NixieChron's Haicom HI-204III GPS Receiver Pucks - Programming them to work with NixieChron

[Sonar Tech]

Hi, folks -

I wanted to drop a quick note that might help someone in the future regarding the Haicom HI-204III GPS pucks that Jeff Thomas used with the NixieChron clocks.  I have 3 of these clocks.  As some of you may know, Jeff programmed the GPS receiver pucks to only output the NMEA $GPRMC string, so it wouldn't blow the buffer on the PIC with a bunch of messages he didn't care about.  Doing this also allowed for more reliable precision enabling him to trigger 1PPS off of the leading edge of the $GPRMC string.

Anyway, the pucks have a small 3V Li-ION battery inside that stores configuration data.  If you unplug the puck for several weeks, the battery will die and the programming and configuration will be lost, returning the puck to factory defaults.  I moved offices and boxed-up my clock for a couple of months.  When I tried to set it back up, all it would do is start counting up, from 00:00.  Additionally, the red LED light in GPS puck itself would never flash, even in the presence of known good GPS signals.  If you have a HI-204III that isn't flashing anymore and won't work with your Nixichron, your puck may have fallen back to a factory default configuration which would prevent it from working with the NixieChron.

To fix the problem, you need to re-program the puck.  Unfortunately, this is a little tedious for people without the right tools, as it requires an RS232 serial dongle with a 5V power source to make the puck work while you re-program it.  You can't just use a FTDI UART cable; you need a RS232 transceiver cable because you need the level shifter for RS232.

Lucky for me, I had a nice dialog with Jeff about this process when one of my pucks reset itself back in 2012, and Jeff was kind enough to supply me with enough details to figure out how to re-program the pucks myself if they ever misbehaved again.

Here's the thread I had with Jeff about this subject, provided here so that the Interwebs will never forget them:

From: Jeff Thomas [mailto:jth...@amug.org]

Sent: Saturday, February 18, 2012 5:24 AM

To: Sonartech

Subject: Re: Bad Nixichron MCU?

  

Hi Sonartech, SiRF Corporation produced a utility program for the peecee that functions as a test and programming environment. It's what I use.

The receivers can be configured with any NMEA setup, or the native SiRF protocol.

To effect a change to NMEA, the receiver must first be placed into SiRF protocol mode (via pulldown menu), then NMEA is selected again in the pulldown menu.

When NMEA is selected, a submenu appears, and you can edit the various NMEA strings being broadcast.

All strings are turned off, except for $GPRMC at once per second. 

 ...I hard wired a dongle from the mini DIN to a DB9, with a small 5v PS "tee'd" into the dongle wire for power.
Your NixiChron schematic shows the connections to the mini DIN. You'll have to wire the DB9 as needed.

At the bottom of the page is a communication log, and it will show the serial data. Only the NMEA $GPRMC will show as being broadcast.

  

Here's a link to a site who has the SiRF demo program:

http://www.falcom.de/support/software-tools/sirf/

  

Regards, Jeff

 

Finally, here's a quick summary of what I did, based on Jeff's instructions above.

First, I purchased an FTDI RS232-to-USB cable from Digi-key, part number 768-1074-ND.  This will provide a USB-based COM port, the level shifter and the required 5V supply all in 1 package.  Next, you need a 6-pin female mini-DIN connector.  I used Digi-Key P/N CP-2160-ND, but CP-2660-ND and CP-2860-ND are also good options.

I used Red, Black, Orange and Yellow 28 AWG stranded wire to make the connections.  You need +5VDC, GND, RX and TX.  I cut each color to a single 4 inch piece and stripped & tinned each end.

Next, I dismantled the RS232-to-USB cable.  Using some Style 2A flat tweezers, I wedged-open the halves of the connector, top and bottom, revealing the circuit board.  There are no screws, only a tight press-fit between the two white plastic halves.  Try not to jam the tweezers into your hand - it kind of hurts and bleeds for a while.

Once inside, carefully remove the DB9 jack soldered to the PCB.  Take care not to lift the pads or overheat; it may make more sense for you to just cut the leads behind the jack body, then use a soldering iron to clean up the pads - less risk doing it this way.

