Hi, 
Guest
Pic
Sign In

time-nuts@lists.febo.com

Discussion of precise time and frequency measurement

View all threads

Ublox M8N - have I a XO or TCXO ?

KG
Kiwi Geoff
Tue, Dec 20, 2016 11:08 PM

Hello All - and Seasons Greetings,

One of the advantages of the recent hobby drone phenomena - it has
brought to the market a lot of low cost GNSS modules that are
lightweight for drone flight control systems. Those of us with other
hobbies, like "Time Nuts" and RTK - these low cost modules can be a
little goldmine for cost effective toys.

However the dark side is that some vendors are re-badging lower cost
modules, printing their own labels and marking as "Ublox M8N" for
example.

I recently bought two M8N modules, and wondered are they genuine - do
they actually have a TCXO (as they should do) or have I a re-badged
plain M8030 etc ?

A check with "u-center" and looking at UBX-MON-VER shows the correct
info, but what about the oscillator - is it a TCXO , good enough to do
RTK with ?

My other new present from Santa ( in the guise of Mark and John) is LH
5 - wow what a wonderful piece of engineering ! - thank you so much,
it has soaked up many a pleasurable hour !

Looking at the ADEV ability of LH5, I wondered if I could use LH to
answer the TCXO question of "my" M8N devices.

I did an overnight run and (hopefully) attached a screen shot, and
have the ADEV results at the end of this message.

While "I think" the graph and data is impressive, I'm not sure
"exactly" what I'm looking at, is the M8N running a "paper clock" and
so it's like "Man with ONE watch knows the exact time" scenario ?

The Ublox M8N was running outside (with a Max-Min thermometer saying
overnight it was 12 - 18 degrees Centigrade), on the graph you can see
the effect of Sunrise on the module, but I'm surprised how little the
clock changed.

Can anyone comment on the following data, and whether they think the
oscillator in "my" M8N is a XO or a TCXO ?

From the attached screen shot, you can see the hole in reception over

the South Pole (yes a Southern Hemisphere Time Nut) - but for the
first time I can see how Glonass makes the hole smaller than GPS alone
!

Also I am using firmware 2.01 in the M8N - so as to do RTK.

Regards, Geoff (Christchurch, New Zealand).

<name>Ublox receiver</name>

Unit type:    Ublox timing receiver

SW:            2.01 (75350)

HW:            00080000

ROM:          2.01 (75331)

07:46:54.000 UTC  20 Dec 2016 - interval 1 seconds

tow accu(ns)  frac(ns) sawtooth temp(C) sats

07:46:54,200831,5.000000,432505.000000,7.672000,0.000000,19
#! new reciever mode: 3D positioning: at tow 200831
#! new minor alarm state 0080:  LEAP PENDING!    : at tow 200831
07:46:55,200832,5.000000,432592.000000,3.694000,0.000000,19
07:46:56,200833,5.000000,432679.000000,-0.254000,0.000000,19
07:46:57,200834,5.000000,432767.000000,-4.261000,0.000000,19

The following shows "clock" adjusting to be within 0.5ms of GPS reference.

07:59:44,201601,6.000000,499848.000000,-1.371000,0.000000,18
07:59:45,201602,6.000000,499935.000000,-4.767000,0.000000,18
07:59:46,201603,6.000000,500022.000000,-8.280000,0.000000,18
07:59:47,201604,6.000000,-499891.000000,8.944000,0.000000,18
07:59:48,201605,6.000000,-499804.000000,5.431000,0.000000,18
07:59:49,201606,6.000000,-499717.000000,1.860000,0.000000,18

End of nights run !

18:53:06,240803,5.000000,67960.000000,6.446000,0.000000,19
18:53:07,240804,5.000000,67821.000000,-0.328000,0.000000,19
18:53:08,240805,5.000000,67682.000000,-6.074000,0.000000,19

Accu ADEV over 40081 points - sample period=1.000 secs

1.000 tau  7.2040e-011 (n=32998)

2.000 tau  4.8366e-011 (n=32996)

5.000 tau  2.8732e-011 (n=32990)

10.000 tau  1.9632e-011 (n=32980)

20.000 tau  1.3585e-011 (n=32960)

50.000 tau  7.5530e-012 (n=32900)

100.000 tau  4.6426e-012 (n=32800)

200.000 tau  2.6046e-012 (n=32600)

500.000 tau  1.3244e-012 (n=32000)

1000.000 tau  8.7194e-013 (n=31000)

2000.000 tau  5.2908e-013 (n=29000)

5000.000 tau  1.9171e-013 (n=23000)

10000.000 tau  1.2432e-013 (n=13000)

Accu HDEV over 40081 points - sample period=1.000 secs

1.000 tau  7.2902e-011 (n=32997)

2.000 tau  4.8859e-011 (n=32994)

5.000 tau  2.8982e-011 (n=32985)

10.000 tau  1.9672e-011 (n=32970)

20.000 tau  1.3796e-011 (n=32940)

50.000 tau  7.7782e-012 (n=32850)

100.000 tau  4.8215e-012 (n=32700)

200.000 tau  2.6956e-012 (n=32400)

500.000 tau  1.3305e-012 (n=31500)

1000.000 tau  8.8066e-013 (n=30000)

2000.000 tau  5.0519e-013 (n=27000)

5000.000 tau  1.8091e-013 (n=18000)

10000.000 tau  1.1809e-013 (n=3000)

Accu MDEV over 40081 points - sample period=1.000 secs

1.000 tau  7.2040e-011 (n=32998)

2.000 tau  3.7837e-011 (n=32995)

