GPSDO or not?

149 views
Skip to first unread message

Mark Phillips

unread,
Apr 4, 2021, 6:35:43 PM4/4/21
to hamsci-grape
I have read with interest one does not necessarily _need_ a GPSDO to operate the GRAPE receiver. This leads me to believe that the quality of data from a non locked RX is adequate for the needs of this project? That said, a GPSDO would no doubt make significant improvements in the quality of the collected data. 

At what point does the reliability (or error/jitter/whatever) of the GPSDO start to make things better and why would the Bodnar device be any better than say the one published in QEX a few months ago?  

I'm coming at this from a cost reduction perspective. I think that for an amateur project this PSWS is still far too expensive. In addition to the RX (about $30) there's a $150 GPSDO, a $70 RPi4, about $25 for a 10MHz dipole antenna and other parts such as coax and reliable power supplies.

With the Bodnar GPSDO coming in at $150 but yet the many, many ones based on a Ublox NEO-x and a Si5351 (with Arduino to control things)  coming in at around $30, a saving of over $100 can be realised immediately.  Indeed, should the Si5351 GPSDO be "good enough for Government work" then it could also be built onto the GRAPE PCB for an all inclusive device. Possibly even with added I2S sound chip ultimately making a RPi HAT?


Mark
G7LTT/NI2O

Dr. Nathaniel A. Frissell Ph.D.

unread,
Apr 5, 2021, 3:31:08 AM4/5/21
to Mark Phillips, hamsci-grape, Steve Cerwin, John Ackermann. N8UR

Hi Mark and the group,

 

The GPSDO is an absolute requirement of the Grape. There was perhaps some uncertainty about this earlier, but recent work by Steve Cerwin WA5FRF has shown that any fixed frequency error translates into a velocity offset that when integrated over time produces a large error in ionospheric height measurements: https://www.youtube.com/watch?v=xfhAxuViTYQ. Similarly, see his excellent earlier paper from the 2020 TAPR DCC examining HF Doppler measurements: https://files.tapr.org/meetings/DCC_2020/2020DCC_WA5FRF.pdf. Kristina KD8OXT’s recent IEEE GRSL paper also shows the importance of the GPSDOs: https://doi.org/10.1109/LGRS.2021.3063361

 

You may have heard Kristina say that for people to participate in HF Doppler measurements that they do not absolutely have to have a GPSDO. This is being said only to encourage people who only have amateur gear to still be able to participate and learn, as well as to provide some roughly qualitative data. We really want to get people away from this, and anyone who will actually be building a Grape should consider a GPSDO a requirement.

 

With that said, other GPSDOs beside the Bodnar will likely perform adequately. The Bodnar was chosen for the Grape v1 because this was an easy, cost-effective COTS solution for the prototype. Both the Grape v2 and the TangerineSDR will be using a custom-built clock module/GPSDO that balances the cost/performance requirements. See John Ackermann N8UR’s recent talk reviewing the current designs: https://www.youtube.com/watch?v=xfhAxuViTYQ. Note that he has three different tiers for performance:

  • Low-Cost based on ublox NEO M9N single band GPS, no pps output
  • Mid-Grade based on ublox NEO M8T single band GPS
  • High-performance based on ublox ZED F9T dual band GPS

 

The Grape v1 does not take advantage of the PPS, so the cheapest option will likely do. This will not be true of the Grape v2. For the TangerineSDR, we need the dual-band GPS in order to make Total Electron Content (TEC) measurements. Note John also spends a lot of time thinking about hold-over requirements.

 

To summarize:

  1. You absolutely do need a GPSDO.
  2. A cheap alternative to the Bodnar GPSDO will probably work great for the Grape v1
  3. We will not be using the Bodnar in the Grape v2 or Tangerine, but instead have a custom design that balances science requirements and cost.

 

73 Nathaniel W2NAF

--
You received this message because you are subscribed to the Google Groups "hamsci-grape" group.
To unsubscribe from this group and stop receiving emails from it, send an email to hamsci-grape...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/hamsci-grape/CA%2Bue%3DO5Fod8nhCkG%2B8u6932z-4phGE%3DF1cdtP%2B6cncdbu4eFjQ%40mail.gmail.com.
For more options, visit https://groups.google.com/d/optout.

Jim Farmer

unread,
Apr 5, 2021, 10:02:17 AM4/5/21
to hamsci...@googlegroups.com

Do we also need to explicitly say that Grape1 expects the GPSDO to output on unusual frequencies, 2.499, 4.999, 9.999, 14.999 MHz?  I have one GPSDO and a second on order, which only put out 10 MHz and 1 pps, no other frequencies possible.

Thanks,

jim

Jim Farmer, K4BSE
Mobile 678-640-0860
jofa...@mindspring.com (spam blocked)
Useless, good-for-nothing retired engineer
Member, Fayette County Amateur Radio Club, http://www.kk4gq.org
Volunteer electrical engineer, Smoke Rise Baptist Church http://www.smokerisebaptist.org

 Love all, † Follow One

"A people that values its privileges above its principles soon loses both."
        Dwight D. Eisenhower, Inaugural Address, January 20, 1953

Dr. Nathaniel A. Frissell Ph.D.

unread,
Apr 5, 2021, 10:14:26 AM4/5/21
to Jim Farmer, hamsci...@googlegroups.com

Hi Jim,

 

Yes, that is important. The Grape1 does need to have a GPSDO input 1kHz below the observation frequency. So, a GPSDO that only puts out 10 MHz and 1 pps will not work. It’s also worth noting that John N8OBJ already has all of the scripts written to do the frequency changes for Leo Bodnar. If you use any other GPSDO, you will also have to take on more responsibility on the software side.

 

73 Nathaniel

Julius Madey

unread,
Apr 5, 2021, 10:16:07 AM4/5/21
to hamsci...@googlegroups.com
I believe that the firmware for the Pi expects a Leo Bodnar mini GPSDO which it programs for those specific frequencies. 
73,
Jules - K2KGJ

Dr. Nathaniel A. Frissell Ph.D.

unread,
Apr 5, 2021, 10:26:20 AM4/5/21
to Julius Madey, hamsci...@googlegroups.com

Ultimately, the Grape v1 is just a direct conversion receiver. So, whatever you put into the GPSDO input port gets mixed with whatever is on the HF antenna port. We select 1 kHz below the observation frequency to put the received carrier in the audio passband.

 

The software on the Raspberry Pi are just scripts that John N8OBJ wrote to handle programming the Leo Bodnar GPSDO frequency easily. As long as you have a mechanism for properly setting the output frequency of whatever GPSDO you are using, you don’t need to use the Bodnar or N8OBJ’s frequency-setting scripts.

 

CAVEAT – Wheneven you change frequencies, regardless of the method, make sure you have FLDIGI set to the correct frequency or the datafiles will be incorrect.

 

73 Nathaniel

 

From: hamsci...@googlegroups.com <hamsci...@googlegroups.com> On Behalf Of Julius Madey
Sent: Monday, April 5, 2021 10:16 AM
To: hamsci...@googlegroups.com
Subject: Re: GPSDO or not?

 

I believe that the firmware for the Pi expects a Leo Bodnar mini GPSDO which it programs for those specific frequencies. 
73,
Jules - K2KGJ

Dr. Nathaniel A. Frissell Ph.D.

unread,
Apr 5, 2021, 10:48:25 AM4/5/21
to hamsci...@googlegroups.com, David Kazdan

Thank you to David for the correction- the Grape is a “Low IF Receiver” and not a Direct Conversion.

 

73 Nathaniel

 

From: David Kazdan <dx...@cwru.edu>
Sent: Monday, April 5, 2021 10:39 AM
To: Dr. Nathaniel A. Frissell Ph.D. <nathaniel...@scranton.edu>
Subject: Re: GPSDO or not?

 

Better, "low IF receiver."  It wouldn't be useful for the Doppler measurement of the carrier if it were direct conversion.

 

          David

 

 

 

Reply all
Reply to author
Forward
0 new messages