room light alone should be enough for a few hundred uA...if you need to see lcd screen you have enough light to power it.
cheaper cells exist
store counter in eeprom in micro..  could use a 6-8 pin micro and separate lcd driver ic

if you can live with 1999 display, dt830 multimeters are 0.5$ in volume ...pull lcd and zebra strip out. job done.
« Last Edit: September 01, 2019, 09:55:54 pm by mariush »
 

Offline Howardlong

  • Super Contributor
  • ***
  • Posts: 4755
  • Country: gb
Re: EEVblog #1242 - Memory LCD+Supercaps+Low Power Design
« Reply #32 on: September 01, 2019, 10:00:21 pm »
could go with a plain solar cell... https://www.digikey.com/product-detail/en/ixys/KXOB25-01X8F/KXOB25-01X8F-ND/9990432
high efficiency ... the room light alone should be enough for a few hundred uA...if you need to see lcd screen you have enough light to power it.
cheaper cells exist
store counter in eeprom in micro..  could use a 6-8 pin micro and separate lcd driver ic


I have a few of these in a couple of different specs. I’ll be straight up with you about them, they’re disappointing in normal inside room lighting. If I get chance I’ll redo some tests and share.
 

Offline Kilrah

  • Supporter
  • ****
  • Posts: 1758
  • Country: ch
Re: EEVblog #1242 - Memory LCD+Supercaps+Low Power Design
« Reply #33 on: September 01, 2019, 10:01:21 pm »
Would expect the device to be enclosed in a dark cabinet and only viewed by an engineer allowed to open the panels when troubleshooting. No reason to think it would get any light.
 

Offline JustMeHere

  • Regular Contributor
  • *
  • Posts: 185
  • Country: us
Re: EEVblog #1242 - Memory LCD+Supercaps+Low Power Design
« Reply #34 on: September 02, 2019, 12:54:02 am »
The specs for the E Ink says .5ua per cm^2. The module is a shade over 2 cm^2.  So it draws at the most 3ua.

LOL, the specs can say whatever, play with one and you'll see what I mean.

LOL, this project is supposed to change once per power cycle.
 

Offline chickenHeadKnob

  • Frequent Contributor
  • **
  • Posts: 815
  • Country: ca
  • doofus programus semi-retiredae
Re: EEVblog #1242 - Memory LCD+Supercaps+Low Power Design
« Reply #35 on: September 02, 2019, 01:57:06 am »
Would expect the device to be enclosed in a dark cabinet and only viewed by an engineer allowed to open the panels when troubleshooting. No reason to think it would get any light.
Doesn't need to, power cycle count is in eeprom. When the cabinet is opened the low power micro and LCD wakes up. I like this idea. I noticed my 37 year old sharp solar calculator now has dead batteries and doesn't work at all with moderate interior natural light however it works fine with high LCD contrast when illuminated with LED lighting.
 

Offline Kilrah

  • Supporter
  • ****
  • Posts: 1758
  • Country: ch
Re: EEVblog #1242 - Memory LCD+Supercaps+Low Power Design
« Reply #36 on: September 02, 2019, 10:47:17 am »
Indeed, that works
 

Offline NANDBlog

  • Super Contributor
  • ***
  • Posts: 4465
  • Country: nl
Re: EEVblog #1242 - Memory LCD+Supercaps+Low Power Design
« Reply #37 on: September 02, 2019, 11:22:31 am »
There is a product here, which has a 3.5 Digit memory LCD driven by an NXP low power LCD driver chip and a micro. The power consumption is about 5-10uA @ 3.3V (it is on a LiSOCl2 cell). Although it is a much bigger screen than what you have, the power consumption is easily the larger than what it comes from the datasheet. Although it is funny, when you disconnect the power source, it keeps on being visible for like 30 seconds with just the small capacitors on the board.
But selecting an off the shelf LCD and designing a project around it is mayor pain in the ass, so I would reconsider it with a button, or just blast the result through an UART every minute or something.
 

Offline Howardlong

  • Super Contributor
  • ***
  • Posts: 4755
  • Country: gb
Re: EEVblog #1242 - Memory LCD+Supercaps+Low Power Design
« Reply #38 on: September 06, 2019, 09:38:07 am »
Yesterday I did some evaluation testing with a few devices, notably PIC16LF1904, PIC16LF1934 and PIC16LF19175 that are largely pin for pin compatible drop in devices. 1ku pricing is $1.18, $1.46 and $1.19 respectively. All are available in 40 pin 5x5mm UQFN, but I used the DIP variants for this.

The best I managed so far at 3V with the display operating but with the CPU in sleep is on a PIC16LF1904, at 410nA plus 14nA per on-segment with the display attached, in static mode. This was using the internal LF oscillator, but no RTC functionality (I don't believe an RTC is a requirement in this case).

Measurements are taken with a Keithley 238 SMU and triaxial cable assemblies to the DUT. The MCUs were on solderless breadboard, and the debugger and any other probing removed. The LCD is DE 117-RS-20/7.5 soldered on a breakout board that had been cleaned in an ultrasonic cleaner.