5.000 tau  2.0161e-011 (n=32986)

10.000 tau  1.3699e-011 (n=32971)

20.000 tau  9.3194e-012 (n=32941)

50.000 tau  4.9118e-012 (n=32851)

100.000 tau  2.9460e-012 (n=32701)

200.000 tau  1.5282e-012 (n=32401)

500.000 tau  8.0902e-013 (n=31501)

1000.000 tau  5.1762e-013 (n=30001)

2000.000 tau  3.3773e-013 (n=27001)

5000.000 tau  9.1508e-014 (n=18001)

10000.000 tau  6.2193e-014 (n=3001)

Accu TDEV over 40081 points - sample period=1.000 secs

1.000 tau  4.1592e-011 (n=32998)

2.000 tau  4.3691e-011 (n=32995)

5.000 tau  5.8201e-011 (n=32986)

10.000 tau  7.9093e-011 (n=32971)

20.000 tau  1.0761e-010 (n=32941)

50.000 tau  1.4179e-010 (n=32851)

100.000 tau  1.7009e-010 (n=32701)

200.000 tau  1.7646e-010 (n=32401)

500.000 tau  2.3354e-010 (n=31501)

1000.000 tau  2.9885e-010 (n=30001)

2000.000 tau  3.8997e-010 (n=27001)

5000.000 tau  2.6416e-010 (n=18001)

10000.000 tau  3.5907e-010 (n=3001)

Frac ADEV over 40081 points - sample period=1.000 secs

1.000 tau  8.6517e-004 (n=32998)

2.000 tau  6.1178e-004 (n=32996)

5.000 tau  3.8695e-004 (n=32990)

10.000 tau  2.7365e-004 (n=32980)

20.000 tau  1.9355e-004 (n=32960)

50.000 tau  1.2250e-004 (n=32900)

100.000 tau  8.6732e-005 (n=32800)

200.000 tau  6.1484e-005 (n=32600)

500.000 tau  3.8542e-005 (n=32000)

1000.000 tau  2.5388e-005 (n=31000)

2000.000 tau  1.7593e-005 (n=29000)

5000.000 tau  1.1647e-005 (n=23000)

10000.000 tau  4.9191e-006 (n=13000)

Frac HDEV over 40081 points - sample period=1.000 secs

1.000 tau  8.6518e-004 (n=32997)

2.000 tau  6.1180e-004 (n=32994)

5.000 tau  3.8698e-004 (n=32985)

10.000 tau  2.7369e-004 (n=32970)

20.000 tau  1.9360e-004 (n=32940)

50.000 tau  1.2258e-004 (n=32850)

100.000 tau  8.6841e-005 (n=32700)

200.000 tau  6.1639e-005 (n=32400)

500.000 tau  3.7430e-005 (n=31500)

1000.000 tau  2.4097e-005 (n=30000)

2000.000 tau  1.7484e-005 (n=27000)

5000.000 tau  1.3218e-005 (n=18000)

10000.000 tau  4.9282e-006 (n=3000)

Frac MDEV over 40081 points - sample period=1.000 secs

1.000 tau  8.6517e-004 (n=32998)

2.000 tau  4.8366e-004 (n=32995)

5.000 tau  2.7905e-004 (n=32986)

10.000 tau  1.9449e-004 (n=32971)

20.000 tau  1.3706e-004 (n=32941)

50.000 tau  8.6694e-005 (n=32851)

100.000 tau  6.1408e-005 (n=32701)

200.000 tau  4.3586e-005 (n=32401)

500.000 tau  2.5653e-005 (n=31501)

1000.000 tau  1.7418e-005 (n=30001)

2000.000 tau  1.2215e-005 (n=27001)

5000.000 tau  8.4702e-006 (n=18001)

10000.000 tau  8.6617e-007 (n=3001)

Frac TDEV over 40081 points - sample period=1.000 secs

1.000 tau  4.9951e-004 (n=32998)

2.000 tau  5.5849e-004 (n=32995)

5.000 tau  8.0555e-004 (n=32986)

10.000 tau  1.1229e-003 (n=32971)

20.000 tau  1.5827e-003 (n=32941)

50.000 tau  2.5026e-003 (n=32851)

100.000 tau  3.5454e-003 (n=32701)

200.000 tau  5.0328e-003 (n=32401)

500.000 tau  7.4054e-003 (n=31501)

1000.000 tau  1.0056e-002 (n=30001)

2000.000 tau  1.4104e-002 (n=27001)

5000.000 tau  2.4451e-002 (n=18001)

10000.000 tau  5.0008e-003 (n=3001)

#*** 708929 Rcvr packets processed.  0 bad packets.

Hello All - and Seasons Greetings, One of the advantages of the recent hobby drone phenomena - it has brought to the market a lot of low cost GNSS modules that are lightweight for drone flight control systems. Those of us with other hobbies, like "Time Nuts" and RTK - these low cost modules can be a little goldmine for cost effective toys. However the dark side is that some vendors are re-badging lower cost modules, printing their own labels and marking as "Ublox M8N" for example. I recently bought two M8N modules, and wondered are they genuine - do they actually have a TCXO (as they should do) or have I a re-badged plain M8030 etc ? A check with "u-center" and looking at UBX-MON-VER shows the correct info, but what about the oscillator - is it a TCXO , good enough to do RTK with ? My other new present from Santa ( in the guise of Mark and John) is LH 5 - wow what a wonderful piece of engineering ! - thank you so much, it has soaked up many a pleasurable hour ! Looking at the ADEV ability of LH5, I wondered if I could use LH to answer the TCXO question of "my" M8N devices. I did an overnight run and (hopefully) attached a screen shot, and have the ADEV results at the end of this message. While "I think" the graph and data is impressive, I'm not sure "exactly" what I'm looking at, is the M8N running a "paper clock" and so it's like "Man with ONE watch knows the exact time" scenario ? The Ublox M8N was running outside (with a Max-Min thermometer saying overnight it was 12 - 18 degrees Centigrade), on the graph you can see the effect of Sunrise on the module, but I'm surprised how little the clock changed. Can anyone comment on the following data, and whether they think the oscillator in "my" M8N is a XO or a TCXO ? >From the attached screen shot, you can see the hole in reception over the South Pole (yes a Southern Hemisphere Time Nut) - but for the first time I can see how Glonass makes the hole smaller than GPS alone ! Also I am using firmware 2.01 in the M8N - so as to do RTK. Regards, Geoff (Christchurch, New Zealand). # <name>Ublox receiver</name> # Unit type: Ublox timing receiver # SW: 2.01 (75350) # HW: 00080000 # ROM: 2.01 (75331) # # # 07:46:54.000 UTC 20 Dec 2016 - interval 1 seconds # # tow accu(ns) frac(ns) sawtooth temp(C) sats 07:46:54,200831,5.000000,432505.000000,7.672000,0.000000,19 #! new reciever mode: 3D positioning: at tow 200831 #! new minor alarm state 0080: LEAP PENDING! : at tow 200831 07:46:55,200832,5.000000,432592.000000,3.694000,0.000000,19 07:46:56,200833,5.000000,432679.000000,-0.254000,0.000000,19 07:46:57,200834,5.000000,432767.000000,-4.261000,0.000000,19 # # The following shows "clock" adjusting to be within 0.5ms of GPS reference. # 07:59:44,201601,6.000000,499848.000000,-1.371000,0.000000,18 07:59:45,201602,6.000000,499935.000000,-4.767000,0.000000,18 07:59:46,201603,6.000000,500022.000000,-8.280000,0.000000,18 07:59:47,201604,6.000000,-499891.000000,8.944000,0.000000,18 07:59:48,201605,6.000000,-499804.000000,5.431000,0.000000,18 07:59:49,201606,6.000000,-499717.000000,1.860000,0.000000,18 # # End of nights run ! # 18:53:06,240803,5.000000,67960.000000,6.446000,0.000000,19 18:53:07,240804,5.000000,67821.000000,-0.328000,0.000000,19 18:53:08,240805,5.000000,67682.000000,-6.074000,0.000000,19 # # Accu ADEV over 40081 points - sample period=1.000 secs # 1.000 tau 7.2040e-011 (n=32998) # 2.000 tau 4.8366e-011 (n=32996) # 5.000 tau 2.8732e-011 (n=32990) # 10.000 tau 1.9632e-011 (n=32980) # 20.000 tau 1.3585e-011 (n=32960) # 50.000 tau 7.5530e-012 (n=32900) # 100.000 tau 4.6426e-012 (n=32800) # 200.000 tau 2.6046e-012 (n=32600) # 500.000 tau 1.3244e-012 (n=32000) # 1000.000 tau 8.7194e-013 (n=31000) # 2000.000 tau 5.2908e-013 (n=29000) # 5000.000 tau 1.9171e-013 (n=23000) # 10000.000 tau 1.2432e-013 (n=13000) # # Accu HDEV over 40081 points - sample period=1.000 secs # 1.000 tau 7.2902e-011 (n=32997) # 2.000 tau 4.8859e-011 (n=32994) # 5.000 tau 2.8982e-011 (n=32985) # 10.000 tau 1.9672e-011 (n=32970) # 20.000 tau 1.3796e-011 (n=32940) # 50.000 tau 7.7782e-012 (n=32850) # 100.000 tau 4.8215e-012 (n=32700) # 200.000 tau 2.6956e-012 (n=32400) # 500.000 tau 1.3305e-012 (n=31500) # 1000.000 tau 8.8066e-013 (n=30000) # 2000.000 tau 5.0519e-013 (n=27000) # 5000.000 tau 1.8091e-013 (n=18000) # 10000.000 tau 1.1809e-013 (n=3000) # # Accu MDEV over 40081 points - sample period=1.000 secs # 1.000 tau 7.2040e-011 (n=32998) # 2.000 tau 3.7837e-011 (n=32995) # 5.000 tau 2.0161e-011 (n=32986) # 10.000 tau 1.3699e-011 (n=32971) # 20.000 tau 9.3194e-012 (n=32941) # 50.000 tau 4.9118e-012 (n=32851) # 100.000 tau 2.9460e-012 (n=32701) # 200.000 tau 1.5282e-012 (n=32401) # 500.000 tau 8.0902e-013 (n=31501) # 1000.000 tau 5.1762e-013 (n=30001) # 2000.000 tau 3.3773e-013 (n=27001) # 5000.000 tau 9.1508e-014 (n=18001) # 10000.000 tau 6.2193e-014 (n=3001) # # Accu TDEV over 40081 points - sample period=1.000 secs # 1.000 tau 4.1592e-011 (n=32998) # 2.000 tau 4.3691e-011 (n=32995) # 5.000 tau 5.8201e-011 (n=32986) # 10.000 tau 7.9093e-011 (n=32971) # 20.000 tau 1.0761e-010 (n=32941) # 50.000 tau 1.4179e-010 (n=32851) # 100.000 tau 1.7009e-010 (n=32701) # 200.000 tau 1.7646e-010 (n=32401) # 500.000 tau 2.3354e-010 (n=31501) # 1000.000 tau 2.9885e-010 (n=30001) # 2000.000 tau 3.8997e-010 (n=27001) # 5000.000 tau 2.6416e-010 (n=18001) # 10000.000 tau 3.5907e-010 (n=3001) # # Frac ADEV over 40081 points - sample period=1.000 secs # 1.000 tau 8.6517e-004 (n=32998) # 2.000 tau 6.1178e-004 (n=32996) # 5.000 tau 3.8695e-004 (n=32990) # 10.000 tau 2.7365e-004 (n=32980) # 20.000 tau 1.9355e-004 (n=32960) # 50.000 tau 1.2250e-004 (n=32900) # 100.000 tau 8.6732e-005 (n=32800) # 200.000 tau 6.1484e-005 (n=32600) # 500.000 tau 3.8542e-005 (n=32000) # 1000.000 tau 2.5388e-005 (n=31000) # 2000.000 tau 1.7593e-005 (n=29000) # 5000.000 tau 1.1647e-005 (n=23000) # 10000.000 tau 4.9191e-006 (n=13000) # # Frac HDEV over 40081 points - sample period=1.000 secs # 1.000 tau 8.6518e-004 (n=32997) # 2.000 tau 6.1180e-004 (n=32994) # 5.000 tau 3.8698e-004 (n=32985) # 10.000 tau 2.7369e-004 (n=32970) # 20.000 tau 1.9360e-004 (n=32940) # 50.000 tau 1.2258e-004 (n=32850) # 100.000 tau 8.6841e-005 (n=32700) # 200.000 tau 6.1639e-005 (n=32400) # 500.000 tau 3.7430e-005 (n=31500) # 1000.000 tau 2.4097e-005 (n=30000) # 2000.000 tau 1.7484e-005 (n=27000) # 5000.000 tau 1.3218e-005 (n=18000) # 10000.000 tau 4.9282e-006 (n=3000) # # Frac MDEV over 40081 points - sample period=1.000 secs # 1.000 tau 8.6517e-004 (n=32998) # 2.000 tau 4.8366e-004 (n=32995) # 5.000 tau 2.7905e-004 (n=32986) # 10.000 tau 1.9449e-004 (n=32971) # 20.000 tau 1.3706e-004 (n=32941) # 50.000 tau 8.6694e-005 (n=32851) # 100.000 tau 6.1408e-005 (n=32701) # 200.000 tau 4.3586e-005 (n=32401) # 500.000 tau 2.5653e-005 (n=31501) # 1000.000 tau 1.7418e-005 (n=30001) # 2000.000 tau 1.2215e-005 (n=27001) # 5000.000 tau 8.4702e-006 (n=18001) # 10000.000 tau 8.6617e-007 (n=3001) # # Frac TDEV over 40081 points - sample period=1.000 secs # 1.000 tau 4.9951e-004 (n=32998) # 2.000 tau 5.5849e-004 (n=32995) # 5.000 tau 8.0555e-004 (n=32986) # 10.000 tau 1.1229e-003 (n=32971) # 20.000 tau 1.5827e-003 (n=32941) # 50.000 tau 2.5026e-003 (n=32851) # 100.000 tau 3.5454e-003 (n=32701) # 200.000 tau 5.0328e-003 (n=32401) # 500.000 tau 7.4054e-003 (n=31501) # 1000.000 tau 1.0056e-002 (n=30001) # 2000.000 tau 1.4104e-002 (n=27001) # 5000.000 tau 2.4451e-002 (n=18001) # 10000.000 tau 5.0008e-003 (n=3001) # # #*** 708929 Rcvr packets processed. 0 bad packets.
Attached image
MC
Mike Cook
Wed, Dec 21, 2016 9:20 AM

Le 21 déc. 2016 à 00:08, Kiwi Geoff geoff36@gmail.com a écrit :

Hello All - and Seasons Greetings,

One of the advantages of the recent hobby drone phenomena - it has
brought to the market a lot of low cost GNSS modules that are
lightweight for drone flight control systems. Those of us with other
hobbies, like "Time Nuts" and RTK - these low cost modules can be a
little goldmine for cost effective toys.

However the dark side is that some vendors are re-badging lower cost
modules, printing their own labels and marking as "Ublox M8N" for
example.

It wouldn't surprise me, but  you have a reference for this?

Can anyone comment on the following data, and whether they think the
oscillator in "my" M8N is a XO or a TCXO ?

I don’t thing that you can get find out from that data. Ublox indicate in the product info sheets that the TCXO option is used to get to a first fix quicker in weak signal conditions. It is not specified and I think that it is logical, that there is an improved timing solution. I would expect that both XO and TCXO versions are the same frequency and the better long term stability of a TCXO not be an influence on the single shot quantization error of the 1PPS.

time-nuts mailing list -- time-nuts@febo.com
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.

"The power of accurate observation is commonly called cynicism by those who have not got it. »
George Bernard Shaw

> Le 21 déc. 2016 à 00:08, Kiwi Geoff <geoff36@gmail.com> a écrit : > > Hello All - and Seasons Greetings, > > One of the advantages of the recent hobby drone phenomena - it has > brought to the market a lot of low cost GNSS modules that are > lightweight for drone flight control systems. Those of us with other > hobbies, like "Time Nuts" and RTK - these low cost modules can be a > little goldmine for cost effective toys. > > However the dark side is that some vendors are re-badging lower cost > modules, printing their own labels and marking as "Ublox M8N" for > example. It wouldn't surprise me, but you have a reference for this? > Can anyone comment on the following data, and whether they think the > oscillator in "my" M8N is a XO or a TCXO ? > I don’t thing that you can get find out from that data. Ublox indicate in the product info sheets that the TCXO option is used to get to a first fix quicker in weak signal conditions. It is not specified and I think that it is logical, that there is an improved timing solution. I would expect that both XO and TCXO versions are the same frequency and the better long term stability of a TCXO not be an influence on the single shot quantization error of the 1PPS. > _____________________________ > time-nuts mailing list -- time-nuts@febo.com > To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts > and follow the instructions there. "The power of accurate observation is commonly called cynicism by those who have not got it. » George Bernard Shaw
BC
Bob Camp
Wed, Dec 21, 2016 12:37 PM

Hi

On Dec 21, 2016, at 4:20 AM, Mike Cook michael.cook@sfr.fr wrote:

Le 21 déc. 2016 à 00:08, Kiwi Geoff geoff36@gmail.com a écrit :

Hello All - and Seasons Greetings,

One of the advantages of the recent hobby drone phenomena - it has
brought to the market a lot of low cost GNSS modules that are
lightweight for drone flight control systems. Those of us with other
hobbies, like "Time Nuts" and RTK - these low cost modules can be a
little goldmine for cost effective toys.

However the dark side is that some vendors are re-badging lower cost
modules, printing their own labels and marking as "Ublox M8N" for
example.

It wouldn't surprise me, but  you have a reference for this?

Can anyone comment on the following data, and whether they think the
oscillator in "my" M8N is a XO or a TCXO ?

I don’t thing that you can get find out from that data. Ublox indicate in the product info sheets that the TCXO option is used to get to a first fix quicker in weak signal conditions. It is not specified and I think that it is logical, that there is an improved timing solution. I would expect that both XO and TCXO versions are the same frequency and the better long term stability of a TCXO not be an influence on the single shot quantization error of the 1PPS.

The TCXO and XO both have similar short term stability at short tau. The only
advantage to the TCXO is faster time to first fix. If anything the XO will “spread”
the quantization error (sawtooth error) better than the TCXO. The TCXO has
more inflections in it’s frequency vs temperature curve. Thus there are more
opportunities for hanging bridges.

In both the TCXO and XO case, the PPS out of the module is based on the
clock edge closest to the PPS estimate. If the clock involved has a period of
20 ns, the error will distribute over ~ +/- 10 ns. The process is identical regardless
of the oscillator. The distribution will be the same with both oscillators. As long as the
phase noise and short tau ADEV are in spec, the PPS estimate will be the same
in both cases.

Bob

time-nuts mailing list -- time-nuts@febo.com
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.

"The power of accurate observation is commonly called cynicism by those who have not got it. »
George Bernard Shaw

time-nuts mailing list -- time-nuts@febo.com
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.

Hi > On Dec 21, 2016, at 4:20 AM, Mike Cook <michael.cook@sfr.fr> wrote: > > >> Le 21 déc. 2016 à 00:08, Kiwi Geoff <geoff36@gmail.com> a écrit : >> >> Hello All - and Seasons Greetings, >> >> One of the advantages of the recent hobby drone phenomena - it has >> brought to the market a lot of low cost GNSS modules that are >> lightweight for drone flight control systems. Those of us with other >> hobbies, like "Time Nuts" and RTK - these low cost modules can be a >> little goldmine for cost effective toys. >> >> However the dark side is that some vendors are re-badging lower cost >> modules, printing their own labels and marking as "Ublox M8N" for >> example. > > It wouldn't surprise me, but you have a reference for this? > >> Can anyone comment on the following data, and whether they think the >> oscillator in "my" M8N is a XO or a TCXO ? >> > > I don’t thing that you can get find out from that data. Ublox indicate in the product info sheets that the TCXO option is used to get to a first fix quicker in weak signal conditions. It is not specified and I think that it is logical, that there is an improved timing solution. I would expect that both XO and TCXO versions are the same frequency and the better long term stability of a TCXO not be an influence on the single shot quantization error of the 1PPS. The TCXO and XO both have similar short term stability at short tau. The only advantage to the TCXO is faster time to first fix. If anything the XO will “spread” the quantization error (sawtooth error) better than the TCXO. The TCXO has more inflections in it’s frequency vs temperature curve. Thus there are more opportunities for hanging bridges. In both the TCXO and XO case, the PPS out of the module is based on the clock edge closest to the PPS estimate. If the clock involved has a period of 20 ns, the error will distribute over ~ +/- 10 ns. The process is identical regardless of the oscillator. The distribution will be the same with both oscillators. As long as the phase noise and short tau ADEV are in spec, the PPS estimate will be the same in both cases. Bob > > >> _____________________________ >> time-nuts mailing list -- time-nuts@febo.com >> To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts >> and follow the instructions there. > > "The power of accurate observation is commonly called cynicism by those who have not got it. » > George Bernard Shaw > > _______________________________________________ > time-nuts mailing list -- time-nuts@febo.com > To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts > and follow the instructions there.
PR
Peter Reilley
Wed, Dec 21, 2016 1:43 PM

You seem to be assuming that the crystal in the TCXO is the same as the
crystal in the XO.
Wouldn't it be likely that the crystal would be higher quality in the
more expensive
product; the TCXO?  How would a cheap crystal vs an expensive crystal
appear different
in the GPS data presented (ignoring the TC part)?

Pete.

On 12/21/2016 7:37 AM, Bob Camp wrote:

Hi

On Dec 21, 2016, at 4:20 AM, Mike Cook michael.cook@sfr.fr wrote:

Le 21 déc. 2016 à 00:08, Kiwi Geoff geoff36@gmail.com a écrit :

Hello All - and Seasons Greetings,

One of the advantages of the recent hobby drone phenomena - it has
brought to the market a lot of low cost GNSS modules that are
lightweight for drone flight control systems. Those of us with other
hobbies, like "Time Nuts" and RTK - these low cost modules can be a
little goldmine for cost effective toys.

However the dark side is that some vendors are re-badging lower cost
modules, printing their own labels and marking as "Ublox M8N" for
example.

It wouldn't surprise me, but  you have a reference for this?

Can anyone comment on the following data, and whether they think the
oscillator in "my" M8N is a XO or a TCXO ?

I don’t thing that you can get find out from that data. Ublox indicate in the product info sheets that the TCXO option is used to get to a first fix quicker in weak signal conditions. It is not specified and I think that it is logical, that there is an improved timing solution. I would expect that both XO and TCXO versions are the same frequency and the better long term stability of a TCXO not be an influence on the single shot quantization error of the 1PPS.

The TCXO and XO both have similar short term stability at short tau. The only
advantage to the TCXO is faster time to first fix. If anything the XO will “spread”
the quantization error (sawtooth error) better than the TCXO. The TCXO has
more inflections in it’s frequency vs temperature curve. Thus there are more
opportunities for hanging bridges.

In both the TCXO and XO case, the PPS out of the module is based on the
clock edge closest to the PPS estimate. If the clock involved has a period of
20 ns, the error will distribute over ~ +/- 10 ns. The process is identical regardless
of the oscillator. The distribution will be the same with both oscillators. As long as the
phase noise and short tau ADEV are in spec, the PPS estimate will be the same
in both cases.

Bob

time-nuts mailing list -- time-nuts@febo.com
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.

"The power of accurate observation is commonly called cynicism by those who have not got it. »
George Bernard Shaw

time-nuts mailing list -- time-nuts@febo.com
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.

time-nuts mailing list -- time-nuts@febo.com
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.

You seem to be assuming that the crystal in the TCXO is the same as the crystal in the XO. Wouldn't it be likely that the crystal would be higher quality in the more expensive product; the TCXO? How would a cheap crystal vs an expensive crystal appear different in the GPS data presented (ignoring the TC part)? Pete. On 12/21/2016 7:37 AM, Bob Camp wrote: > Hi > >> On Dec 21, 2016, at 4:20 AM, Mike Cook <michael.cook@sfr.fr> wrote: >> >> >>> Le 21 déc. 2016 à 00:08, Kiwi Geoff <geoff36@gmail.com> a écrit : >>> >>> Hello All - and Seasons Greetings, >>> >>> One of the advantages of the recent hobby drone phenomena - it has >>> brought to the market a lot of low cost GNSS modules that are >>> lightweight for drone flight control systems. Those of us with other >>> hobbies, like "Time Nuts" and RTK - these low cost modules can be a >>> little goldmine for cost effective toys. >>> >>> However the dark side is that some vendors are re-badging lower cost >>> modules, printing their own labels and marking as "Ublox M8N" for >>> example. >> It wouldn't surprise me, but you have a reference for this? >> >>> Can anyone comment on the following data, and whether they think the >>> oscillator in "my" M8N is a XO or a TCXO ? >>> >> I don’t thing that you can get find out from that data. Ublox indicate in the product info sheets that the TCXO option is used to get to a first fix quicker in weak signal conditions. It is not specified and I think that it is logical, that there is an improved timing solution. I would expect that both XO and TCXO versions are the same frequency and the better long term stability of a TCXO not be an influence on the single shot quantization error of the 1PPS. > > The TCXO and XO both have similar short term stability at short tau. The only > advantage to the TCXO is faster time to first fix. If anything the XO will “spread” > the quantization error (sawtooth error) better than the TCXO. The TCXO has > more inflections in it’s frequency vs temperature curve. Thus there are more > opportunities for hanging bridges. > > In both the TCXO and XO case, the PPS out of the module is based on the > clock edge closest to the PPS estimate. If the clock involved has a period of > 20 ns, the error will distribute over ~ +/- 10 ns. The process is identical regardless > of the oscillator. The distribution will be the same with both oscillators. As long as the > phase noise and short tau ADEV are in spec, the PPS estimate will be the same > in both cases. > > Bob > >> >>> _____________________________ >>> time-nuts mailing list -- time-nuts@febo.com >>> To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts >>> and follow the instructions there. >> "The power of accurate observation is commonly called cynicism by those who have not got it. » >> George Bernard Shaw >> >> _______________________________________________ >> time-nuts mailing list -- time-nuts@febo.com >> To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts >> and follow the instructions there. > _______________________________________________ > time-nuts mailing list -- time-nuts@febo.com > To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts > and follow the instructions there. >
JG
Joseph Gray
Wed, Dec 21, 2016 3:09 PM

Some time ago, I bought two different modules from Reyax on ebay. One
module had a ublox M8N. Recently, I did some reading on several drone
forums about fake ublox modules from China. It seems that modules from
quite a few vendors are not genuine. From the information presented
about identifying the fakes, I am fairly confident that the modules I
bought from Reyax are genuine.

