Author Topic: Making a lower power Moteino  (Read 18398 times)

WhiteHare

  • Hero Member
  • *****
  • Posts: 1297
  • Country: us
Re: Making a lower power Moteino
« Reply #30 on: April 22, 2016, 09:24:50 AM »
@Felix  Looking at the datasheet, it only gives the recommended operating voltage range for 25C.  I don't see any graphs showing the effects of different temperatures.  Do you have any info as to how sensitive those numbers are to temperature? 

Felix

  • Administrator
  • Hero Member
  • *****
  • Posts: 5261
  • Country: us
    • LowPowerLab
Re: Making a lower power Moteino
« Reply #31 on: April 22, 2016, 09:36:22 AM »
@Felix  Looking at the datasheet, it only gives the recommended operating voltage range for 25C.  I don't see any graphs showing the effects of different temperatures.  Do you have any info as to how sensitive those numbers are to temperature?
Unfortunately no, I don't have more than what the datasheet reveals.  :-\

perky

  • Hero Member
  • *****
  • Posts: 846
  • Country: gb
Re: Making a lower power Moteino
« Reply #32 on: April 22, 2016, 09:40:24 AM »
@Felix: Wow, good stuff Felix! I have to say the build quality of yours compared to theirs is so much better, just look at the soldering :-O

@WhiteHare: I think these things work by forward diasing diodes to enable rf through and reverse biasing to block, if that's the case and it isn't just a logic level thing then there could well be a temperature dependence.

Mark.

joelucid

  • Hero Member
  • *****
  • Posts: 869
Re: Making a lower power Moteino
« Reply #33 on: April 22, 2016, 11:03:49 AM »
Felix, have you ever tried to use a RFM69HW at less than 2.5V or is the assumption that it doesn't work based on the data sheet alone?

Felix

  • Administrator
  • Hero Member
  • *****
  • Posts: 5261
  • Country: us
    • LowPowerLab
Re: Making a lower power Moteino
« Reply #34 on: April 22, 2016, 11:25:17 AM »
Felix, have you ever tried to use a RFM69HW at less than 2.5V or is the assumption that it doesn't work based on the data sheet alone?
Not that I can remember.
I know you like to push things to the limit, I do too sometimes (ie Moteino 16mhz @ 3.3v). But with this RFM69 at less than 2.5v it can be very difficult to debug if it starts to edge into some unstable condition, and it's easy for people to point fingers in other directions. That is why I would not run it at less than that.

WhiteHare

  • Hero Member
  • *****
  • Posts: 1297
  • Country: us
Re: Making a lower power Moteino
« Reply #35 on: April 22, 2016, 12:52:21 PM »
Funny how perspectives change.  In the face of uncertainty, keeping the LDO and running from a 9-volt battery "just to be sure" suddenly looks like an attractive, compact, and rational choice if running an HW....  I don't really feel like running a ton more experiments just to test whether the datasheet might be wrong.
« Last Edit: April 22, 2016, 01:11:00 PM by WhiteHare »

joelucid

  • Hero Member
  • *****
  • Posts: 869
Re: Making a lower power Moteino
« Reply #36 on: April 22, 2016, 01:33:56 PM »
Quote
Funny how perspectives change.  In the face of uncertainty, keeping the LDO and running from a 9-volt battery "just to be sure" suddenly looks like an attractive, compact, and rational choice if running an HW....  I don't really feel like running a ton more experiments just to test whether the datasheet might be wrong.

Well lets not jump to conclusions. I'm running a test now: RFM69HW on two aa Eneloops with 3 Ohm in parallel - shouldn't take too long. As of now it's happily transmitting at 17 dBm at 2.2V, creating an RSSI of -34 at the receiver.

One interesting tidbit: The RFM69HW Spec V1.0 specifies minimum voltage at 2.4V. V1.3 has the new minimum voltage of 1.8V for 17dBm. So this was an active modification of the data sheet rather than just an error in the spec.

perky

  • Hero Member
  • *****
  • Posts: 846
  • Country: gb
Re: Making a lower power Moteino
« Reply #37 on: April 22, 2016, 01:38:52 PM »
Mmmm, if I remember the old datasheet said 2.4V specified over the the full power range, whereas the new one breaks it down into finer detail so yes, this is a clarification rather than a correction. Similar to the difference now between the RF69 and sx1231h datasheets.
Mark.

joelucid

  • Hero Member
  • *****
  • Posts: 869
Re: Making a lower power Moteino
« Reply #38 on: April 22, 2016, 04:09:18 PM »
Quote
Well lets not jump to conclusions. I'm running a test now: RFM69HW on two aa Eneloops with 3 Ohm in parallel.