Once the pads are clean, solder the red wire to the MD-60J Mini-DIN socket, pin #2 - +5VCC.  Next, solder the black wire to MD-60J Mini-DIN socket, pin #1 - GND.  Then solder the orange wire to Mini-DIN socket pin #4 - Rx.  Finally, solder the yellow wire to Mini-DIN socket pin #5 - Tx.

Now you're ready to attach the Mini-DIN to the RS232 PCB.  First, route the 4 wires through the jack body and the blue rectangular collar that should have popped-off when you dismantled the plastic case around the cable.  Solder the red wire to the same red wire near the rear of the PCB, where the USB red wire is connected.  This will supply +5V to the puck.  Take care not to allow the original red wire to pop free; you need both wires attached to this same pad.

Then solder the black wire to the same black wire near the rear of the PCB, where the USB black wire is connected.  This will supply GND to the puck.  Take care not to allow the original black wire to pop free; you need both wires attached to this same pad.

Next, solder the orange wire to pad #2 (Rx) on the RS232 transceiver board.  For reference: If you're holding the PCB with the cleaned-up solder pads to your right and you can see 5 pads (top side), the bottom-right pad is pad #1.

Finally, solder the yellow wire to pad #3 (Tx) on the RS232 transceiver board.

Before you proceed to clean-up the plastic and reassemble things, TEST YOUR CABLE with the GPS puck.  Plug the puck into the cable, then plug the cable into the PC.  You should immediately see the red LED glow in the GPS puck base.  If you don't, immediately unplug the cable and check your connections, because something's wrong!  If you have a red solid light, and you have a decent GPS signal, you [may or may not] see one of the blue LED's on the RS232 puck light-up and flash every second as it receives NMEA data from the puck.  If you don't see anything, make sure you have a GPS signal and wait at least 5 minutes to acquire a signal.  If after 5 minutes you still don't see any flashing lights, try swapping the Rx and Tx lines; they may be reversed.  You can safely T/S this once you get the program running below.

Next install the SiRFDemo tool Jeff mentioned.  As of 8/28/2017, it was still available HERE.

Once installed, connect the puck, verify the red LED in the GPS puck is working, then start the SiRFDemo program.  It will immediately ask you for Data Source Setup; go to device manager and figure out what the COM port number is for your USB-to-Serial cable.  Look under Ports (COM & LPT) for something called a "USB Serial Port (COMx)", where "x" is your COM port number.  Select the correct COM port in the Data Source Setup dialog box, and select 4800 for your Baud Rate, then click OK.

If everything's working OK up to this point, you should see valid NMEA strings being shown in the Debug View window, and the Receiver Output View window will show "255 (0xFF) NMEA Data".  If everything's working, and the blue light on the serial cable is flashing once per second or so, it's time to re-program the puck:

1. Pull down the "Action" menu and select "Switch to SiRF Protocol"
2. Wait a few seconds for everything to update; you should see data in all the other windows "come alive" with information.
3. Go back to the "Action" menu and select "Switch to NMEA Protocol..."
4. A new menu should now appear, confirming which NMEA messages you want to see.  Change all of the menu settings as follows:
1. GGA: 0
2. GLL: 0
3. GSA: 0
4. GSV: 0
5. RMC: 1
6. VTG: 0
7. MSS: 0
8. User 8: 0
9. User 9: 0
10. User 10: 0
11. Use Checksums: CHECKED
12. Baud Rate: 4800
5. At this point, the GPS puck red LED should now be flashing once-per-second.  GPS reprogramming is now complete and the GPS puck can be used once more with your NixieChron clock.

That's it!  I tried to be comprehensive enough to get anyone going with this, but if you still need help getting this problem resolved, please contact me and I'll see how I can help.

Cheers,

SonarTech

 

[Terry S]

Very comprehensive -- thank you. Any info on replacing the internal battery?