Joe Gray
W5JG

Some time ago, I bought two different modules from Reyax on ebay. One module had a ublox M8N. Recently, I did some reading on several drone forums about fake ublox modules from China. It seems that modules from quite a few vendors are not genuine. From the information presented about identifying the fakes, I am fairly confident that the modules I bought from Reyax are genuine. Joe Gray W5JG
BC
Bob Camp
Wed, Dec 21, 2016 5:34 PM

Hi

Given the short term stability limits on the frequency source (TCXO or XO) it’s a pretty
good bet that the crystal in both oscillators is very similar. There’s not a lot of difference
between a “normal” crystal and a “TCXO” crystal other than the angle of cut tolerance.
Since the XO has frequency rate of change specs on it for the GPS to work, it’s got
some angle limits that get it pretty close to TCXO land even in that regard.

Bob

On Dec 21, 2016, at 8:43 AM, Peter Reilley preilley_454@comcast.net wrote:

You seem to be assuming that the crystal in the TCXO is the same as the crystal in the XO.
Wouldn't it be likely that the crystal would be higher quality in the more expensive
product; the TCXO?  How would a cheap crystal vs an expensive crystal appear different
in the GPS data presented (ignoring the TC part)?

Pete.

On 12/21/2016 7:37 AM, Bob Camp wrote:

Hi

On Dec 21, 2016, at 4:20 AM, Mike Cook michael.cook@sfr.fr wrote:

Le 21 déc. 2016 à 00:08, Kiwi Geoff geoff36@gmail.com a écrit :

Hello All - and Seasons Greetings,

One of the advantages of the recent hobby drone phenomena - it has
brought to the market a lot of low cost GNSS modules that are
lightweight for drone flight control systems. Those of us with other
hobbies, like "Time Nuts" and RTK - these low cost modules can be a
little goldmine for cost effective toys.

However the dark side is that some vendors are re-badging lower cost
modules, printing their own labels and marking as "Ublox M8N" for
example.

It wouldn't surprise me, but  you have a reference for this?

Can anyone comment on the following data, and whether they think the
oscillator in "my" M8N is a XO or a TCXO ?

I don’t thing that you can get find out from that data. Ublox indicate in the product info sheets that the TCXO option is used to get to a first fix quicker in weak signal conditions. It is not specified and I think that it is logical, that there is an improved timing solution. I would expect that both XO and TCXO versions are the same frequency and the better long term stability of a TCXO not be an influence on the single shot quantization error of the 1PPS.

The TCXO and XO both have similar short term stability at short tau. The only
advantage to the TCXO is faster time to first fix. If anything the XO will “spread”
the quantization error (sawtooth error) better than the TCXO. The TCXO has
more inflections in it’s frequency vs temperature curve. Thus there are more
opportunities for hanging bridges.

In both the TCXO and XO case, the PPS out of the module is based on the
clock edge closest to the PPS estimate. If the clock involved has a period of
20 ns, the error will distribute over ~ +/- 10 ns. The process is identical regardless
of the oscillator. The distribution will be the same with both oscillators. As long as the
phase noise and short tau ADEV are in spec, the PPS estimate will be the same
in both cases.

Bob

time-nuts mailing list -- time-nuts@febo.com
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.

"The power of accurate observation is commonly called cynicism by those who have not got it. »
George Bernard Shaw

time-nuts mailing list -- time-nuts@febo.com
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.

time-nuts mailing list -- time-nuts@febo.com
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.

time-nuts mailing list -- time-nuts@febo.com
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.

Hi Given the short term stability limits on the frequency source (TCXO or XO) it’s a pretty good bet that the crystal in both oscillators is very similar. There’s not a lot of difference between a “normal” crystal and a “TCXO” crystal other than the angle of cut tolerance. Since the XO has frequency rate of change specs on it for the GPS to work, it’s got some angle limits that get it pretty close to TCXO land even in that regard. Bob > On Dec 21, 2016, at 8:43 AM, Peter Reilley <preilley_454@comcast.net> wrote: > > You seem to be assuming that the crystal in the TCXO is the same as the crystal in the XO. > Wouldn't it be likely that the crystal would be higher quality in the more expensive > product; the TCXO? How would a cheap crystal vs an expensive crystal appear different > in the GPS data presented (ignoring the TC part)? > > Pete. > > > On 12/21/2016 7:37 AM, Bob Camp wrote: >> Hi >> >>> On Dec 21, 2016, at 4:20 AM, Mike Cook <michael.cook@sfr.fr> wrote: >>> >>> >>>> Le 21 déc. 2016 à 00:08, Kiwi Geoff <geoff36@gmail.com> a écrit : >>>> >>>> Hello All - and Seasons Greetings, >>>> >>>> One of the advantages of the recent hobby drone phenomena - it has >>>> brought to the market a lot of low cost GNSS modules that are >>>> lightweight for drone flight control systems. Those of us with other >>>> hobbies, like "Time Nuts" and RTK - these low cost modules can be a >>>> little goldmine for cost effective toys. >>>> >>>> However the dark side is that some vendors are re-badging lower cost >>>> modules, printing their own labels and marking as "Ublox M8N" for >>>> example. >>> It wouldn't surprise me, but you have a reference for this? >>> >>>> Can anyone comment on the following data, and whether they think the >>>> oscillator in "my" M8N is a XO or a TCXO ? >>>> >>> I don’t thing that you can get find out from that data. Ublox indicate in the product info sheets that the TCXO option is used to get to a first fix quicker in weak signal conditions. It is not specified and I think that it is logical, that there is an improved timing solution. I would expect that both XO and TCXO versions are the same frequency and the better long term stability of a TCXO not be an influence on the single shot quantization error of the 1PPS. >> >> The TCXO and XO both have similar short term stability at short tau. The only >> advantage to the TCXO is faster time to first fix. If anything the XO will “spread” >> the quantization error (sawtooth error) better than the TCXO. The TCXO has >> more inflections in it’s frequency vs temperature curve. Thus there are more >> opportunities for hanging bridges. >> >> In both the TCXO and XO case, the PPS out of the module is based on the >> clock edge closest to the PPS estimate. If the clock involved has a period of >> 20 ns, the error will distribute over ~ +/- 10 ns. The process is identical regardless >> of the oscillator. The distribution will be the same with both oscillators. As long as the >> phase noise and short tau ADEV are in spec, the PPS estimate will be the same >> in both cases. >> >> Bob >> >>> >>>> _____________________________ >>>> time-nuts mailing list -- time-nuts@febo.com >>>> To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts >>>> and follow the instructions there. >>> "The power of accurate observation is commonly called cynicism by those who have not got it. » >>> George Bernard Shaw >>> >>> _______________________________________________ >>> time-nuts mailing list -- time-nuts@febo.com >>> To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts >>> and follow the instructions there. >> _______________________________________________ >> time-nuts mailing list -- time-nuts@febo.com >> To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts >> and follow the instructions there. >> > > _______________________________________________ > time-nuts mailing list -- time-nuts@febo.com > To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts > and follow the instructions there.
KG
Kiwi Geoff
Wed, Dec 21, 2016 10:30 PM