As usual, you should never trust the datasheet when it comes to low power current draw specs: sometimes it's better, sometimes it's worse, and often the difference is dramatic. Judging by the quality of the quoted figures, I strongly suspect they use unsupervised interns on this work. Measuring sub-uA takes care and attention.

I found some interesting facets regarding the sleep currents. Sleep currents are important because they can be a show stopper before you've turned on a single LCD segment.

The typical sleep current for the PIC16LF1904 is listed as 160nA @ 3V, I was measuring only 17nA. The same applied to the PIC16LF1934 where the datasheet quotes 80nA @ 3V, and I achieved 16nA right off the bat. It's times like this you question your own measurement skills. Replacing the PIC with a 10Mohm resistor gave expected results, allaying any fears of a gross operator error.

On the other side of the coin, the PIC16LF19175 (a recommended replacement for the PIC16LF1934) measured 275nA at 1.8V, but at 3V it was a very different story, with in excess of 3uA (spec says 350nA). It's not clear to me why this is, but after spending half an hour tweaking I put it to one side, and concentrated on the PIC16LF1904, and getting the LCD operating.
 
The following users thanked this post: mariush, thm_w, GeorgeOfTheJungle

Offline Kilrah

  • Supporter
  • ****
  • Posts: 1758
  • Country: ch
Re: EEVblog #1242 - Memory LCD+Supercaps+Low Power Design
« Reply #39 on: September 07, 2019, 07:23:48 am »
As usual, you should never trust the datasheet when it comes to low power current draw specs: sometimes it's better, sometimes it's worse, and often the difference is dramatic.

The typical sleep current for the PIC16LF1904 is listed as 160nA @ 3V, I was measuring only 17nA. The same applied to the PIC16LF1934 where the datasheet quotes 80nA @ 3V, and I achieved 16nA right off the bat.
Well your unit used 17nA but one from the next batch may use what the datasheet says, so you still need to expect and design for the worst case in datasheet figures to still result in a product working to the expected spec. You'll maybe (or rather likely) get lucky, but it would be annoying to expect to get lucky and get bitten manufacturing 1000 units and that reel of chips happens to be on the datasheet's max values...
 

Online GeorgeOfTheJungle

  • Super Contributor
  • ***
  • Posts: 1980
  • Country: pl
Re: EEVblog #1242 - Memory LCD+Supercaps+Low Power Design
« Reply #40 on: September 07, 2019, 10:25:42 am »
The best I managed so far at 3V with the display operating but with the CPU in sleep is on a PIC16LF1904, at 410nA plus 14nA per on-segment with the display attached, in static mode. This was using the internal LF oscillator, but no RTC functionality (I don't believe an RTC is a requirement in this case).

500nA@3V => 0.5e-6*3*24*365= 13 mWh/year
CR2032: 3V, 210mAh => 3*0.21= 630mWh

630/13 => would last 48 years in sleep mode, winner winner chicken dinner.

Now you only need to know the number of wakeup events per year, its runtime and its current :-) here's a spreadsheet:

https://docs.google.com/spreadsheets/d/1EC3Hb3dN89zZMW7C1lywiCgIFoLlzzcPUFokGHGDei4/edit?usp=sharing

Even when the experts all agree, they may well be mistaken.
 

Online Fungus

  • Super Contributor
  • ***
  • Posts: 10114
  • Country: 00
Re: EEVblog #1242 - Memory LCD+Supercaps+Low Power Design
« Reply #41 on: September 07, 2019, 10:29:33 am »
500nA@3V => 0.5e-6*3*24*365= 13 mWh/year
CR2032: 3V, 210mAh => 3*0.21= 630mWh

630/13 => would last 48 years in sleep mode, winner winner chicken dinner.

No need to argue over nanoamps. Once you get down into the single-digit microamp range, the battery's self-discharge starts to dominate.

Now you only need to know the number of wakeup events per year, its runtime and its current :-)

Yep.
 

Online Rerouter

  • Super Contributor
  • ***
  • Posts: 4370
  • Country: au
  • Question Everything... Except This Statement
Re: EEVblog #1242 - Memory LCD+Supercaps+Low Power Design
« Reply #42 on: September 07, 2019, 10:32:24 am »
don't forget to check the leakage current for any ceramic capacitors on your device, larger value ones can be really leaky, e.g. a 47uF cap can be 500nA,

Also the battery self discharge, its not much, but it exists
« Last Edit: September 07, 2019, 10:35:48 am by Rerouter »
 

Online GeorgeOfTheJungle

  • Super Contributor
  • ***
  • Posts: 1980
  • Country: pl
Re: EEVblog #1242 - Memory LCD+Supercaps+Low Power Design
« Reply #43 on: September 07, 2019, 10:51:41 am »
Howardlong, Silicon Labs have also got eight bitters, look : https://www.silabs.com/products/mcu/battery-operation (scroll down!)

« Last Edit: September 07, 2019, 12:58:18 pm by GeorgeOfTheJungle »
Even when the experts all agree, they may well be mistaken.
 