[Sonar Tech]

Unsure about that.  The HI-204III puck it allegedly waterproof, so I imagine they glued it shut all the way around.  Also, there's no FCC ID on my puck, so I can't look up the internal photos on the FCC's web site.  Anyone know the FCC ID for the HI-204III puck?  Also, my battery isn't failing, per se, it's just not holding a charge for months.  I don't know what it's rated at performance-wise, but I know my puck still holds a charge (and thus retains configuration data) for at least a few weeks.  Now that I've made the programming cable and have figured out how Jeff programmed to puck, I'm way less panicky about this part failing.  If the clock internals die, however, I'll have quite the emotional episode!

[Terry S]

I believe the label is still on my puck, I will look this evening.

Lithium batteries are finicky little devices. They like to be stored depleted and charged just before use. Staying charged all the time is probably the worst case scenario for these little cells. I would not be surprised to see them eventually fail in droves. 

Maybe the configuration is maintained in a capacitor? Do you know for a fact there is a lithium cell?

At one time PC's had "PS2" ports and serial ports, both of which could be directly adapted to talk to the Haicomm units. I distinctly recall tinkering with one at one point, perhaps during some similar discussions with Jeff. I may still have the SiRF configuration software installed on an old PC.

Terry

[Sonar Tech]

The only definitive data I have on the internals is the owner's manual, that on Page 10 states that Backup power is provided by a "3V Lithium-Ion rechargeable battery".  Additionally, on Page 26, it says "Commands can be issued to the HI-204III to change the settings of the receiver. The new settings will remain effective on next power-on as long as the on-board rechargeable backup battery is not discharged.  After the backup battery is discharged, factory preset default settings will be used."

Factory defaults are as follows:

Datum: 000 (WGS-84)

NMEA Commands: GGA ON (1 sec), GLL OFF, GSA ON (5 sec), GSV ON (5 sec), RMC ON (1 sec), VTG ON (1 sec) and Checksum ON

Baud Rate: 4800

Elevation Mask: 5°

DOP Mask: DOP Select: Auto, GDOP: 10, PDOP: 10, HDOP: 10.

Receiver Operating Mode: Normal Mode (without 1PPS)

[dixter]

Thanks Sonar Tech...   I originally bought two ( 1 black and 1 silver ) but do not run the two at the same time...  I'm sure the other puck has died by now   good to have these instructions...  now I need to dig into the closet to find the other puck and test it out...    I guess with your USB cable solution I could just keep the extra puck charged up using a USB wart and the custom cable...

later

Dick Bell

[Mitch]

Just letting everyone know that I'll have another ten NixiChron AVR upgrades available shortly. I have around fifteen boards left and five are spoken for. After this, I doubt that I'll make more. Around 33 have been sold so far. They should be ready sometime later this month or early next month.

[MichaelB]

Great information and nice write up. As many of you know I have several of Jeff's clocks in a couple flavors and many of them are bring driven by Haicoms and several BR355's. A couple of months ago I went through all of the GPS's and turned off all the other 'competing' strings. Some of the Haicoms were originally done by Jeff, but most of the others and all the GlobalSats had never been done. No dead batts though. This resulted in a noticeable difference as far the clocks being much more in sync with one another and other clocks. Worth the effort.

[Terry S]

Well it turns out my unit is a Haicomm HI-204S, not 204III.... not sure if that is a significant difference. I have had problems over the years with my Nixichron keeping time poorly even when the Haicomm appears "happy" according to the flashing LED. On the other hand, my Jeff Thomas WWVB clock almost never misses a step.

Terry

[Nicholas Stock]

Terry, may I ask how long you've been running your WWVB? I was always a little concerned with the lifespan of the Z568's in that clock and have refrained from keeping mine 'active', but I may be erring on the overcautious side of things in that regard...

Cheers,

Nick

--
You received this message because you are subscribed to the Google Groups "neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email to neonixie-l+unsubscribe@googlegroups.com.
To post to this group, send email to neoni...@googlegroups.com.
To view this discussion on the web, visit https://groups.google.com/d/msgid/neonixie-l/eee0c42e-4e38-4d4d-b7be-bffd2d21a9cc%40googlegroups.com.

For more options, visit https://groups.google.com/d/optout.

[Terry S]

Nick,

I'm not sure - I think it was around 2003 (PCB is dated 2002). The WWVB clock was newly released by Jeff. It's been running continuously ever since, so probably 14 years. The only problem I've had was some cathode poisoning that Jeff helped me cure.

My advice -- use the clock and enjoy it. The tubes are robust if you keep your hands off them. And they look spectacular.

Terry

On Tuesday, September 5, 2017 at 6:22:13 PM UTC-5, Pramanicin wrote:

Terry, may I ask how long you've been running your WWVB? I was always a little concerned with the lifespan of the Z568's in that clock and have refrained from keeping mine 'active', but I may be erring on the overcautious side of things in that regard...

Cheers,

Nick

On Tue, Sep 5, 2017 at 11:33 AM, Terry S <tschw...@gmail.com> wrote:

Well it turns out my unit is a Haicomm HI-204S, not 204III.... not sure if that is a significant difference. I have had problems over the years with my Nixichron keeping time poorly even when the Haicomm appears "happy" according to the flashing LED. On the other hand, my Jeff Thomas WWVB clock almost never misses a step.

Terry

On Tuesday, August 29, 2017 at 3:41:51 PM UTC-5, Terry S wrote:

I believe the label is still on my puck, I will look this evening.

Lithium batteries are finicky little devices. They like to be stored depleted and charged just before use. Staying charged all the time is probably the worst case scenario for these little cells. I would not be surprised to see them eventually fail in droves. 

Maybe the configuration is maintained in a capacitor? Do you know for a fact there is a lithium cell?

At one time PC's had "PS2" ports and serial ports, both of which could be directly adapted to talk to the Haicomm units. I distinctly recall tinkering with one at one point, perhaps during some similar discussions with Jeff. I may still have the SiRF configuration software installed on an old PC.

Terry

On Tuesday, August 29, 2017 at 1:00:08 PM UTC-5, Sonar Tech wrote:

Unsure about that.  The HI-204III puck it allegedly waterproof, so I imagine they glued it shut all the way around.  Also, there's no FCC ID on my puck, so I can't look up the internal photos on the FCC's web site.  Anyone know the FCC ID for the HI-204III puck?  Also, my battery isn't failing, per se, it's just not holding a charge for months.  I don't know what it's rated at performance-wise, but I know my puck still holds a charge (and thus retains configuration data) for at least a few weeks.  Now that I've made the programming cable and have figured out how Jeff programmed to puck, I'm way less panicky about this part failing.  If the clock internals die, however, I'll have quite the emotional episode!

On Tuesday, August 29, 2017 at 9:27:39 AM UTC-7, Terry S wrote:

Very comprehensive -- thank you. Any info on replacing the internal battery?

--
You received this message because you are subscribed to the Google Groups "neonixie-l" group.

To unsubscribe from this group and stop receiving emails from it, send an email to neonixie-l+...@googlegroups.com.

To post to this group, send email to neoni...@googlegroups.com.

[Nicholas Stock]

Thanks Terry. Sounds like you got yours at the same time as me. I will fire it up again.

Best,

Nick

Sent from my iPhone

To view this discussion on the web, visit https://groups.google.com/d/msgid/neonixie-l/2fa94952-7e01-4e65-8794-4db536c8820c%40googlegroups.com.

[Terry S]

I just received a Globalsat BR-355S4 GPS receiver - $26 shipped on ebay. This unit uses the SiRF Star IV chip set. It's about 1/2 the size of my Haicom unit, maybe 1/4 the volume. It includes a suction cup which adheres it nicely to the window glass. Only apparent down side is that the cable is much shorter at 1.5 meters. My plan was to re-program it per Sonar Tech's write-up, but I went ahead and plugged it right into my Nixichron.

It got a GPS lock within seconds, wayyyyy faster than the Haicom unit ever did. In fact the time was accurate before the receiver even started flashing its' LED. GPS location is rock solid. The Nixichron is very happy. Me too.

Terry