Mike Cook wrote:

It wouldn't surprise me, but  you have a reference for this?

I first read about this issue in the official Ublox forum Mike - try this link,

http://tinyurl.com/Fake-M8Ns

  • at time of writing it leads to some of the messages that discuss
    the issue with photos and examples of M8N's that appear - not to be !

I must say, the two devices I have, appear to be genuine, and do
everything they should, and are a wonderful piece of kit.

My initial (now erroneous) thought was that maybe I could do a quick
overnight run with Lady Heather 5, to show up the M8N oscillator type.
Thanks to the excellent feedback on this thread, I now see why this is
not so, my overnight graph (seen in OP) is looking at a layer of
operation that somewhat hides the actual oscillator behaviour. I still
think it is an interesting graph, and shows the M8N can do a good job
of being a GNSS timing part.

Although as Mark S has said, it requires independent testing against a
reference to be done, rather than my graph - which essentially is the
M8N "self reporting" its quality.

From my reading so far, the current best way to find out if you have a

genuine M8N is to use u-centre , and ask it to poll the UBX-MON-VER
command, which should return:

original firmware (2.01): SWVER 2.01 (75350) HWVER 00080000 EXTENSION
2.01 (75331) PROTVER 15.00 FIS 0xEF4015 (79189) MOD NEO-M8N-0
GPS;SBAS;GLO;BDS;QZSS

after firmware update to (3.01): SWVER EXT CORE 3.01 (107900) HWVER
00080000 EXTENSION ROM base 2.01 (75331) FWVER=SPG 3.01 PROTVER=18.00
FIS=0xEF4015 (100111) GPS;GLO;GAL;BDS;SBAS;IMES;QZSS

Take special note of the "FIS=" (Flash Information Structure) size.

Firmware Update to 3.01 can give you the (newly turned on) Galileo
system, however some people (like me) may choose to stay with Firmware
2.01 because it allows playing the RTK game thanks to the work of:

http://rtkexplorer.com/

Again, thank you for the replies to this thread, I have learned a lot,
and especially Bob Camp, who gave a nice explanation of the oscillator
in such a GNSS module.

Regards, Geoff (Christchurch, New Zealand).

Mike Cook wrote: > It wouldn't surprise me, but you have a reference for this? I first read about this issue in the official Ublox forum Mike - try this link, http://tinyurl.com/Fake-M8Ns - at time of writing it leads to some of the messages that discuss the issue with photos and examples of M8N's that appear - not to be ! I must say, the two devices I have, appear to be genuine, and do everything they should, and are a wonderful piece of kit. My initial (now erroneous) thought was that maybe I could do a quick overnight run with Lady Heather 5, to show up the M8N oscillator type. Thanks to the excellent feedback on this thread, I now see why this is not so, my overnight graph (seen in OP) is looking at a layer of operation that somewhat hides the actual oscillator behaviour. I still think it is an interesting graph, and shows the M8N can do a good job of being a GNSS timing part. Although as Mark S has said, it requires independent testing against a reference to be done, rather than my graph - which essentially is the M8N "self reporting" its quality. >From my reading so far, the current best way to find out if you have a genuine M8N is to use u-centre , and ask it to poll the UBX-MON-VER command, which should return: ----------------------------------------------------- original firmware (2.01): SWVER 2.01 (75350) HWVER 00080000 EXTENSION 2.01 (75331) PROTVER 15.00 FIS 0xEF4015 (79189) MOD NEO-M8N-0 GPS;SBAS;GLO;BDS;QZSS after firmware update to (3.01): SWVER EXT CORE 3.01 (107900) HWVER 00080000 EXTENSION ROM base 2.01 (75331) FWVER=SPG 3.01 PROTVER=18.00 FIS=0xEF4015 (100111) GPS;GLO;GAL;BDS;SBAS;IMES;QZSS ----------------------------------------------------- Take special note of the "FIS=" (Flash Information Structure) size. Firmware Update to 3.01 can give you the (newly turned on) Galileo system, however some people (like me) may choose to stay with Firmware 2.01 because it allows playing the RTK game thanks to the work of: http://rtkexplorer.com/ Again, thank you for the replies to this thread, I have learned a lot, and especially Bob Camp, who gave a nice explanation of the oscillator in such a GNSS module. Regards, Geoff (Christchurch, New Zealand).
Next page
Empathy v1.0   2025 ©emwd.com
Documentation