Interesting - I almost needed to pinch myself. Luckily I had a Moteino with regulator removed and fuses set to no BOD lying around. As I said with this setup at 17dBm the receiver saw the test packages sent 2.2V at -34 RSSI in the beginning. From there is just continued down, very reliably, less than 0.3% packet loss.

It continued all the way down to 1.69V when the 328p just couldn't keep up appearances and suffered from ram corruption. Until then the radio hadn't missed a beat:

Code: [Select]
Apr 22 22:03:18 espgw.lx  [11] [RX_RSSI:-36,1,8,45]
Apr 22 22:03:18 espgw.lx  test4 -> nd:11,nr:285,pw:12,vc:170,rt:3
Apr 22 22:03:21 espgw.lx  [11] [RX_RSSI:-37,1,8,45]
Apr 22 22:03:21 espgw.lx  test4 -> nd:11,nr:286,pw:12,vc:170,rt:3
Apr 22 22:03:24 espgw.lx  [11] [RX_RSSI:-35,1,8,45]
Apr 22 22:03:24 espgw.lx  test4 -> nd:11,nr:287,pw:12,vc:170,rt:3
Apr 22 22:03:27 espgw.lx  [11] [RX_RSSI:-36,1,8,45]
Apr 22 22:03:27 espgw.lx  test4 -> nd:11,nr:288,pw:12,vc:170,rt:3
Apr 22 22:03:30 espgw.lx  [11] [RX_RSSI:-35,1,8,45]
Apr 22 22:03:30 espgw.lx  test4 -> nd:11,nr:289,pw:12,vc:170,rt:3
Apr 22 22:03:33 espgw.lx  [11] [RX_RSSI:-38,1,8,45]
Apr 22 22:03:33 espgw.lx  test4 -> nd:11,nr:290,pw:12,vc:170,rt:3
Apr 22 22:03:36 espgw.lx  [11] [RX_RSSI:-35,1,8,45]
Apr 22 22:03:36 espgw.lx  test4 -> nd:11,nr:291,pw:12,vc:170,rt:3
Apr 22 22:03:39 espgw.lx  [11] [RX_RSSI:-35,1,8,45]
Apr 22 22:03:39 espgw.lx  test4 -> nd:11,nr:292,pw:12,vc:170,rt:3
Apr 22 22:03:45 espgw.lx  [11] [RX_RSSI:-38,1,8,45]
Apr 22 22:03:45 espgw.lx  test4 -> nd:11,nr:294,pw:12,vc:169,rt:17
Apr 22 22:03:51 espgw.lx  [11] [RX_RSSI:-38,6,8,45]
Apr 22 22:03:51 espgw.lx  test4 -> nd:11,nr:296,pw:12,vc:169,rt:30
Apr 22 22:03:54 espgw.lx  [11] [RX_RSSI:-37,1,8,45]
Apr 22 22:03:54 espgw.lx  test4 -> nd:11,nr:297,pw:12,vc:169,rt??.
Apr 22 22:04:00 espgw.lx  [11] [RX_RSSI:-38,1,8,45]
Apr 22 22:04:00 espgw.lx  test4 -> nd:11,nr:299,pw:12,vc:169,rt:33
Apr 22 22:04:06 espgw.lx  FreeHeap: 17440
Apr 22 22:04:06 espgw.lx  [11] [RX_RSSI:-38,1,8,45]
Apr 22 22:04:06 espgw.lx  test4 -> nd:11,nr:301?test4 -> ??#014?B?00?WZ6Q^??
Apr 22 22:04:12 espgw.lx  [11] [RX_RSSI:-36,1,8,45]
Apr 22 22:04:12 espgw.lx  test4 -> nd:11,nr:303,pw:12,vc:169,rtNYV
Apr 22 22:04:15 espgw.lx  [11] [RX_RSSI:-36,1,8,45]
Apr 22 22:04:15 espgw.lx  test4 -> nd:11,nr:304,pw:12,vc:169,rtNYV
Apr 22 22:04:43 espgw.lx  [11] [RX_RSSI:-38,6,8,45]
Apr 22 22:04:43 espgw.lx  test4 -> nd:11,nr:313I?ج?"?A?k???t]#021
Apr 22 22:04:46 espgw.lx  [11] [RX_RSSI:-35,6,8,45]
Apr 22 22:04:46 espgw.lx  test4 -> nd:11,ns:314,pw:12,vc:168,rt???F
Apr 22 22:04:55 espgw.lx  [11] [RX_RSSI:-35,6,8,45]
Apr 22 22:04:55 espgw.lx  test4 -> nd:11,ns:317?-r:2??#006?#6???.
Apr 22 22:04:55 espgw.lx  [11] [RX_RSSI:-36,1,8,45]
Apr 22 22:04:55 espgw.lx  test4 -> nd:11,ns:317?o*'`???#013\??!?|a#021??K


RSSI pretty much unchanged at -36 vs -34 at 2.2V.

