Anyone interested in discussing building DIY Home Automation software and hardware?

[NeverDie]

Of perhaps greater interest to you isn't that, but here (http://www.silabs.com/products/wireless/EZRadioPRO/Pages/si446x.aspx?tab=info), where it says "The Si4468/3 offers exceptional output power of up to +20 dBm with outstanding efficiency. The high output power and sensitivity results in an industry leading link budget of 153 dB allowing extended ranges and highly robust communication links."  It's not 157, but still impressive, no?  Perhaps more importantly, available on a wider range of less expensive modules.

 
It appears that Silicon Labs acquired Energy Micro (maker of EFM32) in 2013.  Silicon Labs' new EZR32 "wireless MCU" appears to integrate the above radio with an a EFM32  (an ARM Cortex M03 or M04 micro-controller).  Digikey sells the chip for  about $5 for quantity 1.I guess I'll wait for some kind of affordable board or module to become available.

 

[tattema]

It was a really smart purchase because energy micro was the only ARM fabricator that architected an ARM based chip designed specifically for low power consumption. While the other companies talk about low power, I would be pleased to know about any device that offers the same level of fine grained power management and control as the EFM32 range.
 
That EZR32 range looks awesome but I'm disappointed with the radio front end and the details they've provided. Unlike Atmel and others who provide free mesh and network stacks, Silicon Labs appear to be lagging hence their involvement in the Thread initiative.  I'm with you on the affordable board aspect. Their evaluation kit is a tad expensive for my budget so perhaps somebody should fabricate an affordable board specifically for the DIY HA space. It's not something I'd do on my own but if others on this site were interested and could pool resources, perhaps we could make some of these devices.

 

[NeverDie]

When you outgrow the LPC810 device and want something different, check out the mbed API's. It's like Arduino for ARM 32 but a whole lot better. They offer an on-line compiler so that you can just build programs in the cloud and then download them to your device over USB. No complex tool chains required coupled with huge community support fo sensors means it's a DIY HA place in heaven. If you then decide you want more control, they give you the source code too so you're not locked into their model. They have a HUGE array of supported devices as you can see here: https://developer.mbed.org/platforms/

Thanks!  Because mbed works across so many different modules, it looks as though it will relieve the pressure I was feeling of having  to pick the "right" module family before even getting started.  Compared to the Arduino IDE (which is what I'm currently familiar with), how easy/hard is it to learn and use mbed?

 

[tattema]

Hi NeverDie
 
Compared to Arduino, mbed is probably at the same level of complexity and supports or the usual peripherals... interrupts, io, motor controls, sensor interfaces so I would be too concerned about it. I began with Arduino but have been happy using mbed. 
 
I've not used ST devices nor the Spark's Photon Wi-Fi but from what i've read, they look like amazing technologies, very comprehensive and potentially complicated. Many of the mbed devices are based on st devices too so there's overlap in that space.
 
There are ways to measure very low current consumption levels without the EFM32 board. As you suggested, use an oscilloscope to record the voltage drop across a 1ohm resistor (ohms law) or use a ucurrent wit an oscilloscope sold by an Aussie who runs the eevblog. http://www.eevblog.com/projects/ucurrent/. It can be used with a multimeter because it converts current consumption to voltage levels and works best with an oscilloscope (not required).
 
I used and had a lot of success with the ucurrent device but recently let the magic smoke out by connecting it around the wrong way. 
 
I'll be interested to know your impressions re the LPC810 when it arrives.
 
Regards
Trev

 

[NeverDie]

Hi NeverDie
 
Compared to Arduino, mbed is probably at the same level of complexity and supports or the usual peripherals... interrupts, io, motor controls, sensor interfaces so I would be too concerned about it. I began with Arduino but have been happy using mbed. 
 
I've not used ST devices nor the Spark's Photon Wi-Fi but from what i've read, they look like amazing technologies, very comprehensive and potentially complicated. Many of the mbed devices are based on st devices too so there's overlap in that space.
 
There are ways to measure very low current consumption levels without the EFM32 board. As you suggested, use an oscilloscope to record the voltage drop across a 1ohm resistor (ohms law) or use a ucurrent wit an oscilloscope sold by an Aussie who runs the eevblog. http://www.eevblog.com/projects/ucurrent/. It can be used with a multimeter because it converts current consumption to voltage levels and works best with an oscilloscope (not required).
 
