I've figured out what I was doing wrong. I expected all my changes would happen with one adjustment, instead each has to be done separately, It's now in the laptop on the balcony. I'll check it in a wee while.

yep, looks good, just got NAV.PVT All I need now is to get openlogger working.

Select to expand quote

decrepit said..
Peter, my understanding of the openlog, is that with only a virgin, fat 32 card in and nothing connected. It will write a config.txt file to the card. Mine is not doing that, that's why I suspect a faulty openlogger.

Sorry, I did not read your previous post carefully enough. Your understanding is correct.

I just tested the three Openlog units I got from China, and none of them works as expected. The LEDs don't light up as when connected to power, no config files are created, and no data are logged. I found one post on a German forum that has similar issues, which were caused by a firmware issue (no or incorrect logging software installed). However, fixing that seemed like a major effort even for someone who has experience programming Arduinos. So for all practical purposes, the the units are useless. Unfortunately, shipping them back to China for a refund will cost more than I paid for them.

I went back to the seller's site and read the item descriptions carefully. The Openlog units that don't work are all black, and have just one row of connections. The other units I have (from Amazon) which work are purple, and have two rows of connections. The different colors also correspond to different software versions. The purple units have the default Openlog software (with full config files), the black units should have "openlog_blackbox_v2.0 firmware". That should not be a problem, but it seems something went wrong somewhere. If you really want to try to get the unit to work, you could try to follow the hints at www.mikrocontroller.net/topic/450438 .. but I certainly won't.

Thanks again Peter for doing that research.
Yes I have the black one with one row of connections. And "black box" was in the item description.
I won't bin it just yet, but I suspect it will just sit in the bottom of my junk box.
The one on order is from "sparkfun" and it's a red or orange with two rows of contacts.

At least we are finding all the traps, and paths not to go down!

Select to expand quote

decrepit said..
Yes I have the black one with one row of connections. And "black box" was in the item description.
I won't bin it just yet, but I suspect it will just sit in the bottom of my junk box.

Mike as mentioned before. Getting a refund from Banggood was fairly painless when I got a credit card phone for about $18 that didn't work. So give them a try. Took just a couple of emails and 2 weeks for the money to be back in my M/C. I did not have to return the phone. The support person was very pleasant.
Gearbest is another matter. They don't even answer emails and support is non-existent.

Thanks David, email sent, we'll see what happens.

Select to expand quote

decrepit said..
At least we are finding all the traps, and paths not to go down!

"We are finding all the traps" - that sounds like famous last words of the archeologist

Select to expand quote

So, here's the reply from Bangood.

cservice@banggood.comTue, Jul 10, 2018 at 9:24 AM
To: mikelayzell
Dear Michael Layzell,
Thank you very much for your order 50460147.
Please upload a video to YouTube and take photos recording to illustrate this issue and send them directly to us, which will allow us to verify the problem and help to resolve it.
We apologize for the inconvenience and look forward to hearing from you.
Best regards, Howard

In my initial mail I had included 3 pics as requested, but how do you show that the logger doesn't record on a video?

I've replied to that effect, so we'll see what they say next.

When you connect power, the LEDs on the logger should light up. The one on mine do not, I bet it's the same for yours. You could send them a video showing that.

OK that will be easier. I'll wait for their next reply, first though.

I cut the broken section off the battery lead today, and hard wired it to a switch, If I get time in the next few days, I'll have another play with the Pi

Today's message from Bangood.

Dear Michael Layzell,

Thank you for your email.

We deeply apologize for what happened with your order 50460147. Would you accept to receive 850 points as a token of our apologies?

Generally speaking, 100 Banggood points are equal to US$1. Points can be exchanged to give up to 15%-20% off your order and can not be used in the following scenarios:- Flash deals products;- Products at unbeatable prices;- Coupons;- Drop-shipping orders;- Pre-order products;- Products with a promo tag;


Sounds reasonable to me, much better than I expected. So I've accepted, think I'll order a battery charger module from them to use it up.

I got a full refund for my 3 units, after a couple of emails from the seller that looked like delay tactics and then asking ebay to step in. So far so good. Only bummer is that the seller also had the purple Openlog units listed for a few cents more, which probably would have worked.

I've just discovered my deal isn't as good as I thought. I can only use a small number of points for each purchase. Means I have to spend a lot more with them to get my money back.

And of course I've had yet another set back with the Pi. The modified battery with switch works, but the Pi records 202bytes of rubbish then stops recording. I suspect my Ucentre settings have something to do with that. So as a test I'll reconnect the cheap Chinese unit and see what that does.

WOOOOOOOOOOOO HOOOOOOOOOOOOOOOOOO,
New openlogger arrived, and I'm getting NAV PVT sentences out of it.
Took a while though, same old problem with broken vcc track where I'd removed the old one, and I didn't notice. The multi-meter spotted the problem, and I quickly fixed that with a short piece of wire. That got the config.txt file on the SD card, but I couldn't get any data.
I had the usb-serial converter in parallel so plugged into the laptop, and nothing coming into U-center, green light flashing away on the M8n. Wondered if having both in parallel was the problem so cut RX and Tx to the openlogger. That made no difference. Was about to give up, but thought I'd better try the last combo, and reconnected the logger without the converter. Yay blue light on the logger starts flashing.

Current drain 60ma, but that's going through the converter, may be less with a direct connection, I'll use the trusty multi-meter tomorrow and find out.
I'd like enough battery to do a 24hr distance, (for people like Kato), but all I could conceivably need is 14hrs.

Tomorrow, I'll see if I can get it all in a small box and take it for a ride.

So looks like the USB to serial converter is finally U/S.

Well done Mike, your persistence deserves a result. Most would have given up by now.

Select to expand quote

Swindy said..
Well done Mike, your persistence deserves a result. Most would have given up by now.

Well I gave up twice today, for an hour or so.

Almost gave up again today!
After fixing everything on to one board, no green light flashing, file saved to card had no GPS data.
Not happy Jan!
Suspected in my fiddling I'd done something nasty to the antenna, so swapped for the cheap Chinese one. When I removed the suspect antenna from the good M8N the plug separated from the coax, not sure how easy it will be to reconnect, so connected the cheap Chinese one, and that works.
Peter, a loose plug on the cable of your dodgy antenna may be the cause of the problem. It only looks like a crimp fit, and I don't see how the inner connector gets crimped.

Here's the results of a stationary test in the back yard. I've removed the start and finish noisy stuff. Grid is 1m



Here's what it looks like at the moment. everything hardwired except the antenna cable, a little slide switch connects battery +ve



Still too big to fit in the little box I bought for it, because I wanted to try a big ground plane.
It does fit inside a mini paqua though. So after lunch I'll strap it and the Gw52 to the masts and take them for a car ride.
Back later.

Good results, the logger matches the GW52 very closely. I don't think two GW52s would be any closer.
So here's the set up.



GW52 in an H20tek pouch and the looger in a midi paqua.



And here's the display in GPSResults, from the resultant files.


The purple/pink lines are the accuracy data, I've never used this before, but the ubx file doesn't display SDoP.
Top is logger bottom GW52.

And here's the text read out for both 2sec results.
GW52
2s:
time,,,,,,,,,,, m ,,,,,[knots] accuracy
13:43:04, 54.0, 52.499, +/-0.139
13:48:00, 48.9, 47.539, +/-0.124
13:45:50, 40.3, 39.132, +/-0.215
13:38:24, 39.0, 37.91,1 +/-0.129
13:40:46, 31.9, 31.012, +/-0.172
13:39:09, 31.2, 30.320, +/-0.154
13:49:43, 30.7, 29.875, +/-0.042
13:36:20, 30.6, 29.758, +/-0.198
13:34:17, 29.7, 28.821, +/-0.086
13:37:06, 29.6, 28.753, +/-0.219
13:35:39, 27.6, 26.778, +/-0.122
13:44:20, 26.8, 26.022, +/-0.153
13:43:54, 22.1, 21.472, +/-0.090
13:33:06, 11.2, 10.887, +/-0.081
13:33:40, 10.7, 10.421, +/-0.079
13:39:58, 7.3, 7.072, +/-0.186
13:40:00, 7.2, 7.023, +/-0.175
13:46:16, 6.8, 6.601, +/-0.109
13:33:12, 6.3, 6.146, +/-0.075
13:51:30, 4.4, 4.236, +/-0.067
13:33:44, 3.7 ,3.583, +/-0.098
13:45:13, 3.2, 3.089, +/-0.275
13:40:03, 2.3, 2.263, +/-0.127
13:41:20, 1.2, 1.176, +/-0.084
13:46:09, 1.1, 1.108, +/-0.062
average[2]: 92.635km/h 50.019knots
average[3]: 85.914km/h 46.390knots
average[5]: 77.077km/h 41.618knots

Logger/M8N 2s
13:43:22, 54.0, 52.510,+/-0.133
13:48:18, 48.9, 47.533,+/-0.151
13:46:07, 40.2, 39.078,+/-0.179
13:38:43, 39.0, 37.941,+/-0.146
13:41:04, 31.8, 30.955,+/-0.153
13:39:27, 31.2, 30.322,+/-0.157
13:50:01, 30.7, 29.876,+/-0.169
13:36:38, 30.6, 29.727,+/-0.151
13:34:35, 29.7, 28.827,+/-0.138
13:37:24, 29.7, 28.824,+/-0.145
13:44:38, 26.8, 26.078,+/-0.149
13:44:12, 22.0, 21.422,+/-0.164
13:33:24, 11.2, 10.866,+/-0.171
13:33:58, 10.7, 10.425,+/-0.152
13:44:23, 7.3, 7.110,+/-0.149
13:40:16, 7.3, 7.059,+/-0.171
13:40:18, 7.2, 6.959,+/-0.163
13:46:34, 6.8, 6.590,+/-0.160
13:33:17, 6.6, 6.384,+/-0.165
13:33:30, 6.3, 6.15,2+/-0.160
13:33:13, 5.7, 5.563,+/-0.162
13:51:48, 4.4, 4.262,+/-0.156
13:34:02, 3.7, 3.579,+/-0.157
13:45:31, 2.5, 2.454,+/-0.186
13:40:21, 2.3, 2.276,+/-0.144
average[2]: 92.640km/h 50.022knots
average[3]: 85.884km/h 46.374knots
average[5]: 77.049km/h 41.603knots

So the GW52's worst accuracy is +/- 0.215 and the logger is +/- 0.186. I'm not sure what that means though as the calculations are different. It's interesting just how close they are, the discrepancy is at low speed where the GW52 has it's best results, contrary to what locosys says. And the open-logger is much more consistent varying from 0.133 to 0.186, whereas the GW52 goes as low as 0.042 and up to 0.215.

Tried to upload the files to KA72, so anybody can see them, but looks like Dylan hasn't enabled .ubx files yet.
GPS Visualizer is processing at the moment??? Well that timed out, maybe GPSVisualizer doesn't do .ubx either?

Think so it just did the .sbp file in a few seconds, but it just sits there with the .ubx

The max current is 50ma, so 24hrs needs 1200mah, my 2,200 is a bit of over kill, think I'll look for a smaller battery, that will make getting it in my little box easier.

Just noticed a huge difference in file size, 223Kb for the GW52 and 6.2mb for the logger.
Checked the file in Ucenter and I've now got a pile of NMEA data as well as the NAV PVT. Not sure how that happened, I'm surprised the logger handled it.
Maybe when the converter was packing up??

Now I'll have to see if I can get the converter working, or wait until a new arrives.

Looks good, Mike! Locosys apparently reduces the SDoP numbers at low speeds, ublox does not. You can sometimes see the Locosys units staying at 0 speed for longer when slowly starting to go faster. Not a problem for speedsurfing, though.

The ubx files should be about 3x larger than the .spb files when recording at the same rate. Your files are 30-fold larger, which is probably from tons of NMEA data, which need more space than the binary data. I have seen one of my ublox chips (the Drotek NEO-M8) adding NMEA data, regardless of what the settings are. None of the other chips does that. With a Pi or bluetooth setup, that's no issue since the GPS gets configured at the start. With the Openlog, it will create large files, like you observed. Theoretically, this could be fixed with custom Openlog software, but that's a tad tricky.

There might be other solutions; it may be worth asking on the u-blox forum. Maybe I'll get around to checking into this in the next few days. The last few days have been too windy .

Select to expand quote

decrepit said..
Tried to upload the files to KA72, so anybody can see them, but looks like Dylan hasn't enabled .ubx files yet.
GPS Visualizer is processing at the moment??? Well that timed out, maybe GPSVisualizer doesn't do .ubx either?

Think so it just did the .sbp file in a few seconds, but it just sits there with the .ubx

I think the Windows version of GPSResults lets you export the files in .sbp format. That may be easiest if it works. I have not tried that recently, though, and there could perhaps be some issues with SDoP / sAcc numbers. These are one byte numbers in Locosys files, but 4 bytes in .ubx, so range checking/limiting would be needed. GPSAction Replay has a problem with that.

Select to expand quote

boardsurfr said..
>>>> I have seen one of my ublox chips (the Drotek NEO-M8) adding NMEA data, regardless of what the settings are. None of the other chips does that. With a Pi or bluetooth setup, that's no issue since the GPS gets configured at the start. With the Openlog, it will create large files, like you observed. Theoretically, this could be fixed with custom Openlog software, but that's a tad tricky.

There might be other solutions; it may be worth asking on the u-blox forum. Maybe I'll get around to checking into this in the next few days. The last few days have been too windy .

Bummer!!!
So with a 24hr attempt, it will be a very large file!
I wonder if that will be a problem.

But I'll check my settings again in case something I've done shifted them. I'm sure at one stage I only saw NAV PVT sentences coming in, but maybe that was just after I set them. I'll check my old files and see.

Select to expand quote

boardsurfr said..
I think the Windows version of GPSResults lets you export the files in .sbp format. That may be easiest if it works. I have not tried that recently, though, and there could perhaps be some issues with SDoP / sAcc numbers. These are one byte numbers in Locosys files, but 4 bytes in .ubx, so range checking/limiting would be needed. GPSAction Replay has a problem with that.

Daffy says he can get the .ubx files to load in GPSVisualizer, but I think he must be more patient than me. changing to .sbp speeds it up no end.
Here's part of the result. Red id the M8N/logger, blue the GW52. The M8N certainly has much better positional accuracy around bends. I wonder if this will affect alpha results.



So now I'll upload the GW52 and logger converted to sbp files to KA72 and see what happens.
well they went in OK, if you want to download, do a "decrepit" search

Select to expand quote

decrepit said.


Daffy says he can get the .ubx files to load in GPSVisualizer, but I think he must be more patient than me. changing to .sbp speeds it up no end.
Here's part of the result. Red id the M8N/logger, blue the GW52. The M8N certainly has much better positional accuracy around bends. I wonder if this will affect alpha results.



So now I'll upload the GW52 and logger converted to sbp files to KA72 and see what happens.
well they went in OK, if you want to download, do a "decrepit" search

Here is the file in GPS-Visulaiser (There is a bug in the speed decimal placing though ):



Here is a screen shot from the Mac version of GPS-Results:

Good result today, but I'm totally mystified.

Bought a new usb-serial converter, and plugged into U-center. Straight away all I had was NAV. PVT, no sign of any NMEA, without touching the configuration.

So reconnected to the open logger, and a few hours later did another balcony test.
Again only NAV. PVT in the resultant file.

So what happened previously when I had reams of NMEA stuff as well as the NAV PVT????? Could it have happened when I had the old serial-usb converter in parallel and it packed up? Doesn't seem likely.

Maybe the message settings doesn't get into flash memory, only BBR? I'll leave it off for a few days and see what happens

Select to expand quote

decrepit said..
Good result today, but I'm totally mystified.

Bought a new usb-serial converter, and plugged into U-center. Straight away all I had was NAV. PVT, no sign of any NMEA, without touching the configuration.

So reconnected to the open logger, and a few hours later did another balcony test.
Again only NAV. PVT in the resultant file.

So what happened previously when I had reams of NMEA stuff as well as the NAV PVT????? Could it have happened when I had the old serial-usb converter in parallel and it packed up? Doesn't seem likely.

Maybe the message settings doesn't get into flash memory, only BBR? I'll leave it off for a few days and see what happens

I would have to check, but I'm too lazy at the moment, but I don't know if the M8N comes with flash. I think it just has the BBR. From what we have seen, where they use a capacitor instead of a battery, the config data will be lost after a short time. I think it takes a day for the config in my module to be lost.

Does that match up with the problem you have seen? I thought that the default out of the box config for the GPS module is that it sends out NMEA only.

Select to expand quote

FormulaNova said..
I would have to check, but I'm too lazy at the moment, but I don't know if the M8N comes with flash. I think it just has the BBR. From what we have seen, where they use a capacitor instead of a battery, the config data will be lost after a short time. I think it takes a day for the config in my module to be lost.

Does that match up with the problem you have seen? I thought that the default out of the box config for the GPS module is that it sends out NMEA only.

Well it kept the 115200 baud rate, and was outputting the NAV PVT sentence as well as the NMEA.

I'll try it in a couple of days and see.

It would be a bugger if it has to be reprogrammed every sail.

And I suspect the converter failed because I had it in parallel with the logger. That would mean a switch to swap from one to the other. Getting a bit beyond a simple, user device.

You could be right formula.

Configuration settings can be modified with UBX configuration messages. The modified settings remain effective
until power-down or reset. If these settings have been stored in battery-backup RAM, then the modified
configuration will be retained, as long as the backup battery supply is not interrupted.

So I hope the one I have is actually a battery not a cap

Seems it depends which chip you have, the above is the MAX -M8

This what it says for the NEO -M8

Configuration settings can be modified with UBX configuration messages. The modified settings remain effective
until power-down or reset. If these settings have been stored in battery-backup RAM, then the modified
configuration will be retained, as long as the backup battery supply is not interrupted.
With the NEO-M8N, configuration settings modified with UBX configuration messages can be saved
permanently. In this case, the modified settings remain effective even after power-down and do not require
backup battery supply.

I think the M8N is a NEO, fingers crossed. This maybe the difference Peter has with the French device?

It's pretty easy to check if the configuration gets saved to BBR and/or Flash. I'd simply change the rate (e.g. to 1 Hz), then save the settings to BBR only. Then change the rate to 5 Hz, save the settings to Flash only.

Next, load the settings from BBR. The rate should revert back to 1 Hz. I usually check how many messages I get in the "Packet" window, that's easier and less error-prone than using the configuration view.

Then, load the settings from Flash. The rate should be 5 Hz again. That should be obvious right away in the "Packet" window, with many more messages going through.

Next step is to unplug and re-plug the GPS to see what rate you get. It's probably the 1 Hz from the BBR, if there is some backup (battery or capacitor). I have not found anything in the ublox docs about priority between BBR and Flash if both are available, though.

Good thinking Peter except it's no longer connected to a usb converter, I'll wait a few days and see what it does.