Online Fungus

  • Super Contributor
  • ***
  • Posts: 10114
  • Country: 00
Re: EEVblog #1242 - Memory LCD+Supercaps+Low Power Design
« Reply #44 on: September 07, 2019, 04:00:44 pm »
Howardlong, Silicon Labs have also got eight bitters, look : https://www.silabs.com/products/mcu/battery-operation (scroll down!)

Finding low power CPUs is easy, the problem is that Dave wanted one with LCD controller.

OTOH he has an LCD that can refresh all by itself so it's not really needed.

Me? I'd use a Tiny85 for this because it's what I know. It's an 8-pin chip so it's easy to work with and I've built stuff that runs for years on a coin cell using them. One gadget I built with a Tiny85 has been running continuously since March 2013 (using a 2540 cell, not a 2032) despite using the watchdog timer to wake up every 2 seconds to check a sensor.

The Mega328 used in Arduinos has a similar power consumption to the Tiny85 but a lot more pins. If I needed more pins I'd use one of those (probably a $2 Arduino Pro Mini with some components removed from the PCB to leave the bare chip).

If you know PIC or ARM or whatever? Use that instead.
« Last Edit: September 07, 2019, 04:10:58 pm by Fungus »
 

Offline Howardlong

  • Super Contributor
  • ***
  • Posts: 4755
  • Country: gb
Re: EEVblog #1242 - Memory LCD+Supercaps+Low Power Design
« Reply #45 on: September 07, 2019, 11:12:18 pm »
Here's my response video.



Code attached.

« Last Edit: September 07, 2019, 11:21:13 pm by Howardlong »
 
The following users thanked this post: thm_w, jaromir, GeorgeOfTheJungle

Online GeorgeOfTheJungle

  • Super Contributor
  • ***
  • Posts: 1980
  • Country: pl
Re: EEVblog #1242 - Memory LCD+Supercaps+Low Power Design
« Reply #46 on: September 08, 2019, 10:40:57 am »
Well done Howardlong, hats off to you !  :-+ :clap:

This padauk PMS154C may work for this? It's got an LCD driver.

https://lcsc.com/product-detail/PADAUK_PADAUK-Tech-PMS154C-S14_C168683.html

It's only 9 cents.

It doesn't seem to have flash/eeprom/NVRAM but, it isn't a must, is it? Unless you want to keep the count with no battery. And a quick battery swap doesn't necessarily mean a reset nor data loss, does it?

I can't read chinese (the datasheet) so I don't know for sure if it's got suitable low power sleep or wake up from sleep modes, but surely it does.
Even when the experts all agree, they may well be mistaken.
 

Offline mariush

  • Super Contributor
  • ***
  • Posts: 3818
  • Country: ro
  • .
Re: EEVblog #1242 - Memory LCD+Supercaps+Low Power Design
« Reply #47 on: September 08, 2019, 12:31:35 pm »
There's also lcd drivers that can be bought separately.
For example 30 cents for Holtek HT1621B :
soic https://lcsc.com/product-detail/LCD-Drivers_HT1621B_C7873.html
lqfp https://lcsc.com/product-detail/LCD-Drivers_HT1621B_C7532.html

or probably same chip, titan micro 1621, 15-20 cents in volume https://datasheet.lcsc.com/szlcsc/Shenzhen-Titan-Micro-Elec-TM1621_C41326.pdf

32 segments and 4 commons, min 2.4v , ~ 1mA at 3v
 

Offline nctnico

  • Super Contributor
  • ***
  • Posts: 17883
  • Country: nl
    • NCT Developments
Re: EEVblog #1242 - Memory LCD+Supercaps+Low Power Design
« Reply #48 on: September 08, 2019, 06:22:53 pm »
As usual, you should never trust the datasheet when it comes to low power current draw specs: sometimes it's better, sometimes it's worse, and often the difference is dramatic.

The typical sleep current for the PIC16LF1904 is listed as 160nA @ 3V, I was measuring only 17nA. The same applied to the PIC16LF1934 where the datasheet quotes 80nA @ 3V, and I achieved 16nA right off the bat.
Well your unit used 17nA but one from the next batch may use what the datasheet says, so you still need to expect and design for the worst case in datasheet figures to still result in a product working to the expected spec. You'll maybe (or rather likely) get lucky, but it would be annoying to expect to get lucky and get bitten manufacturing 1000 units and that reel of chips happens to be on the datasheet's max values...
I agree. The numbers from the datasheet should be used for anything that needs to go in production. The current consumption likely goes up with temperature as well so that is another factor.
There are small lies, big lies and then there is what is on the screen of your oscilloscope.
 

Online Fungus

  • Super Contributor
  • ***
  • Posts: 10114
  • Country: 00
Re: EEVblog #1242 - Memory LCD+Supercaps+Low Power Design
« Reply #49 on: September 08, 2019, 07:30:32 pm »
The current consumption likely goes up with temperature as well so that is another factor.

Sure, but not by an order of magnitude.

I imagine the datasheet was written for the first production samples and they've improved the process since then without changing the datasheet (in case there's any batches of old chips in a warehouse somewhere).
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf