Hayden! Thanks for the video. I was looking into the accuracy of my transverters and sdr radio. I'm trying to learn about this subject. Your video was the first to come up. I like your methodical and straight-forward explanation. See you next time you come over. 73 de Bill
Very timely! I just purchased a BG7TBL GPSDO to use with my IC-9700, and have been pondering what I would do if I ever manage to get my hands on an IC-7610. The distribution amplifier is just the ticket, obviously, so I'll pick up one of those, too. Thanks for another very useful video. Thanks, too, for the link to the attenuator. I've seen reports that even though the 9700 would be overdriven by the 10MHz output of the BG7TBL that it seems to work just fine without attenuation, but me being somewhat nervous about taking risks with such an expensive piece of gear, I was already looking for attenuators.
I bought one of these GPSDO a while ago and it's a great little unit. I have seen where the max power input is 12.9V and have heard that some of these units don't like being fed with 13.8V from a power supply or field day battery so installed a cheap buck booster inside mine so it holds the incoming power voltage at 12.5v regardless of what the voltage is being presented to it. I also fitted an Anderson power pole to the back of it for easier power connection out in the field and less chance of accidental reverse polarization. That also helped with the over crowding on the front panel as well.
A CHEAP solution is you can use a video distribution amplifier. Granted it's 75 ohm and the output level depends on quality of the amps used. I have cheap one and it worked out pretty good. tnx for the video! 73 N8AUM
Thanks for the great video. I checked all the comments but I can't find the solution. I have two questions: 1) can the 10 MHz GPSDO work just off the OCXO, without GPS? 2) I read in specs on eBay that GPSDO can achieve a frequency accuracy of 0.5 parts-per-billion. Is that for real or is it just marketing hype?
I can answer the first question - yes it can work without the GPS, obviously not as accurate but fairly stable. I can’t answer question 2 as I don’t have the test equipment to test
Thanks. But without test equipment, is 1 ppb fairly standard across the industry for these types of devices. I am asking because I am a total newbie and I never came across such a high accuracy. I am a bit surprised.
Hello from Belgium, what is the maximum length of the cable between the 10 MHz output of the BG7TBL and the input of the distribution amplifier? Yours sincerely
I’m. It sure about maximum length. I use a short length of coax. Maximum length would be dependent on loss - even RG-58 you can run a long a way at 10 MHZ with limited loss
I manage to put one of these and the distribution amplifier in the same case, but the 7805 regulator on the amplifier board get quite hot and make the box quite hot by hand… that could be armful for the ocxo? Maybe I must put them back separately
I noticed another unit I have has slightly more output. Not quite 18dBm but about 12 or 13. My measurements are done with a calibrated instruments but it wouldn’t surprise me if there are variations
For the level to be +18dBm, his scope would have to read not 6Vp-p but 10Vp-p. RF instruments standardise on 50 ohm connections and are specified for power. His spectrum analyser's typical and it's seeing +13dBm (which for a 10MHz reference output is also typical). Across 50 ohms, that's very close to 1VRMS (which is about 2.8Vp-p for a sinewave). [The maths is: dBm = 10log(mW), so mW = 10**(dBm/10) = 10**(13/10) = 20 20mW across a 50 ohm load means a voltage of sqrt(0.02 * 50) = 1VRMS. For a sinewave, 1VRMS is sqrt(2) = 1.4Vpk = 2.8Vp-p.] The device's output impedance is also 50 ohms, so it's a Thévenin source of twice that voltage, with half dropped across the output impedance and half appearing across the analyser's input impedance. His oscilloscope's also typical: it displays voltage not power and presents an input impedance of 1 megohm not 50 ohms. That high impedance means the voltage-divider effect is negligible: the 'scope sees the whole Thévenin source of twice 2.8Vp-p. (It's displaying 6Vp-p which is pretty close.) The value's still +13dBm, it's just displayed differently (and in each case correctly) by two different instruments.
