For a month I’ve had Boron LTE and competitor both uploading over cellular at my same remote site.
For September, for the Boron LTE, I uploaded about 20mb of data. (Frustratingly it appears the console doesn’t let me see past usage). This was apparently free because of a 3-month free trial, but it will cost $3 + 17*.4 = $9.8/month afterwards
For September, for the competitor, I uploaded about 15mb of data. This cost me only $4, less than half the eventual Particle cost, and even that seems to be a rounded-up billing.
The competitor charges $3/month for 5mb and then $0.10/mb beyond that. However confusingly my bill from the competitor actually shows even better (a $1 refill gets me “20mb” of data, which is only $0.05/mb!)
I like Particle and would stick around, but how can Particle expect me to do so when their LTE as actually implemented side-by-side in the field is far more expensive?
My sensors must without question upload 20-25mb per month and that is non-negotiable.
Does anybody know if Particle has intentions to lower the price to match competition? Thank you.
(BTW - the competitor’s product uses a BG96 modem with 2G fallback from LTE, and has had better up-time than the Boron despite being mounted at this particular site with a far worse external antenna. The Boron at this site has a $200 extreme quality marine grade high-gain cell antenna and currently has been offline since October 8th, whereas the competitor is currently faithfully connected and uploading).
For self-service customers, the price is fixed and currently dependent on the country/region the SIM is used. There is no price match available.
For enterprise customers there is more flexibility in the cost, including the possibility of larger per-device data allowances, but still no blanket match for any competitor’s charges.
The only small downside for me to the competitor is you can’t write in C but must use their Python-esque scripting language. And you pay $125 per unit instead of $60. Virtually everything else is unfortunately better than Particle. They even have a 5V step up regulator from LiPo.
@RWB, I think you meant 23 minutes. The “competition” (Pycomm I believe) has many reported problems as well, as reported by members who switched to Particle for that reason. Neither platform is a panacea and neither is without its challenges. However, Particle is better setup for managing a fleet of devices IMO, with highly reliable OTA and device monitoring. The cost per MB is not always the only factor to look at unless you are considering a small volume of devices. Nonetheless, I believe you need to chose the platform that suits your requirements.
Not Pycom, which is terrible and highly power consumptive. I don’t know what the keep alive is but clearly it is not costing me more cellular. It sucks because I love Particle and wish Particle had the better product.
@Mjones I take rickkas7’s honest response that price-matching is not offered as a fair indication my professional politeness of initially anonymizing the competitor is, given free market principles, no longer within the requirements of etiquette.
ElectricImp’s imp006 breakout board is what I am evaluating against the Particle Boron.
I came from Pycom in May 2019 as a highly disappointed GPy customer. The Boron was so superior in terms of power consumption and LTE stabilty.
But the Boron still has greatly disappointed me in long-term LTE stability (see my threads) except for the few sites/stations I have where I happened to get “lucky”, where the Boron is performing flawlessly (after painful hours of driving to resolve issues that shouldn’t have been), but the LTE data cost is still 4 or 8 times more expensive than impCellular depending on which understanding of mine is correct. (They advertise $0.1/mb but it actually seems from my bill it’s half that)
The advantages for me of the Boron are:
Can program in C and have support for almost all Arduino libraries (decent)
Smaller/featherwing form factor (minor)
More info on forums (minor)
Can use as I2C slave device whereas imp006 is master-only (minor)
$60 per unit (cheaper) rather than $125 per unit (imp006 breakout)
The advantages for me of the imp006 are:
Radically cheaper cellular (big)
Radically more stable LTE with 2G fallback (huge and massive - literally the most important given the frustrations I’ve had)
More stable electronics/not as “sensitive” as the Boron (big)
Can run straight from very high voltages - e.g. 12V SLA - built in regulator (decent)
LiPo connector with 5V step-up and charging that I assume actually works in the solar use case (decent)
Built-in humidity/temp/accellerometer (minor)
Way more RAM and I believe way more non-volatile storage (big)
A library called “ReplayMessenger” which locally stores/persists my data uploads and reuploads them in case of disconnect/reconnect (decent)
Built-in WiFi capability (minor)
A TON of extra UARTs vs. having zero extra UARTs beyond the first on the Boron (huge)
Both the Boron and the imp006 breakout have the horrendous, evil “u.Fl” flimsy nonsense connectors instead of a proper SMA connector, so that factor of common irrational insanity which has permeated this industry cancels out between the two.
Now the E402 might be a better Particle product for me, but the $0.40/month/mb for me is a sticking point.
Wondering if there are things I am missing if someone could critique my comparison from a pro-Particle viewpoint.
I just recently saw the Cellular Electric Imp board on Digikey or some other electronics place recently.
I remember @hfiennes messaging me saying they were working on a cellular product. I guess this is what he was talking about. It was awhile back.
It saw it ran on Python and I instantly stopped looking into it. I don’t do Python.
Being able to get C++ code and libraries for just about any sensor or chip on the market is priceless to me.
As far as cellular pricing being a tad bit higher with Particle I just plan having the end customer pay for what ever data amount they want. If I get big enough I’ll negotiate better cellular rates.
Reliability is of course of most importance. I have not had reliability issues with Particles or Losant’s cloud platform so I’ve had no reason to look elsewhere.
The only thing I wish particle did was build a board using the ESP32 that has Wifi + BLE plus tons more flash and memory with a lower cost.
Here is may take (and only MY take) on these points:
Perhaps and I know Particle is constantly working to lower their costs though their focus is on Enterprise volumes not one-zies.
Particle is working on stability with the impending release of 2.0.0 LTS (long term support). They specifically chose NOT to have 2G fallback as it requires a large amount of current, negating the low power design. Besides, 2G is getting deprecated everywhere as we speak.
Not sure what you mean by "sensitive". However, you should cite specific examples in a support ticket to have Particle address them.
There is no 5v step-up since the logic on the Boron is 3.3v. There are LOTS of threads in this forum on how to configure the Boron PMIC for solar charging, especially using the newest API provided in the DeviceOS 2.0.0-rc.x versions.
As you said, this is minor
I couldn't find the specs but I will assume yes, I agree and this has been a complaint for years. However, don't be fooled as MicroPython uses a lot of memory that requires frequent garbage collection which can introduce substantial delays in your code if not handled properly. Personally, I wish Particle devices has twice the flash and RAM and they would be perfect. But for now,
That is available with @rickkas7's PublishQueueAsyncRK superb library which even allows you to store the data in external or internal flash.
Yup, minor though it would be nice.
Yup, it would be nice to have extra UARTs though these can be power hogs as well. These can be added externally using an I2C-to-UART bridge.
The imp006 is a dev board which wasn't designed to go into a product which the Boron could (though the featherboard footprint is not ideal). This is why they have the M.2 modules as other production choices.
The uFl is a fact of life given the board real-estate used by a SMA connector.
Perhaps you can explain you use case (eg. high or low volume) and why the data cost is so critical. In a low volume application (< 100) I can understand but with larger volumes, rates can be negotiated with Particle.
To be fair, modules are programmed in Squirrel (syntax close enough to C/C++). In practice I have experienced memory as more roomy on their platform (and no garbage collection issues).
On the original topic, 3MB vs 5MB in itself, is obviously a major difference.
This is true, I use AT&T’s one rate sim’s. $30 / year for 2TB, with a 64K stream cap. I don’t think Particle is going to last though, given the outages and their pricing in the industry, their are far better solutions, maybe not as quick to market, but …