So I guess this should put the hypothesis that the rfm69hw only works down to 2.5V to a conclusive rest.

Joe
« Last Edit: April 22, 2016, 04:12:10 PM by joelucid »

WhiteHare

  • Hero Member
  • *****
  • Posts: 1297
  • Country: us
Re: Making a lower power Moteino
« Reply #39 on: April 22, 2016, 05:27:29 PM »
Are you stopping there, or will you be testing it all the way down to 1.8v?

joelucid

  • Hero Member
  • *****
  • Posts: 869
Re: Making a lower power Moteino
« Reply #40 on: April 22, 2016, 05:34:04 PM »
Reread the post. It ran down to 1.68v!

WhiteHare

  • Hero Member
  • *****
  • Posts: 1297
  • Country: us
Re: Making a lower power Moteino
« Reply #41 on: April 22, 2016, 07:19:07 PM »
Great work!  And fast too!

I'm perplexed though.  The fact that it goes so low is great and definitely more reassuring than if it were a higher number, but, trying to be objective about this:  why is there such a seemingly large disconnect between that result and the minimum voltage shown on the datasheet that Felix linked to?  Can anyone here perhaps shed some light on that?  For example, if, hypothetically, all the MMIC chips ever used on the RFM69 modules were tested using Joe's method, might there be a wide distribution of minimum voltages in the MMIC chips that got manufactured, and perhaps Joe simply had good fortune with the one he happened to test?  Or does it not work like that?  I presume the manufacturer has a large dataset from ongoing QA testing, and so I would think they'd be in the best position to put solid numbers in their datasheet.  I should think it would undermine their sales to state a higher minimum voltage if a lower minimum would work just as well, and so I don't know why they would do that if there were no reason to.
« Last Edit: April 22, 2016, 07:58:25 PM by WhiteHare »

TomWS

  • Global Moderator
  • Hero Member
  • *****
  • Posts: 1735
Re: Making a lower power Moteino
« Reply #42 on: April 22, 2016, 10:15:15 PM »
RSSI pretty much unchanged at -36 vs -34 at 2.2V.

So I guess this should put the hypothesis that the rfm69hw only works down to 2.5V to a conclusive rest.
But, but... RSSI of -36?  Were these two devices like right next to each other?  What's the behavior when the starting RSSI is -54?  Or a more realistic -64?   AGC can mask a wide range of sins.

Tom

joelucid

  • Hero Member
  • *****
  • Posts: 869
Re: Making a lower power Moteino
« Reply #43 on: April 22, 2016, 10:30:52 PM »
WhiteHare, I think we just don't know. Is this a Chinese clone with different specs? Who knows. But for the module: if it's spec'd like a duck and performs like a duck I think one has to assume it's a duck. 3 tiny letters on a component notwithstanding.

Tom, actually the gw still has the agc testing code from recently. The 1 after each measurement shows that agc selected the highest gain again for each packet. Proving again that agc just doesn't work with the current rfm69 library.

I have a fixed version where it does work. In that case I measure RSSI of less than -20 for nodes as close together as these two. But using a rssithresh below noise floor actually has some advantages. I think I wrote all about it in that agc thread.
« Last Edit: April 22, 2016, 10:33:59 PM by joelucid »

Felix

  • Administrator
  • Hero Member
  • *****
  • Posts: 5261
  • Country: us
    • LowPowerLab
Re: Making a lower power Moteino
« Reply #44 on: April 25, 2016, 08:39:57 AM »
joelucid, great testing, it's definitely a good thing to see this work down to 1.8v and even below. I see this as a good margin overclock case, same as the atmega328p running 16mhz at 3.3v without a hitch.
So I guess a very low voltage mote can be achieved, running from a combined 2xAA to depletion without worry that the components can't handle it (at least not the radio).
We should split this topic, what's a good starting post to do that? I'd like to continue the discussion in the direction of narrowing down some specs for such a mote. My interest is growing. :)