*Great Video, what's the difference between a 10MHZ Atomic clock, and a normal OCX crystal oscillator? I've heard that an atomic 10MHZ clock for pro audio purposes wod be over kill because 10MHZ is very accurate in the long term, but OCX is accurate in the short term*
In most cases an atomic clock contains a OCXO for stability as you mention. More info available here -blog.bliley.com/what-the-heck-is-an-ocxo-crystal-oscillator-facts-functionality
There are TCXOs with accuracy that rivals the OCXO. They are very expensive. Either can have very good long term stability. For most testing short term is all that's needed & perhaps more important. GPS satellites contain atomic clocks. That's how GPS works, the receiver compares time differences of the birds it sees to determine where it is in 3D. As such GPS time is excellent. Atomic clocks are arguably more accurate because the satellites & Earth aren't perfect in position. Plus space weather affects the signal. GPS is a convenient way to compare local time to a national standard which is of reach for most consumers. The U.S. standard is accurate to better than 1 second per 300 million years. One could also use a long wave signal, calculate for propagation delay, resulting in fantastically accurate timing.
No mention of the fact that these units lock slightly off frequency? There’s an extensive thread on the eevBlog forum about this issue. Depending on what you use this for, it may not matter, but if you have a GPSDO you expect it to be on 10 MHz to within atomic clock accuracy. That’s the point of using a GPSDO. In my case, I cannot use it as the timebase for my time of day clock because it drifts far too much. If you compare this to another GPSDO like a Thunderbolt you can see the 10MHz waveforms drifting against each other in a few hours. www.ke5fx.com/gpscomp.htm
Star Gazer Interesting and worrisome ...Thanks for raising this point. I note that the discrepancy noted bu EEVblog dates back a few years. I wonder if they have fixed it . Its interesting that it is low drift - but just off of the desired frequency.
@@johnpeterson7264 I don't know if the new units are fixed, I've not read the whole thread. I have one dated 2017-10-09 and it still has the problem, and it had had it for many versions of the unit across several years at that point. Again, not a big deal for someone who just wants a 10 MHz reference to frequency lock a rig or a transverter, but if you're expecting normal GPSDO accuracy over the long term, it doesn't do it. We're only talking 2.11E-11 low here. But still...it should be right. A GPSDO should trend toward 0 phase error in the long term.
The error produced on these GPS's as regards to frequency translates to 2Hz error at 100 GHz... that's more than enough frequency accuracy for most hams as documented on post 2 of the eevblog forum.
@@HamRadioDX And I mentioned that twice. However, you don't NEED a GPSDO if all you're doing is trying to frequency lock your transverter. In the situations where you NEED a GPSDO, you expect GPSDO level accuracy. It should be mentioned that this device doesn't meet that generally assumed criteria. Every other GPSDO I'm aware of does.
Hi, Nice video. I have a couple of questions I hope you can help with. I connected a RS232 cable through a prolific usb cable. It caused the mouse to jump all over the screen. on the cable pins 2,3,5,7 & 8 are connected straight through. Just wondering if I needed to use a null modem cable? The other thing is do you know of any software I can use to adjust the baud rate is, 9600 on the GPSDO? It's too fast for NMEA 0183 which is 4800. I want to hook it up to an old marine HF set.
Hi, I can’t answer the question related to the USB prolific cable. I have identical cables with that chip here and haven’t seen any issues. I’ve also connected using a standard RS232 cable and had similar success. As for the baud rate I am pretty sure that is set in the serial chip of the GPS. However do a search on the EEVBlog (and maybe even post) to see if someone has the answer to that question. Thanks for the comment!
Had the same problem today. Now I use a small CH340 TTL/USB converter, works like a charm. Nevertheless I had to add an TTL inverter stage to mine, because a transistor (Q1) in the GPSDO inverted the signal from the GPS receiver module. But I don't have the GPSDO shown in the video, I own a BG7TBL PLL-GPSDO 2020-02-15.
