Tech Note 096 - Wemos D1 Mini V3 Review
HTML-код
- Опубликовано: 14 май 2018
- This is the latest Wemos D1 Mini V3 board, it adds some useful features.
You can buy the Wemos D1 Mini from here:
Wemos D1 Mini V3: bit.ly/WemosD1MiniV3?p=W314022...
And more Arduino boards/accessories here:
More Arduino Kits: bit.ly/2Hb3WIG?p=W314022082674...
Wemos D1 Mini v3
WiFi ON 70.3mA
WiFi OFF 15.2mA
Sleep 0.24mA
Typical battery life at 1 event per hour = ((70.3*15 + 15.2*15 + 0.24*(3600-15-15))/3600 = 0.59mAHr and 2600/0.59/24 = 182 days 
Наука
Hi Dave. Nice video, but I hope you don’t mind some constructive feedback...
As a tech note I was expecting more technical detail (than just chip part numbers).
eg. Not just saying CH340G is replaced by CH340C with: “slightly different specifications” ??
Technically, the CH340C is simply a pin compatible variant of the CH340G, that doesn’t require an external crystal.
So in your summary, the 26Mhz Xtal does not “replace" the 12Mhz Xtal!
The 12Mhz Xtal referred to, was the CH340G Xtal (not needed for CH340C).
Technically, the ESP8266EX itself can use only a 24Mhz, 26Mhz or 40Mhz Xtal.
The ESP8266EX is simply the regular ESP8266 chip. So technically it doesn’t “replace” a ESP8266S (?).
The shielded ESP-12S module simply contains the same ESP8266EX chip, Xtal, and 4MB flash.
Finally, I believe the V3 now uses the ME6211 LDO regulator, replacing the original V2's RT9013. This also looks possibly a cost saving step, as the RT9013 had a lower typical quiescent current rating of 25uA (vs 60uA for 3.3v ME6211), and also the RT9013 has independent 600ma current limiter.
In summary: V3 basically adds mounting holes + GPIO16 deep sleep link, and (it appears) manufacturing cost savings! :-)
Hopefully the above technical points are helpful, and taken constructively (as intended).
Hi Greg, thanks for the feedback and welcome to my world! For every one message like yours I get 10 or more a day via my web page asking me to not include too much detail as ‘I don’t understand what you mean by...’. I did thoroughly research the difference between the different versions and did consider explaining the ch340 differences but then I have to consider audience retention time of about 2-min maximum so it all ends up getting lost, so from my viewpoint what I have to do is produce a video that appeals to a wide distribution of viewers of which the middle two quartiles are not of an electronic or computer background but enjoy micro-controller programming. Incidentally, the regulator is unchanged throughout the V1-3 evolution. In terms of terminology maybe I should have said the 12s module has been superseded by using only its component parts less the shielding, and by now I would have lost viewers. Conversely I think I provided enough information to those who do understand, for them to quickly determine what was different without buying a board themselves. Overall it was all about cost saving in a market that is flooded with fakes/copies and I understand why Wemos needed to do this plus it removes a lot of fakers who can’t fund the production costs of producing boards that don’t use the 12s and sell as ‘Wemos’ branded item. Finally, the channel is self-funded and thus I don’t get any free samples and rely on a small YT return of typically $10/month if I keep viewer numbers and ‘likes’ up, it’s actually costing me ~$500/year to fund component buys are other parts. You may see Andreas Spiess videos where he has discussed the same problem and whether it’s worth continuing and I’m in the same place as him. My vocation is the transfer of knowledge and information to as many people as I can whilst trying to do that through a request queue currently extending out to 50 more topics like how to use a digital compass, motion sensors, servos, pwm, interrupts, etc. Summary, I will continue to provide the funding for the channel for the foreseeable future but I do need to keep it under review and refine my content in a bid to meet the many demands (audience knowledge and language barriers, RUclips to name a few) and I hope I’ve given you enough of an insight as to why I say what I say, Regards Dave.
Hi Dave. Thanks for the comprehensive reply. I certainly understand your position.
I also agree on the fake Wemos boards issue, having been burned myself with some "LOLIN32 Lite" boards advertised as the genuine thing, only to receive unmarked (non branded) boards, which all had faulty TP4054 Li-Ion charger SOT23-5's.
Fortunately I was able to repair them myself with replacement TP4054. :-)
The official Wemos Store on ALiExpress now has the LOLIN32 Lite back in stock, although all the LOLIN32 boards are now flagged as "Retired". Their new LOLIN D32 appears to currently be their only ongoing ESP32 offering. A pity as I really liked the LOLIN32 Lite's low power battery optimised design (and lower cost).
It's certainly an understandable move for Wemos to rebrand products as LOLIN, and go to lower cost chip level design, so it’s less appealing for fake cloners to profitably replicate.
By the way, the LDO regulator was changed to ME6211 from the V2.2 D1 Mini onwards. Hence my reference to the original V2 RT9013.
For versions prior to V2.2 they used the RT9013, as confirmed by the schematic revisions.
Anyway, thanks again. I’ve given you a Like, and a Sub. :-)
Keep up the good work. YT video production is a time consuming exercise!
Enjoyed your video Dave. I think you have the right level of detail in all your videos. I'm not really interested in ultra-low level technical details and I think you have the right amount of balance. I still enjoy videos from many other micro-processor enthusiasts out there, but skip through the low level/granular stuff to get to the info I need. I can watch your videos from start to finish with no skips.
Thank you for your feedback, I appreciate it.
Another great video Dave. One key positive point of the wemos D1 mini is that, with a constant form factor, all the shields remain compatible with all board versions.
Yes and I think that’s why they have been and remain so successful, also the power consumption for this board is sufficiently low to make it a decent low power development board or to use on battery power for a good number of months, I posted the power consumed in these comments.
Thank you it is great to be able to identify your board by physically looking at it, sorely needed. Helpful in purchasing unless they use the incorrect board pic in their advertising.
It happens a lot with AliExpress as a lot of suppliers were showing pictures of say a v2.3 board but sending out a v1 or 2.2, so a few months ago I spent a long time trying to get a full set, but AliExpress were good as I got a refund for each incorrectly listed board. Unfortunately Wemos do-not maintain a gallery of their older boards to be sure. For now you should only be getting the V3 boards all others are old stock and often more expensive than the new V3 variant.
Very interesting. Thanks
Super Video! :-)
See video notes for the power consumption results
hi, I encountered such a problem, I have 20 wemos d1 mini v3.0.0 boards, but none of them find my networks in client mode, and no device (laptop, smartphone) sees wemos in AP mode. but they see each other (1 wemos create point - 1 wemos scan). Tell me how to deal with such a problem. Has anyone come across?
Thank you very much for your kindness and knowledge and experience we have learned a lot ,
I am planning on using D1 Mini and powering it with 220VAC to DC (HLK-PM01) and connect through to the 5v , or (HLK-PM03) and connect through to 3.3v , which is better ? and do i need any other component ? i.e. capacitor .. etc
For all round ease I would use 5v supply to the on board regulator so connect that to the 5v pin. You can power the board with 3.3v via the 3v pin but it places the regulator and UART on the load. If your planning to use any peripherals these will generally use 3.3v but will nearly always have an on-board regulator so again 5v wins.
Thank you very much , do I need a capacitor with GND or is the board already have one ?
Solar for Green The board has a voltage regulator and full smoothIng/storage capacitors. Just a reasonably smooth power supply like any usb type.
mine has a small led on pin A0
Great video. Have you got the new esp32 Lolin D32? I got mine but haven’t tested yet..there is less pins
Yes got one, I'll try to do a video soon on it. Out performs the original and a surprise feature!
Have you tried to see what the current consumption in deep sleep mode is?
Good question, I should have included these results in the video:
Wemos D1 Mini v3
WiFi ON 70.3mA
WiFi OFF 15.2mA
Sleep 0.24mA
Typical battery life at 1 event per hour = ((70.3*15 + 15.2*15 + 0.24*(3600-15-15))/3600 = 0.59mAHr and 2600/0.59/24 = 182 days
What is the difference between the 3.0 and 3.1 boards? Just the branding change from WeMos to LoLin??
According the to schematic no change and yes just branding to LOLIN like the ESP32 boards too.
is FCC really needed for 2.4 Ghz applications? Thanks Dave
Well it's a vague point, in that imports to the US probably need FCC certification, but who checks! And for most applications RFI noise is not an issue, so agreed not required. Generally most developers and hobbyists should not see this omission as essential. Overall it's been about cost reduction IMO.
You still have to get FCC certification on your final product anyway, even if you use the certified modules. Using a certified module does not mean you don't have to go through compliance testing for the final product, it just means it's shorter (you only have to do unintentional radiator testing I think). It's not relevant for hobbyist devices either way, you're allowed to build up to five of the same device without having to go through testing. Devices sold as a "sub assembly" (like these dev boards) also don't have to go through compliance testing. www.sparkfun.com/tutorials/398
Is there any difference in power draw?
Yes, the mini is not good, the best are the Wemos D32 (70uA) or DFRobot Firebeetle (11uA) in sleep mode.
The "D1 Mini Pro" does not have the Antenna printed on it ? Can this be connected to the WiFi router ?
Yes it has WiFi on-board and can connect to a WiFi router
@@G6EJD Thanks. I found about by uploading a program too.
@@G6EJD I want to reset the wemos D1 mini / Wemos D1 R1 externally just like how I could do it with the Arduino ( i.e connecting the RST and the GND with a tactile switch or push button switch )
I am trying to do the same on Wemos and it does not work !
Is there another way of doing it ?
All I want to do is push a button on the panel and the microcontroller resets itself.
( the reset switch is too small and my microcontroller is enclosed in a box, hard to poke in for a reset )
Appreciate a reply if you could.
and now version 3.1.0 ! Any idea what the very quick update is for Dave? wiki.wemos.cc/products:d1:d1_mini
It's the same schematic I think this is about branding it to Lolin they have dropped wemos.
Interesting to see the evolution photos. Beware when you buy these "cheap" 8266's since there seems to float around lot's of WEMOS/LOLIN knock-offs designs with various versions. Often the brand name and version is removed from the silkscreen. I found that I have bought V1, not exactly what I intended when purchase it. Buy the original.
Sgt-Bond, good advice, only the AliExpress shop for Lolin can ensure a genuine product, a little more expensive but they work reliably and consistently.
Can not be the only one reporting this failure of the WiFi connection on the Wemos and NodeMCU 8266 breakout boards. It first appeared sometime back while testing long-term use. Every morning would check the WiFi connection and for about a couple of weeks, all was fine
I was surprised to see the wemos mini had lost connection. This started happening on a regular basis. Resetting the device brought it back online.
So loaded the same sketch onto the NodeMCU and ran it for some time and eventually the same thing started happening.
Decided to compare the Wifi source so I removed the Asus Rt-AC87u replaced it with Google WiFi hardware. For about a month no problems.
Then all of a sudden started having dropped wifi all over again.
The distance from the wifi source does not come into play I have two outside and two devices inside. The one located inside is very close to the wifi source.
The devices failing have to do with a sketch I run for manual sprinkler control. Meaning a press of the button on the webpage turns on the sprinklers and 15 minutes later turns them off.
Here is the interesting thing about all this if I run the 8266 with "your" Temp/Humidity sketch this does not happen. This device never fails to be available via wifi and using the web interface you showcased herein. BTW the one you posted and the one you modified for me where celsius was changed to Fahrenheit was used.
Here if you are interested is the failing sketch I am using.
pastebin.com/qBy5nkFF
Well in instances like this I would suspect the router and its IP lease time, In nearly every type made (the core software is usually the same open source types) they are set to 24-hours maximum, and to make it a permanent lease set tine to 0 usually.
You should not need a connected server only a Cheney that enables a connection when required. I’ll have a look at your code tomorrow and see if there is any obvious problems, but a cursory look revealed none. The ESP8366 does not have any connection issues or people would be raising the issues. I’ve got 2 ESP8266 servers that have run without error 24/7 for close on 2-years now, just checked and working as expected, so it may be a programme logic issue. Can you monitor the serial port and put lots of diagnostics print statements to see where it is going wrong. I’ll let you know what I can see.
Thank you. To combat the lease issues I have reserved the IP addresses assigned to the devices in question. Also, this does not explain why your Temp/humidity sketch does not drop the Wifi connection. It has run for months on end and on the very same devices that fail on my sketch. Thank you for taking the time to review the code.
I've been through your code twice and can't see a single problem with it. My observations are choosing an IP address in the lower numbers is more likely where the often less than perfect Domain Name Server processes tend to assign IP addresses, usually from 192.168.68.2 upwards and as you've chosen 10 there may well be the odd conflict with another device, also with some Routers you can specifically reserve an IP address for a particular device, on mine it has to be done as IP and MAC address. So what I'd do is move the IP up to about 150 or 200, plus change any associated port forwarding rules so it can still be accessed externally. Other things you can do is switch on your Router log and then see what it says when that IP address is disconnected. Not much else I can find, the code has all the right functions in it I'd expect.
First off thank you for your response. How I am handling the IP address for this device and other test devices is assigning it an IP address. Then I go into the router and reserve the IP address linking it to the MAC address.
The next step will be based on your suggestion assign the devices in questions with 192.168.86.150 and above.
As for the port forwarding as I understand it is only used if you want to connect to a device from outside your LAN. That is inward pointing connection. In this case, I do not allow any connections using the inbound route. So no ports are configured as so. Unless I misunderstood your meaning.
Lastly will check the router log settings and set them up accordingly.
Again Thank you.
OK, I only mentioned the port forwarding in-case you accessed it from the WAN/Internet. The log file should tell you what's happening at the router at least.