Author Topic: Ublox-NEO-M8N GPS navigation signal amplify module for arduino Rasppery PI  (Read 2339 times)

0 Members and 1 Guest are viewing this topic.

Offline Johnny B Good

  • Regular Contributor
  • *
  • Posts: 190
  • Country: gb
I'm looking to building myself a GPSDO on the cheap to add to my current rash of new test gear (SDS1202X-E and a much modded FY6600 DDS AWG). I've trawled the internet for various GPSDO projects to gain some inspiration/track down ready made bargains and have concluded that I could make a start with the unit mentioned in the subject line which is being offered on eBay at what seems an entirely reasonable price, postage free and from a local warehouse as per the link below:

<https://www.ebay.co.uk/itm/Ublox-NEO-M8N-GPS-navigation-signal-amplify-module-for-arduino-Rasppery-PI/312317200298?hash=item48b78e47aa:g:iYcAAOSw1KRb6riz:rk:24:pf:0>

<https://tinyurl.com/y7ypyhnn>

 It looks ideal as a simple starter kit which I can expand into a fully specced GPSDO frequency standard. I intend to wire the 1PPS line on pin 3 to a buffer amp and program the module to clock it at 10MHz for use as a raw 10MHz calibration signal. I understand that the issue in this simplistic approach is that of clock jitter which won't be helped by choosing 10MHz from what I understand is a 48MHz TCXO but it should initially suffice my immediate need since there's no question over its long term stability. I can always reprogram the output for a 1 or 2 or even a 4 MHz output and use an ultra low jitter PLL chip such as the NB3N502 to multiply back up to 10MHz (and even 20 and 30MHz to improve the response of my attempts to recalibrate the 0.1ppm 50MHz TCXO I've installed into my FY6600).

 I'd very much appreciate some advice, if you please, regarding this no name brand module before I splash the cash.  :)

Regards, Johnny B Good
 

Offline Johnny B Good

  • Regular Contributor
  • *
  • Posts: 190
  • Country: gb
Re: Ublox-NEO-M8N GPS navigation signal amplify module for arduino Rasppery PI
« Reply #1 on: February 04, 2019, 09:26:12 pm »
 Well, it looks like no one else has tried this "Ublox-NEO-M8N-GPS-navigation-signal-amplify-module-for-arduino-Rasppery-PI module" so it seems I've been elected as the "Guinea pig Purchaser of the month" to test and report its worth as the core module in a GPSDO project. >:(

 I placed an order last Thursday morning, received a confirmation with a promised delivery of Tue 5th to Wed 6th and it arrived with Saturday's 10am post. I powered it up via the micro-usb connector to give it an initial check out but it took me until the early hours of Sunday morning to sort out the USB pass through with my winXP VM (Virtualbox under LM17.2) to get the u-Blox "u-center_v8.16" software working. I've used VBox's USB pass through years earlier under VBox ver 4 in Ubuntu 12.xx when it was all different yet still tricky to get working. Now it's just tricky to get working in a different way.  >:(

 Anyway, getting the evaluation and configuration software up and running turned out to be a trivial task compared to actually getting it to lock onto satellites using its built in ceramic patch antenna. I had no problems configuring it via the usb interface so it does seem to be functioning ok. However, it does look as though its built in patch antenna is either insufficient for this task or faulty or not properly soldered onto the underside of the circuit board.

 If I monitor for long enough, it will register the presence of satellites but with insufficient C/N ratio to lock onto their data payloads. I've only witnessed two separate satellite passes which managed to raise the C/N ratio above the 5dB minimum mark. The first one, a GPS sat produced a maxima that peaked at 22dB, the second one early this morning wasn't quite so strong at just 12dB or so before disappearing back into the noise threshold of 5dB on the chart. Both passes lasting less than a minute.

 Having read articles about and seen youtube instructional videos on DIYing GPSDOs, the tiny ceramic patch antennas supplied on a 10cm flylead with the tiny press on connectors (smaller than the SMA connector fitted to this GPS module) whilst not ideal, seemed to provide a useful, if downgraded performance in receiving the stronger satellite signals that were managing to infiltrate into the rooms used to build and test these GPSDO units in so it looks like the performance of the directly attached ceramic patch antenna on my unit is either defective, badly attached, being interfered with by stray rf fields on the circuit board or else there's a problem with the Rx circuit in the u-Blox module itself.

 I see no obviously easy way to remove the ceramic patch antenna to look for a bad solder joint so now I'm awaiting delivery of an active LNA with 5 metres of co-axial cable terminated with a male SMA plug  to (hopefully!) bypass the built in antenna. This was an option I was planning on going for anyway,  if only to get more reliable reception of the satellite signals to guarantee uninterrupted service as a GPSDO. I just thought I'd have more time to pick out the best bang for my buck antenna designed to be mounted on a mast (something like those marine GPS active antennas but without the outrageous asking price so typical of such units). Never mind, I've got a 3 quid mag mount unit winging its way to me as I type. ETA for delivery being this Wednesday or Thursday.

 Other than this 'Antenna Issue', it seems ideally suited since there's no need to solder a connection to the PPS signal on the u-Blox chip pin 3 as Scully had to do with his M7M <https://www.youtube.com/watch?v=lbns-FvpzK4> since this signal is brought out on a fifth pin in the serial header. Also, since my unit is the M8N with the flash rom, I don't have to worry about upgrading the 80mF supercap that keeps the cmos sram backed up (erroneously referred to as EEPROM in the video) with a CR2032 cell making this an even better choice for this application. Indeed, with nothing more than a Poundland 1200mAH PowerBank and a 3 quid active mag mount GPS antenna, you can create a crude 10MHz calibration source (crude only in the sense of the inevitable edge jitter you get when deriving a 10MHz signal from a 48MHz TCXO) directly from the module.

 At the moment, I've got it programmed to output a 2Hz 3% duty cycle to the PPS pin (and its red led) when unlocked with a 1Hz 50% duty cycle for when it achieves satellite lock so I'll know straight away whether my antenna upgrade is going to resolve my reception issue and give me a locked signal which I can then reprogram to produce a 10MHz 50% duty cycle pulse for initial calibration purposes.

 Ultimately, if all goes well, I'll reprogram it to output a 2MHz square wave and feed it into an NB3N502 PLL multiplier chip to multiply it back up to a jitter free 10MHz for use as a frequency calibration source. Anyway, that's the situation so far and I'll report back my results with the amplified GPS antenna once I've got some results to report, hopefully before the weekend.

JBG
 

Offline floobydust

  • Super Contributor
  • ***
  • Posts: 2710
  • Country: ca
Re: Ublox-NEO-M8N GPS navigation signal amplify module for arduino Rasppery PI
« Reply #2 on: February 04, 2019, 10:20:43 pm »
I just got a cheap NEO-7M on eBay and perhaps a similar problem.
The NEO-7M has very weak signal, lucky to pick up a satellite or two. I thought it due to snow on the roof, but I stuck a short piece of wire on the SMA center pin and the unit works great.

I believe either the patch antenna or LNA U2 is bad. The patch antenna is yellow instead of pink ceramic. The LNA IC a MAXIM MAX2659 (knock off).

It's the same circuit on this style of GPS module, for the 6m, 7m, 8m.
I've asked the seller for a refund, but did not yet narrow it down to the patch antenna or LNA or a wrong passive part on the PCB.

edit: added LNA IC part number and link
« Last Edit: March 15, 2019, 05:45:15 pm by floobydust »
 
The following users thanked this post: Johnny B Good

Offline cdev

  • Super Contributor
  • ***
  • Posts: 5082
  • Country: 00
Re: Ublox-NEO-M8N GPS navigation signal amplify module for arduino Rasppery PI
« Reply #3 on: February 04, 2019, 11:08:59 pm »
If your units are both all in ones with an integrated patch antenna like the M7N or whatever it is? You can read right in the Ublox integration docs where they tell manufacturers not to put the antenna right over the chip. They shouldn't put the antenna right over the GPS chip because (duh) its noisy!

I think u-center has a noise screen, for just noise. Might be worth looking at and using in your ebay complaints. Also, try putting the GPS over an additional ground plane (flat piece of metal ) without shorting it out. It needs a larger GP to work well. Should be right on top but not shorting any dc connections. Note that the patch antenna needs to be pointing straight up and ideally the GPS should be outdoors with a clear sky view. Also, is there an external antenna connector or footprint for an SMA, etc. on your unit? Dont solder anything to it. Chances are the sellers know they don't work and are just trying to get rid of them, you'll likely get a full refund. Item not as described. If I were you I would just put it aside for now, the units might work okay with an external antenna but its non trivial to make that modification and would probably destroy it. Thats what I accidentally did. I realized then that a preheater is an absolute necessity for working on multilayer boards. I was trying to get the antenna (which turned out to be fine) off and I just toasted it, that heat let the magic smoke out. Got another one that uses an external antenna and that one works fine.
« Last Edit: February 04, 2019, 11:24:40 pm by cdev »
"What the large print giveth, the small print taketh away."
 
The following users thanked this post: Johnny B Good

Offline Johnny B Good

  • Regular Contributor
  • *
  • Posts: 190
  • Country: gb
Re: Ublox-NEO-M8N GPS navigation signal amplify module for arduino Rasppery PI
« Reply #4 on: February 05, 2019, 03:36:02 am »
 To Floobydust:  As you can see from the photos, the J1 socket area is populated with an SMA socket and the patch antenna is virtually identical to the one on your unit, (the only difference being the suffix letter "P").

 I tried a quarter wavelength wire (4.5cm) stuck into the SMA socket with the patch antenna facing down against the aluminium foil groundplane I'd made up by wrapping a small 19 by 15 cm sheet of perspex with kitchen foil but although I could still receive satellites about as badly as before, I wasn't really convinced it was any improvement since there was still some doubt as to whether turning the patch antenna to face the groundplane was really killing off the signals it was receiving.

 It's hard to be certain since it can take ten minutes or longer before new satellites start to reappear. Perhaps I'd have had better luck if I'd kept trimming a few more millimetres off my "quarterwave" antenna to account for the screened inside the socket portion of its resonant length (you didn't have this extra complication in connecting onto the end of the stripline feeder section). BTW, what length did you cut your wire antenna for? Was it quarter or three quarter wave?

 Anyway, Since I don't have any spare SMA connectors to allow me to make a better DIY GPS antenna that I can distance from the potential source of RFI from this board, I decided to give up any further experimentation along these lines and just wait for the active antenna to be delivered.

 I can see where the signal pin from the patch antenna is soldered to the board on my unit (it's directly opposite the solder blob on the top side of the antenna adjacent to pin 4 of the u-Blox module). Since it appears to be linked into the antenna line feeding the antenna input on pin 11, I'm a little concerned that it might effect the signal provided by the SMA socket. If that does prove to be a problem, it does look like it should be possible to desolder the connection using a desoldering suction tool.

 However, there's a microscopically tiny 6 pin chip which looks like it might be some sort of RF switch (PIN diode?) so there's probably no need to take any action. However, for all it's effectiveness as a built in antenna, I think I'd prefer to remove it completely if it can be done without damaging the board. OTOH, fitting a thin copper clad PCB with a Patch antenna shaped aperture to allow it to poke through with clearance holes for the SMA centre pin and the 5 pin header with 4 holes around the SMA socket by which to poke the ground pins through to achieve good ground continuity might actually allow the built in patch antenna to function without interference from the support circuitry around the u-Blox module (or not - it's just a random thought that it might possible become a more useful feature of the module).

JBG

« Last Edit: February 05, 2019, 04:22:01 am by Johnny B Good »
 

Offline Johnny B Good

  • Regular Contributor
  • *
  • Posts: 190
  • Country: gb
Re: Ublox-NEO-M8N GPS navigation signal amplify module for arduino Rasppery PI
« Reply #5 on: February 05, 2019, 04:17:34 am »
 To CDev: Thanks for that input. It did strike that mounting a patch antenna onto the back of a PCB that was hardly any larger was a rather dubious tactic to say the least. I have some familiarity with basic antenna design and the benefit of a decent sized groundplane with certain types of antenna which this appears to be of the same class, going from the advice mentioned in regard of the reception improvement afforded by the use of a foot or so square of a conductive surface placed immediately below such patch antennas.

 I can picture the self interference effect as a miniature of the situation of using a switch mode PSU in the shack to power your thousand quid HF transceiver and switching from using a centre fed balanced dipole strung up between the front and back gardens to just plugging the end of a long wire antenna straight into the antenna socket using the chassis earth as the ground connection.

 In the former case, the relatively high level of RFI from the smpsu in the shack is kept well away from the dipole antenna with any residual interference being well and truly lost in the normal and inescapable levels of atmospheric noise (QRN) versus that noise getting into the antenna socket via the portion of the long wire antenna inside the shack that's exposed to this high level of smpsu generated interference. I can well understand the manufacturer's advice against such a folly.

 I'll sit tight for the next few days until that amplified GPS mag mount antenna I ordered turns up in the post. If that doesn't solve the problem, I'll just return the GPS module for a refund as "Not Fit For Purpose" and look to buy a replacement without this built in patch antenna. The additional 3 quid expense on the LNA antenna won't go to waste since I'd planned on using an external antenna anyway.

JBG
« Last Edit: February 05, 2019, 04:22:30 am by Johnny B Good »
 

Offline floobydust

  • Super Contributor
  • ***
  • Posts: 2710
  • Country: ca
Re: Ublox-NEO-M8N GPS navigation signal amplify module for arduino Rasppery PI
« Reply #6 on: February 05, 2019, 05:06:52 am »
Patch antenna to the LNA input (the way the module is built) I view at most 1-2 satellites  >:(
If I move the patch antenna directly to the NEO-7M (SMA) input I get 12 satellites and good lock, SNR ~23 :o
If I connect 1/2 wavelength wire (9.5cm horizontal) directly to the SMA input I get ~4 satellites and good lock SNR 12-23. Using u-blox u-center S/W. Through 6" snow and wood roof, it's -30°C outside so this is indoors.

So the problem is the LNA is not working. It's one of those generic laser-marked parts from the bowels of china, that 6-pin IC.
SMT marking code LY is MAX2659ELT+, uDFN-6. Measured 3.3V power and 0.8V at the RF input, re-soldered it but no change. The part is also obsolete.

To salvage this GPS board, I might use an (active) antenna with built-in pre-amp at the SMA, or try replace the LNA IC.
Right now, just using the patch antenna alone plopped onto the SMA J1 PCB pad works pretty good. The boards do have a smaller ground-plane than recommended.
« Last Edit: February 05, 2019, 05:08:37 am by floobydust »
 
The following users thanked this post: cdev, Johnny B Good

Offline cdev

  • Super Contributor
  • ***
  • Posts: 5082
  • Country: 00
Re: Ublox-NEO-M8N GPS navigation signal amplify module for arduino Rasppery PI
« Reply #7 on: February 05, 2019, 05:12:59 am »
Brrrr.. OMG..

Thats very very cold

Very useful graphic, thank you!

"Size does matter!"

Re LNA, check the ublox ucenter settings as well as you can, it may be configurable, you may have accidentally turned the LNA off or something. I only have 1 working ublox GPS so I have not spent much time with it but it has all sorts of settings there.

That antenna is passive, before the LNA. Oh, doh.. since the center of the antenna goes straight to the LNA, you may have zapped it with ESD by touching it. I always put something (tape is okay) over the center pin and any other exposed metal.
« Last Edit: February 05, 2019, 05:19:06 am by cdev »
"What the large print giveth, the small print taketh away."
 

Offline Johnny B Good

  • Regular Contributor
  • *
  • Posts: 190
  • Country: gb
Re: Ublox-NEO-M8N GPS navigation signal amplify module for arduino Rasppery PI
« Reply #8 on: February 05, 2019, 07:53:06 pm »
Patch antenna to the LNA input (the way the module is built) I view at most 1-2 satellites  >:(
If I move the patch antenna directly to the NEO-7M (SMA) input I get 12 satellites and good lock, SNR ~23 :o
If I connect 1/2 wavelength wire (9.5cm horizontal) directly to the SMA input I get ~4 satellites and good lock SNR 12-23. Using u-blox u-center S/W. Through 6" snow and wood roof, it's -30°C outside so this is indoors.

So the problem is the LNA is not working. It's one of those generic laser-marked parts from the bowels of china, that 6-pin IC.
SMT marking code LY is MAX2659ELT+, uDFN-6. Measured 3.3V power and 0.8V at the RF input, re-soldered it but no change. The part is also obsolete.

To salvage this GPS board, I might use an (active) antenna with built-in pre-amp at the SMA, or try replace the LNA IC.
Right now, just using the patch antenna alone plopped onto the SMA J1 PCB pad works pretty good. The boards do have a smaller ground-plane than recommended.

 Armed with jeweler's loupe, I took a really close look at my board, in particular the microscopic LNA which showed just 0.61v on the RF in pin. I tested for continuity from the top plate of the cermaic patch antenna through to the solder blob joint on the component side of the board and as expected, it was a full short (1mA test current BTW).

 There's just so little that can go wrong in the fabrication of a ceramic patch antenna, this just leaves, on the balance of probabilities, the LNA (faulty or possibly 'zapped') followed by a a pick and place issue during assembly (erroneously selected or loaded component values eg. Feeltech's 85 ohm attenuator pad cockup),  damaged/faulty smd passives, cracked circuit traces or faulty soldering, working from most to least likely eventuality. The LNA looks the most likely contender for lack of signal from the built in patch antenna. However, I've managed to make this somewhat academic, I'm very pleased to say! :-)

Following upon this close examination, I decided to set it up once more in the middle of my makeshift aluminium foil groundplane taking pride of place on the ledge of a SW facing window. As before, it spent much of its time showing no satellite signal of any sort (not even the more robustly PRN encoded ident modulation that allows you see which satellite is failing to deliver the goods for lack of a decent C/N ratio). Every so often a satellite or two would make its presence known for a few seconds or so before disappearing 'off the grid', leaving me to ponder about whether to try out another 'wire antenna'.

 Having gotten a feel for how badly it was performing with its built in patch antenna, I decided to try out that quarterwave piece of wire again, with the intention of pruning it back millimetre by millimetre to peak it for 1575MHz. Fortunately, I'd lost my "quarterwave antenna" so resorted to fabricating a replacement. Having noted your (seemingly inappropriate) use of a nominally halfwave antenna to achieve an improvement over original patch antenna setup, I elected to go for the more viable 3/4 wave option over that of the 1/4 wave one I'd tried initially.

 So, I duly made up my "3/4λ" antenna (14cm starting length to be pruned to resonance) and inserted it into the SMA socket, leaving the patch antenna face down on the groundplane, and backed my way out of the bay window corner (there's a desk, temporarily removed from an adjacent 'guest bedroom', that's been sat in the window bay for the past 6 months or more, making access a little awkward to this GPS favoured end of the bay windowsill).

 By the time I caught my first view of the satellite signal history window, it was already showing a rapidly growing list of acquired satellites with blue and mostly green strength bars which has remained more or less fully populated ever since, typically with 10 to 12 satellites in the green (see the attached pictures). I stood transfixed by a sight I'd thought I'd never get to see until that active GPS mag mount antenna I'd ordered finally arrived this Wednesday or Thursday (or, knowing my luck, more like next week or perhaps even not at all).

 It just goes to show you what a difference a mere extra 10cms of length can make!  ;D In the case of simple wire antennas, "Three" would seem to be "Magic" indeed as far as λ/4 base is concerned.

 Enjoy the pictures!  :)


PS

  The active antenna arrived in this morning's post (Wed 6th Feb). It was packed into an ordinary, unpadded, envelope with an invoice from Revtronic of Bradford W. Yorkshire. Despite the lack of padding, the antenna and its 5 metre length of co-ax (with male SMA plug) had survived its transit through the Royal Mail letter post system unharmed.

 To be fair, the antenna with its cable and SMA plug is a pretty rugged item to be be dispatching via the RM postal service so I wasn't put off by the lack of the more typical Ebay "Item in a padded envelope" experience for things like Arduino Nano and GPS module circuit boards.

 Needless to say, I had that antenna kit unpacked and connected up within 10 minutes of it coming through the letterbox (after verifying that the 5 metre cable was just a mere 2 or 3 cm shy of its claimed length) and I was observing an extra 4 or 5 satellites just by placing it on my aluminium foil groundplane right where I'd previously positioned the GPS module and its makeshift wire antenna.

 The signals look to be a good 10 to 15 dB stronger despite the 5 metres of rather thin (and presumably quite lossy at 1.5GHz) co-axial cable. I now typically show 14 to 18 locked satellites on the plot which, considering the restricted view of the sky through the window, seems a remarkably high number (and this doesn't include the three to five sats it hasn't fully locked onto).

 I've already configured the GPS module to output 10MHz on the PPS line and made use of the third harmonic to beat the FY6600 AWG against using my Kenwood TS140S HF transciever to monitor the frequency. No need to use the 500Hz narrowband CW filter to reduce modulation sideband interference as I have to do with the R4 LW broadcast on 198KHz, enduring the 800Hz sidetone as a consequence whenever I want to monitor the severity of the modulation interference which does detract from the timing accuracy using the S meter to mark the start/stop transition points. Instead, I've been able to monitor in AM mode listening to the more relaxing "waves breaking on a tropical beach" sounds (along with back ground beep noise interference from the USB signals modulating the carrier).

 Obviously, I need to buffer the output with Schmitt triggered gates and utilise an ultra low jitter PLL to eradicate such residual noises before filtering my 10MHz calibration signal to get rid of unwanted odd harmonics. Mind you, the third harmonic is proving useful in that I can more readily and immediately detect the changes by ear when trying to adjust to within a couple of ppb of error.

 Believe me, trying to trim a 0.1ppm TCXO to within one or two ppb is no easy feat at the best of times. TBH, I'm astounded at just how closely I did manage to calibrate it when my only stable reference was a MW broadcast on 1485Hz which has since proved to be about 13ppb adrift from its assigned frequency (dead stable as if it were using a GPSDO reference, only with a seemingly deliberate 13ppb offset applied).

 Whilst I should combine it with a good quality VCOCXO for a truly stable, jitter free reference, relying on its own built in 48MHz TCXO will suffice my immediate needs for the time being. I can now contemplate a GPSDO solution of my own design as a future project, armed with the data I'll now be able to gather from the current setup which is already revealing the half hourly ppb variations of the AWG's 50MHz TCXO frequency reference despite the constant 18.3 deg C room temperature readings. The beat period has dropped from 23 seconds (at 30MHz) down to just 4 seconds as I type - something I wasn't readily able to detect using the 198KHz Rubidium disciplined R4 LW broadcast before I had this GPS module setup.

 For anyone reading this thread with a view to making up a GPS based calibration reference on the cheap, the only problem with these boards seems to be the waste of space built in patch antenna which looks very much like an LNA chip issue. Since external active patch antennas with 3 or 5 metres of cable are so cheap (just £2.99 for mine from a local UK supplier!), there's no real need to let this put you off.

 If you do buy one that's deaf, it's easy enough to test whether an external antenna will be the cure by simply plugging a 14cm 3/4 wave stiff wire antenna into the SMA socket (or shove it into the hole were an absent SMA socket would have been soldered to the board - not necessary if it's going into a larger box with a panel mounted antenna connector you can solder onto the vacant connector position). If that lets you receive sat signals ok, there's no need to ship it back for an exchange or refund and you can order a proper external antenna, assuming you hadn't already 'jumped the gun' on this part of the project.


PPS

 I've just placed a ballasted biscuit tin on the roof of my bay window with the mag mount active antenna stuck on the lid to see what, if any improvement a clearer view of the sky would afford (the location doesn't offer a full 360 deg view, it's more like 200 degrees). The signals came up another 15 dB or so as one might expect. However, that's not really news. What's really newsworthy is the fact that when I disconnected the SMA connector and, as a kindness to the GPS module, I set it back up on the groundplane, instead of the expected total loss of satellite signals, I was still recieving almost the full complement, albeit some 20 dB or so down on what I'd been getting with the active antenna prior to relocating it into the clear.

 I don't quite understand why the integrated patch antenna should suddenly start receiving signals strongly enough to let me lock onto the satellites instead of merely indicating their presence below the 5dB C/N threshold. Was there an intermittent contact/dry joint or was it waiting for an externally connected active antenna to let the firmware/control software unlatch the on-board LNA out of a disabled state? Perhaps the on-board antenna just needs the user to connect a dummy amplifier load to the ext ant input to cycle it into an active state. I don't know, just hazarding a guess.

 Bad news, it looks like some sort of intermittent problem after all that. I've just disconnected the ext ant to see whether it would go back to locking onto sats at a reduced level from the built in patch antenna and all bar one completely disappeared with the one remainer showing a blue strength bar until that too disappeared. Never mind, this might offer sufficient clue as to what the actual problem might be to anyone with a mind to sort it out.

JBG

« Last Edit: February 07, 2019, 03:19:39 am by Johnny B Good »
 

Offline Johnny B Good

  • Regular Contributor
  • *
  • Posts: 190
  • Country: gb
Patch antenna to the LNA input (the way the module is built) I view at most 1-2 satellites  >:(
If I move the patch antenna directly to the NEO-7M (SMA) input I get 12 satellites and good lock, SNR ~23 :o
If I connect 1/2 wavelength wire (9.5cm horizontal) directly to the SMA input I get ~4 satellites and good lock SNR 12-23. Using u-blox u-center S/W. Through 6" snow and wood roof, it's -30°C outside so this is indoors.

So the problem is the LNA is not working. It's one of those generic laser-marked parts from the bowels of china, that 6-pin IC.
SMT marking code LY is MAX2659ELT+, uDFN-6. Measured 3.3V power and 0.8V at the RF input, re-soldered it but no change. The part is also obsolete.

To salvage this GPS board, I might use an (active) antenna with built-in pre-amp at the SMA, or try replace the LNA IC.
Right now, just using the patch antenna alone plopped onto the SMA J1 PCB pad works pretty good. The boards do have a smaller ground-plane than recommended.

 Armed with jeweler's loupe, I took a really close look at my board, in particular the microscopic LNA which showed just 0.61v on the RF in pin. I tested for continuity from the top plate of the cermaic patch antenna through to the solder blob joint on the component side of the board and as expected, it was a full short (1mA test current BTW).

====snip bit about the 14cm 3/4 wave antenna test====

PPS

 I've just placed a ballasted biscuit tin on the roof of my bay window with the mag mount active antenna stuck on the lid to see what, if any improvement a clearer view of the sky would afford (the location doesn't offer a full 360 deg view, it's more like 200 degrees). The signals came up another 15 dB or so as one might expect. However, that's not really news. What's really newsworthy is the fact that when I disconnected the SMA connector and, as a kindness to the GPS module, I set it back up on the groundplane, instead of the expected total loss of satellite signals, I was still recieving almost the full complement, albeit some 20 dB or so down on what I'd been getting with the active antenna prior to relocating it into the clear.

 I don't quite understand why the integrated patch antenna should suddenly start receiving signals strongly enough to let me lock onto the satellites instead of merely indicating their presence below the 5dB C/N threshold. Was there an intermittent contact/dry joint or was it waiting for an externally connected active antenna to let the firmware/control software unlatch the on-board LNA out of a disabled state? Perhaps the on-board antenna just needs the user to connect a dummy amplifier load to the ext ant input to cycle it into an active state. I don't know, just hazarding a guess.

 Bad news, it looks like some sort of intermittent problem after all that. I've just disconnected the ext ant to see whether it would go back to locking onto sats at a reduced level from the built in patch antenna and all bar one completely disappeared with the one remainer showing a blue strength bar until that too disappeared. Never mind, this might offer sufficient clue as to what the actual problem might be to anyone with a mind to sort it out.

JBG

 I decided to try an experiment with that LNA yesterday. Basically, after starting with a 1M resistor, followed by a 51K resistor and then a 10K finally landing up with a 3.3K resistor, I was able to pull the RF pin up from its 0.61v setting right up to 0.81v, just 0.02v short of the "Typical" value quoted in the data sheet (there's no min or max values quoted) when the /SHDN pin is high (as it is in this case) which actually seems to have restored some sensitivity to the built in Patch antenna.

 In view of the comments regarding getting usable signals with a plug in patch antenna that bypasses any on board LNA, I suspect it's still not entirely up to par but it seems to work about as good as that simple 14cm piece of wire did. That being the case, I can at least save some wear and tear on the SMA connector's centre pin from inserting and removing that wire antenna.

 I haven't actually looked for a replacement LNA chip, in spite of my recent shopping spree on eBay for clock multiplier chips and solder-less prototyping breadboards as well as SO/SSO to DIP adapter boards and other bits 'n' bobs I need to stock up on but having had some success at getting the LNA to function after a fashion, I'm going to look out for a replacement LNA chip in the hope that I can effect a more satisfactory fix.

 Anyhow, I've attached a couple of photos to show the results of my experimental repair work (bodge). These were taken over an hour after I'd cold started it on the patch antenna. I did cheat slightly by hooking the roof antenna up to kick start it after seeing no sats for 5 or 10 minutes. However, I have had it cold start ok (it takes a bad 5 to 15 minutes for it to start locking onto satellites) without such a helpful nudge.

 On previous occasions whenever I've disconnected the roof antenna to see what, if any satellites it might be able to cling onto, I've only seen it reporting the most marginal of signals before it gave up completely less than half an hour later. This is the first time I've seen it hang onto the satellite signals indefinitely and, more importantly, acquire them using only the built in patch antenna.

 It's currently locked onto 4 or 5 satellites as I type this. I'll leave it running overnight to see whether it manages to hang onto them and perhaps even add to its collection if the rain abates. I suspect the wet slates and brickwork isn't helping any right now.

JBG

 
 

Offline floobydust

  • Super Contributor
  • ***
  • Posts: 2710
  • Country: ca
How do you put an external antenna connector in the middle of the pcb trace run from the LNA IC output to the GPS module?
I'm suspicious of the whole idea - an idle LNA and SMA antenna both driving the same node, one like a stub, and the (antenna) PCB layout looks abysmal for a microwave design.
This might explain the odd results.
 

Offline Johnny B Good

  • Regular Contributor
  • *
  • Posts: 190
  • Country: gb
How do you put an external antenna connector in the middle of the pcb trace run from the LNA IC output to the GPS module?
I'm suspicious of the whole idea - an idle LNA and SMA antenna both driving the same node, one like a stub, and the (antenna) PCB layout looks abysmal for a microwave design.
This might explain the odd results.

 Just about my thoughts too!  >:(  I do wonder what the manufacturer was hoping to achieve with this ineffectual 'accessory'. As I've already mentioned, it would have been far better to save the expense of the on-board patch antenna and the LNA chip to cut the manufacturing costs to make a more profitable/competitive product.

 Even forcing bias voltage into the LNA with that 3.3k resistor didn't offer quite the hoped for improvement and I landed up finally removing the ridiculously microscopic LNA chip (WTF does it have to be so fekin' small?!!). BTW, the LNA's /SHDN pin is hard wired to the Vcc line so can never be shut down (with a NOT label on that control pin, I imagine it has an internal pullup which will float it to the Vcc rail anyway even if it was left open circuit whether by dry joint or design).

 Like you, I did wonder about its effect on the Ext Ant input circuit even though, rather surprisingly, just plugging a 14cm 3/4 wavelength wire antenna in, it managed to perform credibly well indoors, typically allowing the module to get locks on some 10 to 12 satellites most of the time. Since I've removed the LNA, my 3/4 wavelength antenna seems to perform even better. I was able to get almost as good a signal using a 1/4 wavelength antenna, unlike the first time I'd tried this. Whilst experimenting with such wire antennas, I also tested 5/4 and 7/4 wavelength antennas, all of which gave reasonable results but the 3/4 wavelength seemed to just have the edge.

 I like the convenience of the final GPSDO having a built in backup antenna but I think I'll leave that option to a later stage of the project. I did checkout replacement LNA chips and active patch antennas but decided I'd be better off not sullying the antenna input circuit with such interference, electing to buy a 5 metre antenna cable instead for the time being so have just ordered one which should arrive in about a week's time. That will extend the existing 5 metre cable run to let me roam further away from the window bay and onto my test/work bench. It'll probably knock about 5 or 6 dB off the signals (I haven't bothered looking up the loss figures for RG174 at 1.5GHz so that's just a guesstimate and I can afford even a 10dB drop anyway).

 In the meantime, I've noticed the onset of regular phase shift jumps in the 2MHz signal at about two per second comparing it on the scope against the 2MHz sinc wave from the FY6600. Prior to that, I couldn't make up my mind as to whether the previous more subtle phase shift jump behaviour (very small jumps once every 2 to 5 seconds) was an artefact of the DSO technology or an innate behaviour of the GPS module's disciplining of its own internal 48MHz VCTCXO. I've now come to the conclusion that it's innate to the GPS module and now only shows such obvious disciplining on account of my reseting the module back to defaults causing it to lose its VCTCXO calibration settings.

 Hopefully, it'll recalibrate after a day or two of good satellite signals. It'll be interesting to see how it free runs when I induce it to lose satellite lock. I did originally have it set to output 3Hz at 7% duty cycle when not locked so as to give me an indication on the PPS LED but I've reprogrammed it to output 2MHz at 33.3333% duty cycle instead so I can compare the free running frequency against the locked frequency output.

 After observing the rather sedate free running drift between the sig genny and the GPS when unlocked and the result of the phase jumping to hold it on frequency when locked, it became all too obvious as to the cause of those previously subtle jumps I'd observed the day before I started trying to sort out that pesky LNA chip. Hopefully, it'll get its calibrated status back again in the next day or so and I'll be able to reconsider whether I can forego the extra expense and complications of an add on VCOCXO for my 'Poor Man's GPSDO' calibration reference. It's all a learning experience and if I do upgrade to a VCOCXO it won't be just because "It's the done thing".  :)
 

Offline Johnny B Good

  • Regular Contributor
  • *
  • Posts: 190
  • Country: gb
 A quick follow up to my previous post.

 In no particular order. What I surmised about the /SHDN pin having an internal pull might possibly be wrong. A very close look at the vacated LNA placement shows that there is a link between pads 4 and 5 (Vcc and the /SHDN pins). The data sheets (two different ones) say nothing whatsoever about the /SHDN pin other than one of those data sheets showing that the Digital Input Current is a maximum of 1 microamp immediately below the entries for the Digital Input Logic high and low voltage levels.

 You need to use deductive reasoning to work out that the only pin in the package that meets that description is the /SHDN pin which is labelled on the pin out diagram on the previous page as the "Logic Control". Since it doesn't mention the polarity of this 1 microamp, it looks very much as though this pin can't be left to float on its own and needs to be tied to a logic 0 or 1. The link between the pads for pins 4 and 5 serves to satisfy this requirement to permanently enable the LNA. IOW, a dry joint, contrary to my initial assumption, could cause erratic behaviour if not actually allow it to shut itself down.

 As a further check, I managed to find a page describing a development board designed to allow testing by potential customers which confirmed the need to direct pin 5 to ground via a 10K resistor or directly to Vcc (no simple pullup resistor with a shorting jumper to ground shutdown option here). Although it's just about possible for a failed solder joint for pin 5 in manufacturing, it seems extremely improbable. Very likely, the LNA on mine was simply a bad unit (witness the low .61v on the the RF in pin 3).

 TBH, I'm not at all enthusiastic about getting the on-board patch antenna working since the way the LNA output has simply been linked into the antenna socket line seems so much like a horrible kludge, it might never had worked anyway. Better safe without it than to reinstate a working LNA chip is how I feel about it, especially after my attempts to get it to lose satellite lock on the testbench this afternoon to check the free running drift.

 I did eventually lose satellite lock but it took a good ten to fifteen minutes of waiting for the 7 or 8 satellites to suddenly vanish after seeing the signal strength bars hovering around the 8dB mark by plus or minus 5 dB with no antenna of any sort plugged in. Presumably, unplugging the antenna will result in an instant loss of signal once I've got it assembled into a nice metal box.

 Witnessing this response made me wonder just how much contribution was being made by that patch antenna even when I'd added the 3.3K bias resistor to ginger up the LNA when, presumably the exposure of some 18mm's worth of stripline was enough to still pick up very weak but usable signals.

 Having finally gotten the module's TCXO to free run, I was able to confirm my initial assumption about the phase jumps. Furthermore, the frequency and level of shift hadn't improved overnight so I'm not so sure about the supposed self calibration. I think the variations in the disciplining rate are more down to room temperature rather any recalibration effect. It might be keeping the average frequency spot on but with far less subtlety than a GPSDO using a VCOCXO with a very long time constant PLL control circuit in charge of the gentle persuasion form of disciplinary action. I can see myself shopping for a decent, yet cost effective, VCOCXO module some time soon (but only after I've done some more experiments).

JBG
« Last Edit: March 17, 2019, 06:33:38 am by Johnny B Good »
 

Offline Johnny B Good

  • Regular Contributor
  • *
  • Posts: 190
  • Country: gb
 Since my last update,  I've done some more investigation into acquiring a cheap patch antenna but it seems my previous purchase of that active mag mount antenna with a 5 metre cable terminated with an SMA male plug for just £2.99 from a UK supplier was just pure dumb luck on my part for snagging such a bargain. Even the cheapest active patch antennas with just half a metre of cable were over 5 quid and my previous bargain purchase gave me a rather jaundiced regard of such "bargains". All I was really after was a cheap passive patch antenna just like the on-board one which I could plug into the ext ant socket just to see how effective it can be compared to an active antenna.

 I then had the bright idea of trying an eBay search for passive patch antennas but the only suppliers at sub three quid pricing were all in China, Hong Kong or Singapore with delivery times in the 12 to 33 day ball park. However, I did spot the fact that all these patch antennas sport a peel away protective backing for the adhesive film that's used to secure them to the board so decided to have a go at peeling the uBlox module's patch antenna off the board after unsoldering the feed pin with a desoldering suction tool and eventually, with the aid of a craft knife, managed to lift it off the board without too much damage to the through plated hole that I hadn't entirely managed to suck all of the solder out of (and also without damaging the patch antenna).

 The double sided adhesive layer had chosen to remain attached to the board but it wasn't too difficult to peel it away, leaving the board nice and clean, not only of any sticky residue but also, and more importantly, free of its encumbrance of an ornamental antenna. Though that side of the board was mostly ground plane, there were a few circuit traces to be seen which may well have been interfering with the satellite signals (I have no other explanation for what followed).

 Having liberated the antenna, I was now able to repurpose it as a slightly more remote patch antenna option via a short 15cm or so length of micro-coax I'd been holding onto for the past 25 years. It's PTFE insulated and you can see the silvery tightly woven outer braiding under the translucent PTFE insulation. The centre core is stranded and covered with similarly translucent PTFE insulation of about 2mm in diameter. I know it's PTFE on account it's a real pig to strip using normal wire strippers since unless you very carefully score the insulation with a sharp blade where you'd like it to part company, it simply stretches rather than snap apart like PE would.

 Anyhow, I soldered it onto the feed connection pin and soldered the braid onto the bottom plate which seems to act like a solder magnet (the slightest touch with a blob of molten solder causes it to immediately wet - no problems creating a reliable solder joint, on the contrary, the problem is more the issue of creating accidental solder joints!). Not having any spare SMA plugs of any description, I simply formed the end of the coax into a coax like plug shape that I could insert into the SMA socket to try my freshly liberated Patch antenna in its new configuration.

 Bearing in mind that I'd already tried linking the antenna pin straight across to the incoming antenna strip line feed with a 4mm jumper link to no effect after getting rid of the LNA, I was astounded to discover that it actually outperformed my 3/4 wavelength antenna by a significant margin.

 In the end, suspecting that the shield contact in the socket might have been allowing to act like a 3/4 wavelength antenna that just happened to have a lump of ineffective patch antenna dangling off the end, I soldered it onto the ground and cente pins of the SMA socket to prove the signal was being guided in from the patch antenna's feed point rather than stray signal pickup on the screen being conducted in (also, it eliminates the problem of the push in connection falling apart at the slightest provocation).

 I've now discovered that it's virtually impossible to attenuate the satellite signals enough to lose lock by grasping the antenna in my hands, at least in my first floor work room - I have to go to the ground floor before I can induce sufficient loss of signal. Recovery now only takes a few seconds, unlike before when I'd gingered up the LNA with a 3.3K bias resistor where I'd have to place it in view of the sky and wait quarter of an hour or more.

 So, it seems the Patch antenna is just fine and possibly even the LNA (although I still have doubts about that). The best explanation I can come up with is that it was just being swamped by interference from those circuit traces immediately underneath the base plate which only relied upon capacitive coupling to the ground plane (and those circuit traces). Given a board lacking such circuit traces on the underside, with the whole area being ground plane, it might have worked passingly well but, I suspect the designers may have simply overlooked the presence of these traces which may not have existed in an earlier design of these boards.

 I learned several new things today. One, these patch antennas are only held in place by a thin double sided adhesive pad and a single solder joint to the feedpoint. Two, they absolutely rely on not being sat on top of sources of strong interfering signals. Three, with a little bit of finnessing it's entirely possible to remove them intact from and without damage to the board given a suitable craft knife as a lever and effective use of a quality vacuum pump desoldering tool to clear the solder out of the through plated hole. And, four, there's nothing wrong with the patch antenna (as I said earlier, there's not a lot can go wrong with a ceramic patch antenna).

 If the manufacturer had only realised the futility of such a patch antenna on this particular design of the board, he could have saved himself needless expense and the grief of handling the returns by customers less adventurous than you and I. Obviously, with the patch antenna now soldered to the antenna socket, it will interfere to some extent with the signal feed from an external antenna but when that feed is a good 20 to 30dB stronger, it doesn't, as I've already verified, really matter so much. However, once I do get hold of the short SMA flyleads I've ordered, I'll be able to sacrifice one as a plug in antenna lead.

 If I do incorporate the patch antenna back into the final design of the GPSDO, I'll be looking for a much better way to switch the feeds to eliminate the possibility of destructive interference between the built in and the external antenna feed options.

 Now that I've solved the riddle of the ornamental patch antenna, I can concentrate on my next step which is to confirm whether that clock multiplier chip will actually generate output frequencies below the minimum 14MHz as stated in the data sheet.  >:(

 That little fact doesn't get mentioned in the feature headline of "Up to 190MHz (at 5v or else 120MHz at 3.3v). It's only after seeing mention of a minimum of 5MHz crystal oscillator frequency or 2MHz clock input that you then see mention of this 14MHz lower output frequency limit which is a little contrary to the concept of 2MHz clock input minimum with a PLL that can be programmed with multiplier as low as 2. I guess it's high time I did some testing to see just how strict this 14MHz minimum limitation actually is.

JBG
« Last Edit: March 18, 2019, 06:11:27 am by Johnny B Good »
 

Offline AlienRelics

  • Contributor
  • Posts: 37
  • Country: us
    • OlyMEGA
I am on a similar search for accuracy. What I am thinking of doing is using a GPS module with TCXO as the clock signal source to replace the crystal on a Teensy. Then when it isn't locked, I still get TCXO accuracy, but when it is, ppb GPS timing accuracy. I was thinking I'd program it for 8MHz output and double it.

Thank you for the tip on the clock multiplier. Much simpler than how I was planning on doubling the clock.

I have a DIY Mall branded USB GPS module labeled G28UF8FUSB, which came up in a lot of Chinese language searches. They seem to indicate that this does have a TCXO on board. There are pads on the board labeled for a separate connector including PPS. This will do for now and is simpler to reprogram with the built-in USB.
 

Offline Johnny B Good

  • Regular Contributor
  • *
  • Posts: 190
  • Country: gb
I am on a similar search for accuracy. What I am thinking of doing is using a GPS module with TCXO as the clock signal source to replace the crystal on a Teensy. Then when it isn't locked, I still get TCXO accuracy, but when it is, ppb GPS timing accuracy. I was thinking I'd program it for 8MHz output and double it.

Thank you for the tip on the clock multiplier. Much simpler than how I was planning on doubling the clock.

I have a DIY Mall branded USB GPS module labeled G28UF8FUSB, which came up in a lot of Chinese language searches. They seem to indicate that this does have a TCXO on board. There are pads on the board labeled for a separate connector including PPS. This will do for now and is simpler to reprogram with the built-in USB.

 I was just about to follow up that last post of mine to refine various of my then untested hypotheses before discovering your follow up post. If you're planning on buying a 3N502 clock multiplier chip as per my suggestion, don't bother (unless you need to add an external 10MHz reference socket to something like an FY6600 which uses an internal 50MHz reference).

 Sadly, it doesn't solve the issue of the "Leap 48MHz TCXO clock period adjustments" it uses to keep the PPS synchronised to GPS or UTC time (regardless of the programmed output frequency). If you can monitor the third harmonic of an 8MHz PPS clock setting using a multimode HF receiver or transceiver using the narrow band FM mode, you can clearly hear these "leap cycles" as ticks every few seconds to 3 or 4 a second, depending on the actual frequency of the 48MHz TCXO clock. The 3N502, good as it is for providing very low jitter clock frequency multiplication, can do nothing about this peculiar source of phase noise as I eventually learned from my experiments using the 3N502 chip after finally getting round to setting one up in a SOT8 to 8 pin DIP adapter board to plug into a solderless breadboard lashup yesterday.

 It was only this afternoon, after completing the last tests with this rather fragile lash up, that I finally came to realise the scale of the problem and started a fresh search for a suitable VCOCXO module, not because "It's the done thing" but because it's the only sensible way to proceed in this case if you want to go beyond substituting for an absent 10MHz WWV frequency calibration source.

 Whilst the raw 10MHz PPS output option is a very good substitute for the now impossible to receive WWV 10MHz broadcasts (sunspot minima period) to effectively trim an HF rig's internal reference oscillator to within half a Hertz, it's certainly no use as an external 10MHz clock reference for communications gear and arbitrary wave function generators and the like. The plain fact is that there's no viable alternative to that of using a VCOCXO, disciplined by the GPS module as your low phase noise carrier source.

 You might get away with the much cheaper option of a VCTCXO but the issue of "retrace" rears its ugly head in this case. Still, a VCTCXO might be worthy of consideration since, other than spending real money on a new VCOCXO (or lucking out on a cheap used unit), it should be a straight forward upgrade option, assuming you've made allowances, spacewise, in your home brewed GPSDO to take the older more bulky VCOCXO modules. You'll have already dealt with the issues of designing and assembling the long time constant PLL circuitry by which to discipline your VCTCXO module which will certainly be just as effective a way to deal with those pesky phase correcting jumps in the raw PPS output.

 I had no Idea what a "Teensy" was until I thought to try googling it just now. :-) If, as I gathered from the results, you only need it for precision timing where these "leap phase adjustments" shouldn't be an issue, then you can use the raw PPS output. If you'd used a u-blox NEO 7N or M8N, you could even have done away with the clock multiplier chip since they can be programmed to output 16MHz directly.

 However, as I've just discovered by testing this, with a divide by 3 factor involved, it lands up alternating between a duty cycle of 33.333 and 66.666 percent on every leap phase adjustment! I tried it with an 18MHz setting which works fine, giving a similarly jittery trace to the 10MHz one (ISTR Scully of scull.com fame discovering the limit being just over the 19MHz mark, hence my bothering to test the 16MHz option).

 Wow! This GPS module never ceases to surprise me. :-)  This issue with alternating duty cycles every few seconds to 3 or 4 times per second, may, or may not, be a significant factor in your application. I think, if your GPS module works similarly (and with a 48MHz TCXO), you might be needing that 3N502 mulitplier chip anyway. The resulting phase jumps correspond to 30 deg at 8MHz which, using the 3N502, will become 60deg jumps at 16MHz. Again, this could prove problematic in your usage case (but less so than alternating between 33/66 duty cycle on every corrective phase jump when the module is programmed to generate the 16MHz directly on its PPS pin).

 Continuing, for the benefit of anyone else interested in this particular NEO M8N module, I did eventually find out why the built in patch antenna wouldn't deliver the goods. Basically, it amounted to the manufacturer neglecting the presence of circuit traces scattered within the ground plane upon which the antenna had been glued, using the very common (as I eventually discovered) method of a very thin adhesive pad.

 It was only when I started searching for cheap passive patch antennas to test with, that I spotted the protective peel away backing tab displayed in all the illustrations. It was this which inspired me to desolder the feed pin joint so I could use a thin craft knife blade to prise my "prize", the free patch antenna, away from the module. I hadn't been completely successful with the desoldering tool but the only collateral damage was to the through plated hole used by the feed pin which is of no further consequence anyway, with my having discovered the true cause of the antenna failure in what lay out of sight beneath it (pictures to follow).

 I soldered a short 15cm length of RG174 type co-ax to my prize, formed the other end of this co-ax into a co-ax plug shape, plugged it into the antenna socket and got even stronger signals than my 3/4 wave wire antenna had given me. Indeed, after temporarily soldering it to the socket terminals, it was almost impossible to lose satellite lock by shielding it with both hands in my first floor office/workshop, now 'man cave'. To do that I had to go to a ground floor room in this three storey Victorian house and it now locks back onto the satellites in a matter of seconds as opposed to quarter of an hour when I'd bodged a solution to the 'deaf LNA' with a 33K bias resistor to raise the RF input pin bias voltage from its low of 0.61v to 0.81v to approximate the 0.83v 'typical' figure given in the data sheet, which bodge had allowed the GPS module to just about manage to receive usable satellite signals from the built in patch antenna.

 This bodge solution still couldn't match the results I was able to get just by sticking a 14cm 3/4 wave piece of wire into the antenna socket so I landed up removing the suspect LNA chip altogether. I had hoped to discover a dry joint but gave up trying to solder it back onto the board when I realised the utter hopelessness of my achieving such an impossible task in the face of such insane micro-miniaturisation (the damn thing was barely bigger than the smallest smd resistor FFS!).

 I'm guessing that previous revisions of this module hadn't been cursed by such signal carrying circuit traces on the 'back side' of the board, allowing such a misplaced patch antenna to work tolerably well. Manufacturing inertia seems to have led to this egregious error in mounting a patch antenna right on top of strong sources of RFI, rendering it as nothing more than an ornamental adornment.

 The patch antenna base plate relies on capacitive coupling alone to the module's ground plane which works well provided that it doesn't also capacitively couple to strong sources of RFI as was obviously the case here. In all likelihood, there wasn't anything wrong with the microscopic LNA chip or any of the other components involved but I'm unlikely to know despite keeping hold of that much abused LNA chip in a small glass jar as testament to the folly of such extreme miniaturisation.

 Having done pretty well all of the research on this particular GPS module that I'd originally been seeking advice on, my conclusion is that, aside from the waste of a perfectly good patch antenna and, most likely, the microscopic LNA chip as well, since it's using the NEO M8N GPS module which is very well specified for the task. It's RF sensitivity is such that the use of an on board LNA with the patch antenna is hardly necessary since it can retain lock just from the stray pickup on the strip line connection to the EXT ANT socket, receiving almost as many satellites with a passive patch antenna indoors as it does with an externally placed active patch antenna.

 I think the only reason why the RFI was only upsetting the built in patch antenna and not the receiver in the u-blox module itself was simply because it had overloaded the LNA chip so severely it had totally desensitised it to the point where very little signal was able to get passed onto the output pin. The low bias voltage on the RF input pin was very likely just the side effect of this overloaded condition. It seems just dumb luck that this mistake was so huge as to effect damage limitation to just that of the patch antenna feed, preserving the main receiver section for externally connected signal sources.