I've used a pre-owned video distribution amp. .....1990's era......analog NTSC/PAL type with lot's of BNC outputs. On e-bay they are dirt cheap these days. I swapped the 75ohm output resistors for 49.9 ohm types.(50 ohm). Some of them even have socketed opamps so there's an option for swapping, but not likely necessary because their bandwidth is already around 100Mhz. I am not sure is there perfect phase matching, to the original GPS 10Mhz source, if that matters.......but like I said, cheap cheap,cheap !!!
Hello everyone, I recently purchased a device like this but it takes me a long time to block the GPS, I mean more than 15 minutes sometimes. How long is the normal time to be operational? Can you change the antenna for a better one, with more gain or angle? Thank you very much
Hi Gerardo, sometimes it depends on the GPS chip or receiver, but 15 minutes seems a bit excessive. The GPS’s in the video only take about 20-25 seconds to lock with a good signal. You can try an external antenna outside for better performance?
Perhaps the GPS is doing a cold start each time. Two primary causes. 1 backup battery has failed. 2 long intervals between power ups. In both cases the GPS receiver needs to re-download the current constellation almanac which can often take 15min or more before handing off to user land
@@robertklug3863the output from the GPSDO is +13dBm, so if you use a 20dB attenuator that will mean -7dBm is feeding into the 9700. I think from memory the specs say to feed maximum of -10, but I have seen some put the full 13dBm in with no problems. To be safe I'd use a 30dB attenuator instead of a 20.
Does anyone know what the blinking green LED on the bottom of the board rear between the BLOX module and the board corner labeled D2 means. Mine seems to blink at a 1hz rate, but it may mean something else?
We know from the spectrum analyser reading that the power level is 13.88 dBm at 10 MHz. Assuming the input impedance of the analyser is exactly 50 ohms at this frequency, let's back things up. To determine the power dissipated, we calculate the antilog of 13.88/10, i.e. 10^(13.88/10). This results in 24.43 mW. Divided by 1000 gives 0.02443 W. Transposing P=E^2/R to solve for E, which will be the RMS value of the voltage, we end up with E=SQRT(P*R). SQRT(0.02443*50) = 1.105 Vrms. Multiplying this result by SQRT(2) to arrive at the peak voltage of 1.56. Multiplying Vp by 2 gives us 3.12 Vpp. And that's right in the ballpark for what I'm measuring on a 2014 model.
nice video. i'll be sure to pick one of these up on ebay. it would be nice if all the hams had these for their radios. then SSB would sound great, since we are all locked to gps standard. 73s see my channel for more radio fun.
Hayden! Thanks for the video. I was looking into the accuracy of my transverters and sdr radio. I'm trying to learn about this subject. Your video was the first to come up. I like your methodical and straight-forward explanation. See you next time you come over. 73 de Bill
Bill, legend! Thanks mate, are you looking at getting into the higher bands? Let me know 👍
Very timely! I just purchased a BG7TBL GPSDO to use with my IC-9700, and have been pondering what I would do if I ever manage to get my hands on an IC-7610. The distribution amplifier is just the ticket, obviously, so I'll pick up one of those, too. Thanks for another very useful video. Thanks, too, for the link to the attenuator. I've seen reports that even though the 9700 would be overdriven by the 10MHz output of the BG7TBL that it seems to work just fine without attenuation, but me being somewhat nervous about taking risks with such an expensive piece of gear, I was already looking for attenuators.
I was going to mention the attenuator is recommended! Good luck!
2:51 it does. It's CTS. They're a good make! It's almost certainly a second hand OCXO that used to be in an old cellular base station.
@2m50s: "...doesn't really have a brand..." It's CTS Corporation, a major supplier of oscillators and related parts.
I bought one of these GPSDO a while ago and it's a great little unit. I have seen where the max power input is 12.9V and have heard that some of these units don't like being fed with 13.8V from a power supply or field day battery so installed a cheap buck booster inside mine so it holds the incoming power voltage at 12.5v regardless of what the voltage is being presented to it. I also fitted an Anderson power pole to the back of it for easier power connection out in the field and less chance of accidental reverse polarization. That also helped with the over crowding on the front panel as well.
I've used a buck converter on most of my units that are constantly powered, although we have run them in the field from 13.8V without any issues.
For polarity protection, rig up a P Channel mosfet and you'll sleep better for it. Look up polarity protection mosfet and try it out
Thanks Hayden! Just bought one on EBay for freq. control of my ham rig.
Excellent, they are great value for money and a must in the shack
Ham Radio DX it’s all set up and working perfectly.
A CHEAP solution is you can use a video distribution amplifier. Granted it's 75 ohm and the output level depends on quality of the amps used. I have cheap one and it worked out pretty good. tnx for the video! 73 N8AUM
Does the fact that it is 75 ohms at BNC affect the use of distributing 10Mhz to 50 ohm test gear? How would one convert 75ohm to 50ohm?
Thanks for the great video. I checked all the comments but I can't find the solution. I have two questions:
1) can the 10 MHz GPSDO work just off the OCXO, without GPS?
2) I read in specs on eBay that GPSDO can achieve a frequency accuracy of 0.5 parts-per-billion. Is that for real or is it just marketing hype?
I can answer the first question - yes it can work without the GPS, obviously not as accurate but fairly stable. I can’t answer question 2 as I don’t have the test equipment to test
Thanks.
But without test equipment, is 1 ppb fairly standard across the industry for these types of devices. I am asking because I am a total newbie and I never came across such a high accuracy. I am a bit surprised.
Thanks for this nice review
Thanks!
Does the PPS when there is no GPS lock ? Is the phase of the PPS behaving with the 10MHz or is it jumping around Trig PPS and let 10MHz on 2nd trace ?
Hello from Belgium, what is the maximum length of the cable between the 10 MHz output of the BG7TBL and the input of the distribution amplifier?
Yours sincerely
I’m. It sure about maximum length. I use a short length of coax.
Maximum length would be dependent on loss - even RG-58 you can run a long a way at 10 MHZ with limited loss
I manage to put one of these and the distribution amplifier in the same case, but the 7805 regulator on the amplifier board get quite hot and make the box quite hot by hand… that could be armful for the ocxo? Maybe I must put them back separately
I think 6 volts p2p would equate more like 18DBm or there about. I use a 20 db pad on mine and that brings it down to around .7 volts or 2 DBm.
I noticed another unit I have has slightly more output. Not quite 18dBm but about 12 or 13. My measurements are done with a calibrated instruments but it wouldn’t surprise me if there are variations
For the level to be +18dBm, his scope would have to read not 6Vp-p but 10Vp-p.
RF instruments standardise on 50 ohm connections and are specified for power. His spectrum analyser's typical and it's seeing +13dBm (which for a 10MHz reference output is also typical).
Across 50 ohms, that's very close to 1VRMS (which is about 2.8Vp-p for a sinewave). [The maths is:
dBm = 10log(mW), so
mW = 10**(dBm/10)
= 10**(13/10)
= 20
20mW across a 50 ohm load means a voltage of sqrt(0.02 * 50) = 1VRMS.
For a sinewave, 1VRMS is sqrt(2) = 1.4Vpk = 2.8Vp-p.]
The device's output impedance is also 50 ohms, so it's a Thévenin source of twice that voltage, with half dropped across the output impedance and half appearing across the analyser's input impedance.
His oscilloscope's also typical: it displays voltage not power and presents an input impedance of 1 megohm not 50 ohms. That high impedance means the voltage-divider effect is negligible: the 'scope sees the whole Thévenin source of twice 2.8Vp-p. (It's displaying 6Vp-p which is pretty close.)
The value's still +13dBm, it's just displayed differently (and in each case correctly) by two different instruments.
Thanks Hayden, a nice video and very well presented. Jamie VK2YCJ
*Great Video, what's the difference between a 10MHZ Atomic clock, and a normal OCX crystal oscillator? I've heard that an atomic 10MHZ clock for pro audio purposes wod be over kill because 10MHZ is very accurate in the long term, but OCX is accurate in the short term*
In most cases an atomic clock contains a OCXO for stability as you mention. More info available here -blog.bliley.com/what-the-heck-is-an-ocxo-crystal-oscillator-facts-functionality
There are TCXOs with accuracy that rivals the OCXO. They are very expensive. Either can have very good long term stability. For most testing short term is all that's needed & perhaps more important.
GPS satellites contain atomic clocks. That's how GPS works, the receiver compares time differences of the birds it sees to determine where it is in 3D. As such GPS time is excellent.
Atomic clocks are arguably more accurate because the satellites & Earth aren't perfect in position. Plus space weather affects the signal. GPS is a convenient way to compare local time to a national standard which is of reach for most consumers. The U.S. standard is accurate to better than 1 second per 300 million years. One could also use a long wave signal, calculate for propagation delay, resulting in fantastically accurate timing.
No mention of the fact that these units lock slightly off frequency? There’s an extensive thread on the eevBlog forum about this issue. Depending on what you use this for, it may not matter, but if you have a GPSDO you expect it to be on 10 MHz to within atomic clock accuracy. That’s the point of using a GPSDO. In my case, I cannot use it as the timebase for my time of day clock because it drifts far too much. If you compare this to another GPSDO like a Thunderbolt you can see the 10MHz waveforms drifting against each other in a few hours.
www.ke5fx.com/gpscomp.htm
Star Gazer Interesting and worrisome ...Thanks for raising this point. I note that the discrepancy noted bu EEVblog dates back a few years. I wonder if they have fixed it . Its interesting that it is low drift - but just off of the desired frequency.
@@johnpeterson7264 I don't know if the new units are fixed, I've not read the whole thread. I have one dated 2017-10-09 and it still has the problem, and it had had it for many versions of the unit across several years at that point. Again, not a big deal for someone who just wants a 10 MHz reference to frequency lock a rig or a transverter, but if you're expecting normal GPSDO accuracy over the long term, it doesn't do it. We're only talking 2.11E-11 low here. But still...it should be right. A GPSDO should trend toward 0 phase error in the long term.
The error produced on these GPS's as regards to frequency translates to 2Hz error at 100 GHz... that's more than enough frequency accuracy for most hams as documented on post 2 of the eevblog forum.
@@HamRadioDX And I mentioned that twice. However, you don't NEED a GPSDO if all you're doing is trying to frequency lock your transverter. In the situations where you NEED a GPSDO, you expect GPSDO level accuracy. It should be mentioned that this device doesn't meet that generally assumed criteria. Every other GPSDO I'm aware of does.
Hi, Nice video. I have a couple of questions I hope you can help with. I connected a RS232 cable through a prolific usb cable. It caused the mouse to jump all over the screen. on the cable pins 2,3,5,7 & 8 are connected straight through. Just wondering if I needed to use a null modem cable? The other thing is do you know of any software I can use to adjust the baud rate is, 9600 on the GPSDO? It's too fast for NMEA 0183 which is 4800. I want to hook it up to an old marine HF set.
Hi,
I can’t answer the question related to the USB prolific cable. I have identical cables with that chip here and haven’t seen any issues. I’ve also connected using a standard RS232 cable and had similar success.
As for the baud rate I am pretty sure that is set in the serial chip of the GPS. However do a search on the EEVBlog (and maybe even post) to see if someone has the answer to that question.
Thanks for the comment!
Had the same problem today. Now I use a small CH340 TTL/USB converter, works like a charm. Nevertheless I had to add an TTL inverter stage to mine, because a transistor (Q1) in the GPSDO inverted the signal from the GPS receiver module. But I don't have the GPSDO shown in the video, I own a BG7TBL PLL-GPSDO 2020-02-15.
I've used a pre-owned video distribution amp. .....1990's era......analog NTSC/PAL type with lot's of BNC outputs. On e-bay they are dirt cheap these days. I swapped the 75ohm output resistors for 49.9 ohm types.(50 ohm). Some of them even have socketed opamps so there's an option for swapping, but not likely necessary because their bandwidth is already around 100Mhz.
I am not sure is there perfect phase matching, to the original GPS 10Mhz source, if that matters.......but like I said, cheap cheap,cheap !!!
what impedance are all the exits of both devices? High or 50 Ohm?
50 ohms
Hello everyone, I recently purchased a device like this but it takes me a long time to block the GPS, I mean more than 15 minutes sometimes. How long is the normal time to be operational? Can you change the antenna for a better one, with more gain or angle? Thank you very much
Hi Gerardo, sometimes it depends on the GPS chip or receiver, but 15 minutes seems a bit excessive. The GPS’s in the video only take about 20-25 seconds to lock with a good signal. You can try an external antenna outside for better performance?
Perhaps the GPS is doing a cold start each time. Two primary causes. 1 backup battery has failed. 2 long intervals between power ups. In both cases the GPS receiver needs to re-download the current constellation almanac which can often take 15min or more before handing off to user land
@@RobB_VK6ES thank you very much!
Download the news version of the software, it is nicer.
So where again does the 2W BNC attenuator go exactly? Between GPSDO and Icom 9700 , I'm thinking?
Hi Robert, yes you are correct. Between the GPSDO and the 9700 to reduce the level going into the radio.
@@HamRadioDX And that is a 20 Db attenuator, correct? for the Icom 9700?
@@robertklug3863the output from the GPSDO is +13dBm, so if you use a 20dB attenuator that will mean -7dBm is feeding into the 9700. I think from memory the specs say to feed maximum of -10, but I have seen some put the full 13dBm in with no problems. To be safe I'd use a 30dB attenuator instead of a 20.
Does anyone know what the blinking green LED on the bottom of the board rear between the BLOX module and the board corner labeled D2 means. Mine seems to blink at a 1hz rate, but it may mean something else?
Is it in sync with the GPSlock LED or one of the other LEDs on the front panel? Maybe a heart beat LED?
What is the peak to peak if driven into 50ohm?
Mine is 2.16V p-p into 50 ohms. It has a date code of 2017-10-09. There are many variations of this device made over the years.
We know from the spectrum analyser reading that the power level is 13.88 dBm at 10 MHz. Assuming the input impedance of the analyser is exactly 50 ohms at this frequency, let's back things up. To determine the power dissipated, we calculate the antilog of 13.88/10, i.e. 10^(13.88/10). This results in 24.43 mW. Divided by 1000 gives 0.02443 W. Transposing P=E^2/R to solve for E, which will be the RMS value of the voltage, we end up with E=SQRT(P*R). SQRT(0.02443*50) = 1.105 Vrms. Multiplying this result by SQRT(2) to arrive at the peak voltage of 1.56. Multiplying Vp by 2 gives us 3.12 Vpp. And that's right in the ballpark for what I'm measuring on a 2014 model.
No mentioning that the distribution amplifier puts out a 13dBm signal!!!! Way too much for using it with a radio like the IC-9700!
A 20 to 30dB attenuator is recommended to reduce the level
The amp should be easy enough to make yourselves, surely.
anyone know what the "ref in" on the kenwood ts890 would need? sine or square wave?
thanks
I paid about $70 for mine last year.
nice video. i'll be sure to pick one of these up on ebay. it would be nice if all the hams had these for their radios. then SSB would sound great, since we are all locked to gps standard. 73s see my channel for more radio fun.
you should buy us all one LOL :)
NMEA ≠ "Nema".
NEMA, NMEA?
I paid about $70 for mine last year.