Onethinx Core Module

We’re still working on the low-power modes and provide an update asap! If possible we want to use hibernate mode which puts the micro into sub uA current consumption.

We will probably provide the following APIs: LoRaWAN_Sleep(), LoRaWAN_DeepSleep() and LoRaWAN_Hibernate().

1 Like

@Rolf1 The low-power mode API you describe sounds very promising to me. Really looking forward to seeing this implemented as I had planned to use the module for battery powered nodes.

In the meanwhile I could get the current consumption of my dev kit down to 678 µA by disabling the LEDs and entering hibernate mode inbetween transmissions like this:

Cy_SysPm_SetHibWakeupSource(CY_SYSPM_HIBERNATE_RTC_ALARM); Cy_SysPm_Hibernate();

But I guess with this I only hibernate the user core but not the core where the LoRaWAN stack runs? :thinking:

onethinx-module-hibernated-current

That’s correct. Unfortunately the stack on the M0+ currently doesn’t have the possibility to get into any of the low power modes.

At the moment we’re working hard on these power modes and we’ll provide updates asap.

btw: by measuring current using the header connections, the ground current of the LDO will be added to the readings. We provided the testpoint TP_M1 and TP_M2 to measure the current through the module only (don’t forget to remove the R6 zero ohms bridge).

I’ve added the schematics for your convenience: OT_LoRaWAN_Kit_01.pdf (514.8 KB)

1 Like

Thank you for the hint with the testpoints. I removed R6 and measured current through TP_M1 and TP_M2 now. Surprisingly readings at my multimeter are slightly higher than before. Maybe I need better measurement equipment? Any recommendations?

Probably the current through the LDO is very small (1uA or so), so you currently won’t see the difference.

Spreading in measurements can currently be due to the difference of temperature or just measurement accuracy (repeatability).

Mine arrived yesterday - now just need to find time to get to grips with it! Oh, and forgot to order a CR2450 Battery so slight delay there too! :frowning:image

Might delay until low power libs/Api released…after as quick test of course!..:wink:

1 Like

Current status of the PSoC6 hibernate current (without SX1261).

Sub microamp!! :ok_hand::ok_hand:

Goal reached. We’re continuing to work on the stack and prepare the low power API function calls for next release.

Comparison characteristics:
ARM Cortex M4: 3.40 CoreMark/MHz
Power consumption: 40uA/MHz
Flash: 1MB
RAM: 256KB

image image

:open_mouth::clap::ok_hand:

Batches of 2450 batteries finally delivered yesterday…perhaps get play time scheduled in for next week now! :wink:
image
:thinking: if power consumption as low as suggested and shown above then looks like many years worth!..might need to get a few more modules if initial results and experience ok…otherwise internal self discharge likley to render batteries redundant before consumed :rofl:

2 Likes

This is my temporary enclosure :slight_smile:

IMG_2696

2 Likes

Mine is ordered and I cant wait !!

Looking to develop with the mbed OS, should be exciting.

Serge

anyone have one in a use case … experiences ? :roll_eyes:

Still working on it, basically I am waiting for two things now. MBED support in terms of being able to flash firmware, I can compile already but because of the trickery of using one core for the Lorawan stack and locking some memory zones flashing is not straight forward.

Second is the low power API.

I know Rolf and his team are working on this.

1 Like

in fact, this is crucial.

1 Like

Thanks for your concern and yes, we’re working hard to get low power finalized.

I like to share the latest updates:

  • Yesterday we fixed a tough error which was bothering us for more than a week. it was caused by (undocumented) stackpointer problems. The ones who already have our module will notice we used a PSoC6 ES3 (engineering sample). Still facing the fact that we’re early PSoC6 adopters :face_with_raised_eyebrow:
  • Previous week we got testresults from the EMC testhouse which made us really happy: spurious emmisions are well below limits (-15dB worst case!) and antenna radiation pattern is pretty ‘isotropic’ which means we can use maximum power while receiving well in every direction.
  • We fixed some programming issues related to ModusToolBox 1.1 in combination with the ES3 silicon.
  • We worked on the US915 stack and sending out lots of modules this week to the US.
  • We’re in contact with the Cypress / Mbed OS team to see if we are able to provide a Mbed working on our module soon.

And last but not least, we’re probing to see if there’s interest to set up a beta-testers program. Drop me an email when interested (rolf at onethinx dot com).

3 Likes

:rocket: We got some exciting news for you! :rocket:

Onethinx is now launching a 𝗖𝗟𝗢𝗦𝗘𝗗 𝗕𝗘𝗧𝗔 of our Onethinx Core LoRaWAN module. As you already might know the Onethinx Core is an unique LoRaWAN module, which excels in security, ease of use and reduces time-to-market for developers. It uses (as one of the few modules in the current market) Semtech’s latest LoRaWAN chipset (SX126x) and Cypress’ PSoC6 MCU which is tailored for IoT. It offers a LoRaWAN stack which is locked down, Bluetooth BLE 5.0 and an high efficient integrated antenna.

Our goal is to deliver significant advantages to the developer by providing a LoRaWAN module which is extremely easy to use, while providing a truly secure, advanced and robust platform for LoRaWAN IoT applications.

Therefor we need a group of people from the community who will be eager to test our product, share feedback and help us bring the Onethinx Core module to the next level of world’s coolest LoRaWAN module.

We offer you the Onethinx Core module with the beta firmware mounted on our development kit for free in return for your valuable contribution.

Drop us a mail at info(at)onethinx(dot)com and join the leading LoRaWAN developer team.

Be quick as the program closes 14th of June!

1 Like

any results from the closed beta ? :sunglasses:

Yes, we now have a nice team looking forward to get hands-on with our module.

Still working on the latest features though but it’s gonna be a pretty cool update!

We did extensive testing on a lot of networks and gateways, still super stable since the beginning: immediate join on all SFs even over 14 kilometers (@SF12) :grinning:

1 Like

Is this different to the performance of an SX1278 ?

The SX1262 can go to 22dBm, but sensitivity is listed as the same as SX1278.

1 Like

:rocket: Update :rocket:

The low power version of the stack is currently available together with lots of new features!

To mention some:

  • several sleep modes, easy configurable for wake-up on time, delay and external wakeup IO
  • ULTRA low hibernate power (410nA without RTC and 600nA with RTC @ 1.8V)
  • FlashWrite and FlashRead simplified API
  • Setting preferred US subband channels
  • Enhanced join procedure for fast join (optimized datarate / channel polling algorithm)

Take a look at our GitHub page for using the newer version of the module with the latest M4 API.

2 Likes