I used and had a lot of success with the ucurrent device but recently let the magic smoke out by connecting it around the wrong way. 
 
I'll be interested to know your impressions re the LPC810 when it arrives.
 
Regards
Trev

I took a stab just now at using the uCurrent and an oscilliscope to measure the current on an NRF24L01+, but there was too much noise.  The 1 ohm sense resister approach had less noise, but still too much noise.  I suspect long wires are contributing to the problem.  I'm inclined to think some kind of onboard measurement by the arduino itself is the way to go, as then the wire lengths are short.  Not sure how others are doing it, but so far it seems a lot easier said than done.

 

[tattema]

That's disappointed for you. I don't get any noticable noise measuring voltage drop with my scope so without actually seeing the setup, I'm not sure what to say.
 
You could try putting balun filters on the VDD into the NRF24L01 to knock out high frequencies. If you haven't already tried, switch on the scope bandwidth limiting function. 
 
I'm surprised that there's more noise with the ucurrent given that the voltage swing on the outputs is almost 1v whereas the voltage drop across 1 ohm is in the millivolt range.
 
I hope you can solve it because I'm keen to see what you discover regarding current consumption on the NRF chips.

 

[NeverDie]

I'm surprised that there's more noise with the ucurrent given that the voltage swing on the outputs is almost 1v whereas the voltage drop across 1 ohm is in the millivolt range.

Wait, how do I  get a 1 volt swing on the output of the uCurrent?  It must be something blindingly obvious that I'm not seeing.  The uCurrent came with absolutely zero instructions, but looking at the specifications it says:
.  • 3 Current ranges:
o +/- 0-300mA (70µV / mA burden voltage typical)
o +/- 0-1000µA (10µV / uA burden voltage)
o +/- 0-1000nA (10µV / nA burden voltage)
• Output Voltage Units:
o 1mV/mA
o 1mV/µA
o 1mV/nA
 
Because I'm expecting the current during transmit will likely be in the 10-20ma range, I have the uCurrent set to the milliamp range, which gives just 1mv output per milliamp of current, and so it would seem I'm no better off (aside from the burden voltage issue) than using a 1ohm sense resistor without a uCurrent.
 
This is literally the first time I'm using the uCurrent.  Am I doing it wrong?  If I were to put it on the microamp range, then wouldn't the output voltage saturate and not give me a true measurement (which would require an output of 10-20V,which the uCurrent can't do)?

 

[tattema]

Mine also arrived without instructions and it was challenging finding information about it. Annoying because its not cheap. Here's something from the maker: http://www.eevblog.com/projects/ucurrent/
 
I sit corrected, the voltage swing on mine was in the mv range, not v range so I don't think you're doing anything wrong from what you say. I was using mine in a similar way until i connected the inputs to the outputs. grrr.
 
The ucurrent has no internal protections so setting it to the ua range with ma loads or greater could damage it. 
 
Can you describe the noise you see on your scope? Frequency? amplitude? voltage?

 

[NeverDie]

Mine also arrived without instructions and it was challenging finding information about it. Annoying because its not cheap. Here's something from the maker: http://www.eevblog.com/projects/ucurrent/
 
I sit corrected, the voltage swing on mine was in the mv range, not v range so I don't think you're doing anything wrong from what you say. I was using mine in a similar way until i connected the inputs to the outputs. grrr.
 
The ucurrent has no internal protections so setting it to the ua range with ma loads or greater could damage it. 
 
Can you describe the noise you see on your scope? Frequency? amplitude? voltage?

Thanks for the link!  The uCurrent would have been better if it had more ranges, so it could amplify in the manner you were thinking.  Also, I don't think Dave Jones ever says what will happen if you exceed the currents in each range: is it destroyed or does it just not measure well at the higher currents?  It would be good to know, in case I want to measure a mix of high and low currents on the lower range setting.
 
I'll see if I can take a picture of what's on the oscilloscope screen.  It seems fancy enough that it should be able to do that.  Maybe the "ultravision" will help with that.  It's a fresh out of the box Rigol DS1054, my very first oscilloscope experience.
 
On my first attempt last night I used an Arduino Uno to drive the NRF24L01+.  I eventually noticed that even without plugging the USB cable into my computer that the USB cable was a big source of noise when plugged into the Uno. When I did plug the USB cable into my computer, the noise got much worse. So, the USB cable has got to go.
 
New plan for the second attempt::

1. Use an RFToy, instead of Uno, because it's a more compact way to plug-in the NRF24L01+ without having wires dangling all over, and because it can run off a coin cell battery, again without wires.
2. Plug the NRF24L01+ into "tall" female header pins, and then plug those into the usual 2x4 header on the RFToy. Why? Because the leg of the tall VCC header pin I'm going to snip and (hopefully) replace with a 1 ohm sense resister. With luck there will be enough room on either side of the inline sense resister to connect up the oscilloscope and thereby do measurements without introducing long wires that might pick up noise.

If that's too tight to make it all fit, then I guess I'll have to build a little adapter board to accomplish the same thing. That seems like more effort though.  I hope I can find some tall female headers laying about somewhere....  

 

[tattema]

Firstly, congratulations on buying the Rigol DS1054. In my opinion, it's value for money cannot be beaten! I have a similar scope from the same manufacturer that I hacked so that its bandwidth was boosted to 300MHz instead of the 75 it came with. Rigol manufacture exactly the same probe front end circuitry for all the DS1000 series and DS2000 series so a simple firmware upgrade has been known to convert them into more expensive/powerful models.
 
To measure current, I change the acquisition mode on my DS to HighRes. (Acquire button then Acquisition then HighRes). I turn on Anti-Aliasing from the same menu. You want to make sure that your channel, in my case CH1, is set to DC coupling and that BW Limit is on.
 
With my ucurrent, I exceeded the current ratings per setting by a small scale and it seemed to survive but as expected, plugging power directly into the inputs caused smoke to be released and it ceased working. The internal op-amps have been destroyed and I've been meaning to buy a set with my next digikey order but haven't done it yet.
 
I think your idea to eliminate noise sources by running off a battery is excellent. That way you'll avoid another potential problem that can fry small circuits, namely, is the voltage difference between the oscilloscope earth on the probe and the power-supply ground pin. They should be the same voltage (0) relative to each other but if not, the potential difference may destroy small circuits, or worse, cause the probe to explode for higher currents. A simple test is to check the voltage across the oscilloscope ground and your circuit ground with a meter. If it reads near 0v then you're safe. Cheap wallwarts, specifically Chinese SMP's are renowned for noise and misreferenced 0v outputs.
 
Good luck sorting it it.

 

[NeverDie]

To measure current, I change the acquisition mode on my DS to HighRes. (Acquire button then Acquisition then HighRes). I turn on Anti-Aliasing from the same menu. You want to make sure that your channel, in my case CH1, is set to DC coupling and that BW Limit is on.

Thanks, that gives me an idea of what to aim for.  I didn't get that evolved just yet on taking the measurement, but I did manage to eliminate enough of the noise from my first attempt that I may have gotten some useable results.  I posted them starting here:
http://forum.mysensors.org/topic/1664/which-are-the-best-nrf24l01-modules/108
and below it I show how I did the setup to take the measurement.

 

[tattema]

Thanks, that gives me an idea of what to aim for.  I didn't get that evolved just yet on taking the measurement, but I did manage to eliminate enough of the noise from my first attempt that I may have gotten some useable results.  I posted them starting here:
http://forum.mysensors.org/topic/1664/which-are-the-best-nrf24l01-modules/108
and below it I show how I did the setup to take the measurement.

 
 
Hi NeverDie, this video about noise and digital oscilloscopes might be helpful. https://www.youtube.com/watch?v=Znwp0pK8Tzk
 
Thanks for the link. I enjoyed reading the analysis and investigation you're doing on the mysensors forum. Of concern are the fake NRF modules on sale. The one with the potted chip on board appears to be missing a PI filter on the output stage so I'd hate to see what spurious harmonics and emissions in the 2.4ghz range are being transmitted, especially if the chip is a clone and has dubious operational parameters.
 
Of interest re cloned NRF chips.. http://zeptobars.ru/en/read/Nordic-NRF24L01P-SI24R1-real-fake-copy
 
I have one of those 'fake' boards so I'm inclined to wire it up and help verify the current consumption. Nice work BTW with your boards and in-line resistor.
 
I must admit that I'm also a bit impressed that somebody was able to clone the chip!! It reminds me of a story I'd like to share about counterfeit products in China. A couple of years ago I was in Guangzhou heading back to the hotel from a meeting when I was approached by a street vendor selling Samsung clones. Insistent that they were original 'copies', I purchased one to tear down and mess with later on. Sure enough, it was a complete fake. The ARM cpu was a rock chip device but the phone looked identical down to the packaging. The display was high resolution, retinal level, and the device looked and functioned like an original Samsung. Back in Australia it wouldn't register on the network as the IMEI number wasn't recognised so I managed to flash it with an old nokia IMEI number that was laying around and it worked perfectly, for a few weeks, util the batteries packed it in.

 

[pete_c]

I got bit a couple of years ago relating to those tiny USB to NIC devices sold as Apple USB to NIC devices with a look identical to the real McCoy.  The description mentioned that it was an Asix device and included Asix network drivers which really did work.
 
Here I was using them for some little projects mostly removing the board and cutting the wires to solder them to a motherboard.  Mostly just purchased them for their tiny size.  I did purchase them in bulk of 10-20 at a time really cheap. (less than $1 USD each purchased in bulk)
 
After using / soldering up a few of these noticed that the network performance was way less than stellar and I wasn't really getting much more than 5Mbps per device.  Later did read that they were Asix clones and being sold as Asix devices with copied disfunctional chips.  I did build a few of the devices using the clones and never really paid attention other than the slow performance.  I did find geniuine Asix USB to NIC devices a bit later and was able to purchase them sans a case which is what I wanted.  It was though a PITA to unsolder all the clone devices and replace the USB to NIC devices with the genuine devices.  I did also make a deal of of this with Ebay and got my money back on the last purchase of the clones.

 

[NeverDie]

I must admit that I'm also a bit impressed that somebody was able to clone the chip!! 

 
That was my first reaction as well.  Then what next blew me away was finding out  there are at least a dozen different known clones of it: http://www.testandmeasurementtips.com/embedded/pocketpico-p200-powermate/
plus who knows how many other clones out there  "in the wild" as it were.
 
The original motivation for that mysensors thread was me being disappointed in some NRF24L01+ modules that I had purchased.  I had heard from Sparkman about how great they were but the ones I had seemed to have fairly miserable range.  Then I read a thread alerting people to be wary of fakes, and, indeed, it turned out that mine very likely were fakes based on their tracecode.  So, next I purchased some modules from Itead, who people thought were using genuine Nordic Semiconductor chips, and after getting those from a distributor, I was greatly disappointed to find out that they too seemed to have (at least in my testing) fairly miserable range.  At that point I suspected the Itead modules were likely fakes as well, and with no way to know what was real or fake I sampled modules from a bunch of different vendors in an attempt to find some that performed well and were therefore (presumably) real.  The odd twist in all this is that I did find some high performing modules that stood out from the rest, and they were actually the cheaper, definitely fake modules modules with the epoxy blob on them.  Compared to the others, its range is just fantastic, and they are nowhere near as sensitive to antenna orientation as the other modules.  Today's testing does show that they draw roughly 50% higher currentr during transmission (which, to me, is an acceptable tradeoff for their superior range) than a genuine chip would.  I need to test if their sleep current is sufficiently low, but if so, then I'm sold on them as being actually better than the original, at least for my purposes of using them in sensors for home automation.  Ironically, the Itead modules probably are, in fact, genuine--I sent photos of the chip tracecodes to Nordic Semiconductor, who responded and said they don't correspond to any "known fakes."  The very recent transmission current testing also suggests the Itead modules are likely to be genuine. Nordic Semi offered to x-ray them, but it's now  a moot point if the "blob modules" actually perform better.  In the end it may turn out that the transmission current will help differentiate fakes from the genuine article, but the curious twist is that I may have found a cheap fake that I actually prefer!

 

[NeverDie]

Hi NeverDie, this video about noise and digital oscilloscopes might be helpful. https://www.youtube.com/watch?v=Znwp0pK8Tzk

Thanks for the video link.  
 
Yesterday I posted a query on the eevblog forum, and Dave Jones quickly responded with another relevant, new video:
https://www.youtube.com/watch?v=DMXiD3dKYJc
Quite timely!