 If this hadn't been the case, the whole module would have been 'deaf' to signal reception on the EXT ANT socket as well, leaving the only (reversable) solution as removal of the output coupling capacitor link to the main ANT input line to break the signal path. Luckily, this wasn't the case, allowing a simple experiment with a 14cm 3/4 wavelength wire antenna to prove the rest of the GPS module as being fit for purpose.

 There may well be other brands and/or models of GPS modules similarly afflicted by this "Deaf Integrated Patch Antenna" syndrome so my experience of the problem might well apply to those as well. With care, such modules could equally benefit from a "Patchectomy", allowing the excised patch antenna to be repurposed as a separate plug in "Test antenna" via a short 15 to 30 cm fly-lead with co-ax plug of your choosing.

 The trickiest part of the patchectomy operation ime being the successful desoldering of the one and only through plated hole connection to the antenna feed pin. As long as the antenna itself isn't damaged, some damage to this through plated hole in the PCB is of no consequence since you'd have to be insane to even contemplate fitting a replacement antenna back onto the board once you've tested the extracted antenna with a soldered on fly-lead connection (aside from gross physical damage, there's little that can go wrong in such a simple design of antenna).

 Rather intriguingly with the u-blox control centre software, I noticed a section on disciplined oscillator options (external and internal oscillators) with half a dozen DAC types and protocols on offer, suggesting that, with the right hardware in place to control a VCXO (TC or OC type) it might be possible to create a simplified GPSDO without resorting to separate microcontroller and PLL hardware. Unfortunately, the "Help" file has been no help in explaining this or any of the other, many and varied options available in the configuration menu. I guess some more search engine activity is called for.

 One other curious observation I've made relates to the additional 5 metres of RG174 antenna extension cable which arrived a few days ago to supplement the 5 metre external active mag mount patch antenna I've got placed on the flat roofed bay window of my 'man cave'.

 I'd checked out the dB loss rates per hundred feet/metres and had extrapolated from the 1GHz loss figures an attenuation figure of some 5 to 6 dB which I'd thought was on the optimistic side. When I did comparative tests, the extra five metres of RG174 seemed to be only offering an additional 2 to 3dB of loss. This was a far smaller loss figure than I was expecting. I'm absolutely certain that a 5 metre length of even the finest RG174 can't have less than 5dB of loss at 1.6GHz which rather suggests that the input impedance is significantly higher than the nominal 50 ohm rating of RG174 cable, allowing for a mismatch induced voltage magnification at the receiver front end. Possibly the active antenna's five metre feeder mismatch loss was causing a modest voltage drop of around a dB or two with the additional 5 metre cable introducing a mismatch induced voltage magnification of another dB or two which could account for the overall reduction in observed loss compared to the expected 5 or 6 dB based on a reasonably close impedance match between the receiver input and the co-ax.

 I suppose it's just possible that the apparently haphazard LNA chip T connection to the antenna input line may have been the result of a design to keep the impdances close to the nominal 50 ohm match to the antenna feeder cable, possibly putting it at the lower impedance tolerance limit until I disconnected the LNA from the antenna input circuit, raising the impedance sufficiently for just such an unexpected result to arise. I guess I'll be able to get a more reliable assessment of the true cable loss when I can get hold of a 10 to 16 dB SMA attenuator pad to plug into the module's antenna socket to provide a more consistent 50 ohm termination for testing the antenna cable losses.

 Normally, such additional attenuation would never be deemed a benefit in a GPS antenna setup hence my not having such items to hand. I don't think my collection of F plug/socket attenuator pads are rated beyond 1GHz so rather than buy F to SMA connector adapters, I think I'd be better off ordering some SMA pads instead if I were to be so inclined as to do such testing (I might if the price is right - it's not vital, just a case of satisfying my curiosity over this mystery of a seemingly ultra low loss antenna extension cable).

 Pictures as promised. The first three are self explanatory. The last two IMG_9484 and IMG_9486 are the "Before" images for comparison.

JBG


 
« Last Edit: April 19, 2019, 01:26:46 am by Johnny B Good »
 

Offline AlienRelics

  • Contributor
  • Posts: 37
  • Country: us
    • OlyMEGA
"If you'd used a u-blox NEO 7N or M8N, you could even have done away with the clock multiplier chip since they can be programmed to output 16MHz directly."

The documentation I'd found on the U-Blox website seems to indicate a max output of 10MHz. I have a module here that is labeled G28U8FUSB. It shows up in the U-Blox software as an 8, and on searching for the number I get a lot of Chinese language sites. All that I've looked at contain the letters TCXO, so that should make this an M8N under the hood.

I had nothing here to connect it to, but when I set it to 10MHz, the PPS LED dims as you'd expect. When I tried setting it to 16MHz, it goes back to blinking once per second. I'm uncertain what that means. I have another GPS module that is labeled M8N, I will try with that, too. I need to connect these up to a 'scope to see what is going on.

Thank you very much for all of your measurements. Very interesting to know about the glitches and possible method of stabilizing by VCTCXO and slow PLL.
 

Offline Johnny B Good

  • Regular Contributor
  • *
  • Posts: 190
  • Country: gb
"If you'd used a u-blox NEO 7N or M8N, you could even have done away with the clock multiplier chip since they can be programmed to output 16MHz directly."

The documentation I'd found on the U-Blox website seems to indicate a max output of 10MHz. I have a module here that is labeled G28U8FUSB. It shows up in the U-Blox software as an 8, and on searching for the number I get a lot of Chinese language sites. All that I've looked at contain the letters TCXO, so that should make this an M8N under the hood.

I had nothing here to connect it to, but when I set it to 10MHz, the PPS LED dims as you'd expect. When I tried setting it to 16MHz, it goes back to blinking once per second. I'm uncertain what that means. I have another GPS module that is labeled M8N, I will try with that, too. I need to connect these up to a 'scope to see what is going on.

Thank you very much for all of your measurements. Very interesting to know about the glitches and possible method of stabilizing by VCTCXO and slow PLL.

 In the "What do you want from a GPSDO?" thread, I posted the results of those tests along with a few more tests which revealed that my M8N module (at least) could generate a 19.2MHz output and that any frequency options below that which relied only on odd integer factors in the divisor needed to produce them from the 48MHz TCXO were destined to produce alternations in the duty cycle on every corrective phase jump (eg 5 and 7 divisor values were tested for 9.6 and 6.857143 MHz outputs).

 If you're going to use this to produce a very long term accurate clock for your "Teensy", you'd best avoid such gross changes in duty cycle or any relatively large phase shifts, especially if you're overclocking it. Obviously, bus clock signals can tolerate some phase noise or else PC MoBos wouldn't  offer any spread spectrum options just to reduce the impact of any RFI leaking out of the box.

 If you just need a relatively clean and stable clock with the long term stability of an atomic clock, the good old fashioned long time constant PLL will effectively filter out these phase disturbances. Modern GPSDOs tend to use a microprocessor to control the disciplined oscillator since this makes it possible to keep the oscillator on frequency in a hold over mode whenever it loses lock with the satellites. You may not require this feature if you can be confident of rarely suffering a total loss of lock on the satellites (at least for any longer than a minute or two at a time as a rare event or three each day).

 You'll certainly learn a lot by examining the waveforms with a 'scope. I'd not have been able to work out what the GPS module was doing without one. For this work, there's no doubt that the 'scope is an indispensable tool.

JBG

« Last Edit: March 28, 2019, 01:38:28 pm by Johnny B Good »
 

Offline AlienRelics

  • Contributor
  • Posts: 37
  • Country: us
    • OlyMEGA
This gentleman also talks about the GPS clock jumping a cycle to correct at intervals.
https://youtu.be/B_Y971uY3Nk

I have a few VCTCXO oscillators. I was thinking about having the Teensy 3.x itself count the period from the PPS output and use one of the 12 bit DAC outputs to adjust the VCTCXO. I'd use long term averaging in software.

I could even have a D18B20 temperature sensor stuck on the oscillator and have it store average correction factors vs temperature, then have it use those values when GPS lock isn't available. This will also allow me to use any garden variety GPS module.

I found the thread you mentioned, I'll be reading that over carefully. Thank you.

Of course, an Arduino also uses a 16MHz crystal, so this applies equally to that. Then you'd need an external DAC, but those are not expensive.
 

Offline Johnny B Good

  • Regular Contributor
  • *
  • Posts: 190
  • Country: gb
@ AlienRelics

 The gentleman in that video was referring to a Rubidium standard which might be using a similar approach to decimating its frequency down to the 10MHz reference output. However, that seems unlikely in view of the fact that that would introduce an unwanted glitch if used to drive the 10MHz external oscillator reference ports that are now adorning most, if not all, of the higher quality amateur and professional radio gear as well as high end test and measurement kit.

 He's either got a problem with his Rubidium oscillator or else one with his connecting cables. Anyway, that's beside the point. I have to say that his attitude to the state of that VCOCXO module seemed rather casual to say the least. That's one module that was not given the TLC it deserved in its extraction from the scrapped kit it had been salvaged from, looking as it did, like it had been knocked into the skip with a sand wedge. Much gentler abuse than that so often proves detrimental to the stability of the crystal resonator as to make it useless as a VCOCXO. That's yet another reason not to trust a far eastern based Ebay supplier of such (ab)"used" VCOCXOs and look to the likes of Digikey for a new undamaged unit, even if it means parting with a non-trivial lump of cash that's likely to account for two thirds or more of the total BOM costs in a "cheap" home brewed GPSDO project. :'(

 There is some truth behind his thoughts on the pre-aging aspects of buying second-hand VCOCXO modules but the evident physical damage to that particular example rather negates that possible benefit. Even when salvaged and shipped with the TLC such modules deserve, there's still the unanswered question as to how much margin is left for further ageing before it drifts beyond range of the Vfc  tuning voltage limit.

 Although I have seen claims being made that the latest VCTCXO modules have stability performance levels comparable to that of VCOCXO modules, I take such claims as simply being advertising guff, not worth any credibility on the basis that all advertising is simply the art of lying by omission taken to the extreme.

 The reason why the VCOCXO option is invariably chosen over the VCTCXO option when designing a GPSDO is simply on account that a TCXO cannot compensate for "retrace" which the OCXO can suppress simply by holding the quartz resonator at a constant temperature, neatly sidestepping this question of "retrace" altogether. If you are going to go to all that trouble of marrying a GPS module to a voltage tunable XO, then you might as well choose the most frequency stable variant and make all that effort pay off rather than take half measures with a mere VCTCXO. If a job's worth doing, then it's worth doing well or else not at all.

JBG
« Last Edit: March 29, 2019, 03:37:48 am by Johnny B Good »
 

Offline texaspyro

  • Super Contributor
  • ***
  • Posts: 1267
NO!  All OCXOs have retrace issues.  For some, they may be small, but they are there.

To buy a new OCXO with the specs of some $30 Ebay OCXOs can set you back $1000+ in small quantities.   It can be worth it to buy a few cheapies and sort them for coefficient of natural goodness.  I've bought dozens of Ebay OCXOs and only had a couple of real stinkers.  I got a few that were amazingly good.   You can expect the same even with new-from-the-factory OCXOs.  OCXOs are like snowflakes... no two are alike and they all have their own personalities.
 

Offline Johnny B Good

  • Regular Contributor
  • *
  • Posts: 190
  • Country: gb
NO!  All OCXOs have retrace issues.  For some, they may be small, but they are there.

To buy a new OCXO with the specs of some $30 Ebay OCXOs can set you back $1000+ in small quantities.   It can be worth it to buy a few cheapies and sort them for coefficient of natural goodness.  I've bought dozens of Ebay OCXOs and only had a couple of real stinkers.  I got a few that were amazingly good.   You can expect the same even with new-from-the-factory OCXOs.  OCXOs are like snowflakes... no two are alike and they all have their own personalities.

@texaspyro

 Thanks for that insight. I'm no expert (how can I be when I lack expertise - I only developed my interest in the subject just a few short months ago on acquiring a Feeltech FY6600 AWG.  :)  ).

 BTW, before going any further, I'd like to make it crystal clear that I currently have no interest whatsoever in OCXO modules since they're of no use whatsoever in a GPSDO. The only type of OCXO that I'm trying to get my hands on, and it seems I can't emphasise this point enough, is the voltage tunable variety, aka a VCOCXO. It's been bugging me that all the helpful advice so far on buying cheap oscillator modules on Ebay has taken the form of links to sources of OCXOs rather than VCOCXOs.

 Apologies for being so forceful in emphasising this point but I'm trying to break out of this circular furrow we appear to be ploughing like ants trapped in a looped pheromone trail of death by exhaustion. For all our sakes, enough of these OCXOs already! I'd like to move towards getting hold of a VCOCXO now, if you don't mind.  :)

 I suppose I should have qualified my statements about "retrace" (caveats and all that) to account for the fact that it's not entirely eradicated by eliminating, in an ovenised XO, the one factor that temperature compensation strives imperfectly to compensate for, namely frequency variations with temperature. I suppose I should have emphasised that what was being almost entirely eliminated were the temperature variations which are the main contributor to this issue of retrace.

 I am aware of the ageing process which can exhibit 'stiction' like micro jumps even with ovenised quartz resonators but this retrace issue is far less of a problem than that of allowing the quartz resonator to track variations of ambient temperature and then hope to compensate for this with a calibrated tuning voltage control that relies solely upon temperature sensor input which is akin to trying to compensate for a rubbing voice coil by pre-distorting the signal in the vain hope of cancelling this defect out of the system.

 Also of course, I assumed that it was understood that I was referring to a steady state condition where the VCOCXO is never powered down and allowed to temperature cycle on a regular basis, which process reintroduces the issue of 'retrace' in even the finest of ovenised oscillator modules.

 No system of control can ever be absolutely perfect but, in this case generally speaking,  the VCOCXO solution just happens to be significantly less imperfect than the VCTCXO one. I take your points about manufacturing variations in the quality of brand new units versus the lottery of buying second hand (ab)used high quality double ovened (VC)OCXOs but trying to second guess which is the best option in this case seems to me to be a bit of mug's game unless you can get hold of such parts for chump change. Sadly for me, that doesn't appear to be an option.

 Lacking the experience of buying second hand used units by the box load as I do, I have to rely on the probability of my getting hold of a good unit being best served by the purchasing of a brand new unused VCOCXO rather than a used unit of unknown  provenance from some trader who is essentially a scrap dealer in China. I'd prefer to purchase from American suppliers but the shipping charges to the UK tend to kill such deals.  :(

 The thing with Digikey UK is that for orders worth £33 or more, the shipping on such parts is free, so not only do I acquire a brand new item at a reasonable price, I also avoid the problems of trying to collect on a Chinese warranty should the need arise. However, since I'm now going to be visiting a major radioham rally in just over four week's time where I'm hoping to acquire electronic components in face to face deals with actual traders the good old fashioned way before the internet spoilt everything, I think I'll hang fire on blowing some 60 or 68 quid plus vat on one of the two Connor-Winfield DIP packaged 5v sine output VCOCXOs that I've currently got my sights on (the OH100-61005SV-010.0M and the OH200-71005SV-010.0M respectively). I just might luck out at the rally next month so it's worth the wait (I'm the master of procrastination, I'll have you know!  :)  ).

 The reason why I've chosen 5 volt parts is simply that I've decided to use cheap and disposable pound shop (2 quid actually!) 5v 2.1A USB wallwarts along with cheap and disposable pound shop 1200mAH battery banks to ride out any hour long mains outages which saves either wasting power in a 3.3v LDO regulator to push the required 1A heater warm up current or the complication of adding a switching converter to reduce the load on the 5v rail to a less demanding 700mA.

 Squeezing a small mains voltage smpsu into a GPSDO box to make it all self contained, just doesn't seem very optimal to my mind. Why would you want to complicate the business of replacing a blown psu board (cannibalised out of a cheap wallwart anyway -cheapest source of small 5v smpsu boards) by having to open up the box (not to say the possible need to cannibalise yet another cheap wallwart if you've managed to misplace your spare which had likewise been cannibalised from an identical wallwart in a genius stroke of forward planning)?  If you must pull the GPSDO box apart to repair it, please let it not be for the sake of a PSU board that could have been swapped out in less than a minute as an external unit plugged straight into a wall outlet.

JBG
 

Offline texaspyro

  • Super Contributor
  • ***
  • Posts: 1267
It's generally assumed that OCXOs have a EFC input (unless otherwise specified) and they are usually referred to as just OCXOs and not VCOCXOs.

Also, retrace describes how an oscillator freqency changes after a disturbance like a power failure.  This change is usually not due to just temperature changes.
« Last Edit: March 29, 2019, 10:27:47 pm by texaspyro »
 

Offline texaspyro

  • Super Contributor
  • ***
  • Posts: 1267
Oscillators like clean power supplies, otherwise power noise can get into the output.   Cheap 5V supplies and switchers are not known for being quiet.
 

Offline texaspyro

  • Super Contributor
  • ***
  • Posts: 1267
Also cheap OCXOs are not known for having low phase noise, etc.   You get what you pay for... or not.  New, quiet, stable, high performance  OCXOs can easily set you back over $1000... some are over $30,000.  Price goes up exponentially with performance.,  a very few extra dB can easily double the price.  I'd say start with a cheap used one and then go from there.

You are not going to get quiet, low phase noise out of a cheap OCXO or wall wart.
« Last Edit: March 29, 2019, 10:37:42 pm by texaspyro »
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf