|
Andy G4JNT
I'm using this TCXO in my 3.4GHz transverter and for a relative low cost frequency source, it's excellent. But I can't work out why there is a steady frequency drift from turn-on. It's only about 0.1ppm so is within spec, and occurs over the first 20 - 30 minutes of operation.
Isn't the whole point of a TCXO that it's temperature compensated ? Perhaps it's just compensation loop-gain is too low.
Guess I'm expecting too much :-)
I've just bought a batch of each of the low cost TCXOs that Farnell & RS have available in several different frequencies, 19.2, 26 and 40MHz and will investigate their performance as references. As modern Fract-N synths prefer a higher Fref than 10MHz, wondering if these will do, and they can always be locked to 10MHz separately.
|
|
|
I'd be interested to see what you find from those TCXOs and the frac-n chips, which give better performance. I was looking for a cheap source of the 26MHz ones for a 13cms transverter I hope to complete soon ... just need to complete a bit of the switching logic and the various enable/pwrdn bits. Should do 0.5 watt out and uses the SAW filters from Golledge ... the extra SMA is a feed of rx if to feed to a SDR dongle
I'm using this TCXO in my 3.4GHz transverter and for a relative low cost frequency source, it's excellent. But I can't work out why there is a steady frequency drift from turn-on. It's only about 0.1ppm so is within spec, and occurs over the first 20 - 30 minutes of operation.
Isn't the whole point of a TCXO that it's temperature compensated ? Perhaps it's just compensation loop-gain is too low.
Guess I'm expecting too much :-)
I've just bought a batch of each of the low cost TCXOs that Farnell & RS have available in several different frequencies, 19.2, 26 and 40MHz and will investigate their performance as references. As modern Fract-N synths prefer a higher Fref than 10MHz, wondering if these will do, and they can always be locked to 10MHz separately.
-- Best regards,
Robin Szemeti
Redpoint Consulting Limited
E: robin@...
T: +44 (0) 1299 405028
M: +44 (0) 7971 883371
CONFIDENTIALITY NOTICE
The information contained in this e-mail is intended only for the
confidential use of the above named recipient. If you are not the
intended recipient or person responsible for delivering it to the
intended recipient, you have received this communication in error
and must not distribute or copy it.
Please accept the sender's apologies, notify the sender immediately
by return e-mail and delete this communication.
Thank you.
|
|
|
Andy
Two effects happen when you turn on an OCXO. The first is getting the oven up to temperature, which can be quite quick, the other is the ‘ageing’ of the crystal. A quartz crystal will continually age with time which manifests as
a small change in frequency. From switch on the rate of change steadily reduces with time but is still measurable after a year.
Some (many!) years ago I worked in the quartz crystal division of Marconi’s when they were supplying the frequency references for the satellite tracking dishes. The prototype sat in the corner running continuously and after two
years the change was down to around 1 part in 10^11/year.
Ed G3VPF
toggle quoted messageShow quoted text
From: UKMicrowaves@groups.io <UKMicrowaves@groups.io> on behalf of Robin Szemeti - G1YFG <robin@...>
Sent: Friday, June 12, 2020 12:25:32 PM
To: UKMicrowaves@groups.io <UKMicrowaves@groups.io>
Subject: Re: [UKMicrowaves] TCXO Behaviour
I'd be interested to see what you find from those TCXOs and the frac-n chips, which give better performance. I was looking for a cheap source of the 26MHz ones for a 13cms transverter I hope to complete soon ... just need to complete a bit of the switching
logic and the various enable/pwrdn bits. Should do 0.5 watt out and uses the SAW filters from Golledge ... the extra SMA is a feed of rx if to feed to a SDR dongle
I'm using this TCXO
in my 3.4GHz transverter and for a relative low cost frequency source, it's excellent. But I can't work out why there is a steady frequency drift from turn-on. It's only about 0.1ppm so is within spec, and occurs over the first
20 - 30 minutes of operation.
Isn't the whole point of a TCXO that it's temperature compensated ? Perhaps it's just compensation loop-gain is too low.
Guess I'm expecting too much :-)
I've just bought a batch of each of the low cost TCXOs that Farnell & RS have available in several different frequencies, 19.2, 26 and 40MHz and will investigate their performance as references. As modern Fract-N synths prefer a higher Fref than 10MHz,
wondering if these will do, and they can always be locked to 10MHz separately.
--
Best regards,
Robin Szemeti
Redpoint Consulting Limited
E: robin@...
T: +44 (0) 1299 405028
M: +44 (0) 7971 883371
CONFIDENTIALITY NOTICE
The information contained in this e-mail is intended only for the
confidential use of the above named recipient. If you are not the
intended recipient or person responsible for delivering it to the
intended recipient, you have received this communication in error
and must not distribute or copy it.
Please accept the sender's apologies, notify the sender immediately
by return e-mail and delete this communication.
Thank you.
|
|
|
Andy G4JNT
OCXOs yes, but TCXOs shouldn't have a warm-up They should be on frequency instantly, hence my query about a [warm-up] drift each time they're turned on..
The frequency of a TCXO is controlled via a forward correction loop, either from a simple bit of analogue circuitry including a thermistor in the older ones, or a digital look-up table. Not sure what these use, but I suspect the digital route is cheaper.
On Fri, 12 Jun 2020 at 16:14, Ed G3VPF < g3vpf@...> wrote:
Andy
Two effects happen when you turn on an OCXO. The first is getting the oven up to temperature, which can be quite quick, the other is the ‘ageing’ of the crystal. A quartz crystal will continually age with time which manifests as
a small change in frequency. From switch on the rate of change steadily reduces with time but is still measurable after a year.
Some (many!) years ago I worked in the quartz crystal division of Marconi’s when they were supplying the frequency references for the satellite tracking dishes. The prototype sat in the corner running continuously and after two
years the change was down to around 1 part in 10^11/year.
Ed G3VPF
I'd be interested to see what you find from those TCXOs and the frac-n chips, which give better performance. I was looking for a cheap source of the 26MHz ones for a 13cms transverter I hope to complete soon ... just need to complete a bit of the switching
logic and the various enable/pwrdn bits. Should do 0.5 watt out and uses the SAW filters from Golledge ... the extra SMA is a feed of rx if to feed to a SDR dongle
I'm using this TCXO
in my 3.4GHz transverter and for a relative low cost frequency source, it's excellent. But I can't work out why there is a steady frequency drift from turn-on. It's only about 0.1ppm so is within spec, and occurs over the first
20 - 30 minutes of operation.
Isn't the whole point of a TCXO that it's temperature compensated ? Perhaps it's just compensation loop-gain is too low.
Guess I'm expecting too much :-)
I've just bought a batch of each of the low cost TCXOs that Farnell & RS have available in several different frequencies, 19.2, 26 and 40MHz and will investigate their performance as references. As modern Fract-N synths prefer a higher Fref than 10MHz,
wondering if these will do, and they can always be locked to 10MHz separately.
--
Best regards,
Robin Szemeti
Redpoint Consulting Limited
E: robin@...
T: +44 (0) 1299 405028
M: +44 (0) 7971 883371
CONFIDENTIALITY NOTICE
The information contained in this e-mail is intended only for the
confidential use of the above named recipient. If you are not the
intended recipient or person responsible for delivering it to the
intended recipient, you have received this communication in error
and must not distribute or copy it.
Please accept the sender's apologies, notify the sender immediately
by return e-mail and delete this communication.
Thank you.
|
|
|
Andy G4JNT
Made up these two test modules for 40MHz and 26MHz TCXOs from Farnell (different makes of module, chosen more-or-less at random)
I did a quick frequency test of the 40MHz one first of all, and over a 10 minute period (after it had been on for about 5 minutes already while I was adjusting the frequency). The drift was about 3Hz - it may have been less.
The 26MHz one has now been on for several hours. on a frequency counter while I've been out for a walk. Came back to find it just 0.8Hz off where I'd left it. However, opened the window to cool the shack down a bit, and teh frequency immediately dropped by 1.3Hz.
All well within spec, but for use in a microwave transverter you do want to keep them away from sudden temperature changes like drafts of wind. It'll be a while before I connect them up to a synth to check what teh short term multiplied performance is like
For anything above 5.76GHz I'd use an OCXO without question, if the LO is not going to be used locked to an external reference. There are some quite nice small ones doing the rounds at quite reasonable prices (and they're NOT chinese)
But these TCXOs do seem to offer a remarkably good performance for a few quid each.
One trouble with adjusting these is that all my 10-turn presets, being wirewound, adjust in steps. At the moment I just stick the whole preset across the 3.3V rail. To get really fine adjustment, the tuning voltage range from the preset needs to be more restricted by additional R top and bottom.
On Fri, 12 Jun 2020 at 12:25, Robin Szemeti - G1YFG < robin@...> wrote: I'd be interested to see what you find from those TCXOs and the frac-n chips, which give better performance. I was looking for a cheap source of the 26MHz ones for a 13cms transverter I hope to complete soon ... just need to complete a bit of the switching logic and the various enable/pwrdn bits. Should do 0.5 watt out and uses the SAW filters from Golledge ... the extra SMA is a feed of rx if to feed to a SDR dongle
I'm using this TCXO in my 3.4GHz transverter and for a relative low cost frequency source, it's excellent. But I can't work out why there is a steady frequency drift from turn-on. It's only about 0.1ppm so is within spec, and occurs over the first 20 - 30 minutes of operation.
Isn't the whole point of a TCXO that it's temperature compensated ? Perhaps it's just compensation loop-gain is too low.
Guess I'm expecting too much :-)
I've just bought a batch of each of the low cost TCXOs that Farnell & RS have available in several different frequencies, 19.2, 26 and 40MHz and will investigate their performance as references. As modern Fract-N synths prefer a higher Fref than 10MHz, wondering if these will do, and they can always be locked to 10MHz separately.
--
Best regards,
Robin Szemeti
Redpoint Consulting Limited
E: robin@...
T: +44 (0) 1299 405028
M: +44 (0) 7971 883371
CONFIDENTIALITY NOTICE
The information contained in this e-mail is intended only for the
confidential use of the above named recipient. If you are not the
intended recipient or person responsible for delivering it to the
intended recipient, you have received this communication in error
and must not distribute or copy it.
Please accept the sender's apologies, notify the sender immediately
by return e-mail and delete this communication.
Thank you.
|
|
|
Ian White
The ageing is a long-term effect, but Andy is
asking about changes on shorter timescales of minutes to a few
hours.
This takes us back to the dark art of temperature
compensated VFOs, and I believe the frequency changes happen for
the same reason. The three main frequency-determining elements -
the crystal, the transistors (which are
the main source of internal heating) and the
temperature compensating components (usually NTC capacitors) -
do not occupy a single point in space. Either internal or
external temperature changes arrive at those three elements at
different times, so even if the temperature compensation could
be perfect over the longer term, there will still be
shorter-term frequency excursions.
toggle quoted messageShow quoted text
On 12/06/2020 16:14, Ed G3VPF wrote:
Andy
Two effects happen when you turn
on an OCXO. The first is getting the oven up to temperature,
which can be quite quick, the other is the ‘ageing’ of the
crystal. A quartz crystal will continually age with time
which manifests as a small change in frequency. From switch
on the rate of change steadily reduces with time but is
still measurable after a year.
Some (many!) years ago I worked
in the quartz crystal division of Marconi’s when they were
supplying the frequency references for the satellite
tracking dishes. The prototype sat in the corner running
continuously and after two years the change was down to
around 1 part in 10^11/year.
Ed G3VPF
I'd be interested to see what you find from those TCXOs
and the frac-n chips, which give better performance. I was
looking for a cheap source of the 26MHz ones for a 13cms
transverter I hope to complete soon ... just need to
complete a bit of the switching logic and the various
enable/pwrdn bits. Should do 0.5 watt out and uses the SAW
filters from Golledge ... the extra SMA is a feed of rx if
to feed to a SDR dongle
I'm using this TCXO
in my 3.4GHz transverter
and for a relative low cost frequency source, it's
excellent. But I can't work out why there is a steady
frequency drift from turn-on. It's only about 0.1ppm
so is within spec, and occurs over the first 20 - 30
minutes of operation.
Isn't the whole point of a TCXO that it's temperature
compensated ? Perhaps it's just compensation loop-gain
is too low.
Guess I'm expecting too much :-)
I've just bought a batch of each of the low cost TCXOs
that Farnell & RS have available in several
different frequencies, 19.2, 26 and 40MHz and will
investigate their performance as references. As
modern Fract-N synths prefer a higher Fref than 10MHz,
wondering if these will do, and they can always be
locked to 10MHz separately.
--
Best regards,
Robin Szemeti
Redpoint Consulting Limited
E: robin@...
T: +44 (0) 1299 405028
M: +44 (0) 7971 883371
CONFIDENTIALITY NOTICE
The information contained in this e-mail is intended only for the
confidential use of the above named recipient. If you are not the
intended recipient or person responsible for delivering it to the
intended recipient, you have received this communication in error
and must not distribute or copy it.
Please accept the sender's apologies, notify the sender immediately
by return e-mail and delete this communication.
Thank you.
|
|
|
Andy G4JNT
Even that argument doesn't really apply here. The whole TCXO module is about 2.5 x 2mm, weighs less than a gram, and consumes just 2mA at 3V supply, so I can't believe differential heating between components over that time scale is remotely feasible..
I can only assume it's imperfect temperature compensation. It is within spec, but still a bit unexpected. The TCXO really ought to be renamed PTCXO; Partially Temperature Compensated XO.
On Fri, 12 Jun 2020 at 21:09, Ian White < gm3sek@...> wrote:
The ageing is a long-term effect, but Andy is
asking about changes on shorter timescales of minutes to a few
hours.
This takes us back to the dark art of temperature
compensated VFOs, and I believe the frequency changes happen for
the same reason. The three main frequency-determining elements -
the crystal, the transistors (which are
the main source of internal heating) and the
temperature compensating components (usually NTC capacitors) -
do not occupy a single point in space. Either internal or
external temperature changes arrive at those three elements at
different times, so even if the temperature compensation could
be perfect over the longer term, there will still be
shorter-term frequency excursions.
On 12/06/2020 16:14, Ed G3VPF wrote:
Andy
Two effects happen when you turn
on an OCXO. The first is getting the oven up to temperature,
which can be quite quick, the other is the ‘ageing’ of the
crystal. A quartz crystal will continually age with time
which manifests as a small change in frequency. From switch
on the rate of change steadily reduces with time but is
still measurable after a year.
Some (many!) years ago I worked
in the quartz crystal division of Marconi’s when they were
supplying the frequency references for the satellite
tracking dishes. The prototype sat in the corner running
continuously and after two years the change was down to
around 1 part in 10^11/year.
Ed G3VPF
I'd be interested to see what you find from those TCXOs
and the frac-n chips, which give better performance. I was
looking for a cheap source of the 26MHz ones for a 13cms
transverter I hope to complete soon ... just need to
complete a bit of the switching logic and the various
enable/pwrdn bits. Should do 0.5 watt out and uses the SAW
filters from Golledge ... the extra SMA is a feed of rx if
to feed to a SDR dongle
I'm using this TCXO
in my 3.4GHz transverter
and for a relative low cost frequency source, it's
excellent. But I can't work out why there is a steady
frequency drift from turn-on. It's only about 0.1ppm
so is within spec, and occurs over the first 20 - 30
minutes of operation.
Isn't the whole point of a TCXO that it's temperature
compensated ? Perhaps it's just compensation loop-gain
is too low.
Guess I'm expecting too much :-)
I've just bought a batch of each of the low cost TCXOs
that Farnell & RS have available in several
different frequencies, 19.2, 26 and 40MHz and will
investigate their performance as references. As
modern Fract-N synths prefer a higher Fref than 10MHz,
wondering if these will do, and they can always be
locked to 10MHz separately.
--
Best regards,
Robin Szemeti
Redpoint Consulting Limited
E: robin@...
T: +44 (0) 1299 405028
M: +44 (0) 7971 883371
CONFIDENTIALITY NOTICE
The information contained in this e-mail is intended only for the
confidential use of the above named recipient. If you are not the
intended recipient or person responsible for delivering it to the
intended recipient, you have received this communication in error
and must not distribute or copy it.
Please accept the sender's apologies, notify the sender immediately
by return e-mail and delete this communication.
Thank you.
|
|
|
Out of curiosity, did you test the higher frequency TCXOs and come to any conclusions?
I'm using this TCXO in my 3.4GHz transverter and for a relative low cost frequency source, it's excellent. But I can't work out why there is a steady frequency drift from turn-on. It's only about 0.1ppm so is within spec, and occurs over the first 20 - 30 minutes of operation.
Isn't the whole point of a TCXO that it's temperature compensated ? Perhaps it's just compensation loop-gain is too low.
Guess I'm expecting too much :-)
I've just bought a batch of each of the low cost TCXOs that Farnell & RS have available in several different frequencies, 19.2, 26 and 40MHz and will investigate their performance as references. As modern Fract-N synths prefer a higher Fref than 10MHz, wondering if these will do, and they can always be locked to 10MHz separately.
--
Robin Szemeti - G1YFG
|
|
|
Hi Robin and others, just for info I’m using a series of these in various applications. (e-bay number
132652769518) The 25mhz one I’m using (via an LC filter) to stabilise my QO-100 LNB, it is 10hz (as measured by G4DDK) out but very stable so 10Hz x 390 means my RX
is 3.9KHz out..... but hey-ho, beggars and all that.
Also use a 101MHz example to lock my 70cm ICENI (The xtal I used is not particularly good) and a 116MHz example in my ongoing 2m transverter project.
BTW I run the 25MHz ref and the 595MHz osc in my QO-100 downconverter 24/7/365.
Enjoy this lovely sunny Sunday, my missus says I have to put the soldering iron down and take her out !
Regards
Colin G8LBS.
Sent from
Mail for Windows 10
toggle quoted messageShow quoted text
From: Robin Szemeti - G1YFG
Sent: 06 September 2020 00:35
To: UKMicrowaves@groups.io
Subject: Re: [UKMicrowaves] TCXO Behaviour
Out of curiosity, did you test the higher frequency TCXOs and come to any conclusions?
in my 3.4GHz transverter and for a relative low cost frequency source, it's excellent. But
I can't work out why there is a steady frequency drift from turn-on. It's only about 0.1ppm so is within spec, and occurs over the first 20 - 30 minutes of operation.
Isn't the whole point of a TCXO that it's temperature compensated ? Perhaps it's just compensation loop-gain is too low.
Guess I'm expecting too much :-)
I've just bought a batch of each of the low cost TCXOs that Farnell & RS have available in several different frequencies, 19.2, 26 and 40MHz and will investigate their performance
as references. As modern Fract-N synths prefer a higher Fref than 10MHz, wondering if these will do, and they can always be locked to 10MHz separately.
--
Robin Szemeti - G1YFG
|
|
|
Interesting ... I have an almost identical device in my 817 as a "high precision" reference. It is the second one I have had as the first died after a couple of years. I can't say the one in the 817 is particularly stable. Would you say yours is significantly better in terms of freq stability than a standard crystal?
On Sun, 6 Sep 2020 at 10:46, Colin Ranson < g8lbs@...> wrote:
Hi Robin and others, just for info I’m using a series of these in various applications. (e-bay number
132652769518) The 25mhz one I’m using (via an LC filter) to stabilise my QO-100 LNB, it is 10hz (as measured by G4DDK) out but very stable so 10Hz x 390 means my RX
is 3.9KHz out..... but hey-ho, beggars and all that.
Also use a 101MHz example to lock my 70cm ICENI (The xtal I used is not particularly good) and a 116MHz example in my ongoing 2m transverter project.
BTW I run the 25MHz ref and the 595MHz osc in my QO-100 downconverter 24/7/365.
Enjoy this lovely sunny Sunday, my missus says I have to put the soldering iron down and take her out !
Regards
Colin G8LBS.
Sent from
Mail for Windows 10
Out of curiosity, did you test the higher frequency TCXOs and come to any conclusions?
in my 3.4GHz transverter and for a relative low cost frequency source, it's excellent. But
I can't work out why there is a steady frequency drift from turn-on. It's only about 0.1ppm so is within spec, and occurs over the first 20 - 30 minutes of operation.
Isn't the whole point of a TCXO that it's temperature compensated ? Perhaps it's just compensation loop-gain is too low.
Guess I'm expecting too much :-)
I've just bought a batch of each of the low cost TCXOs that Farnell & RS have available in several different frequencies, 19.2, 26 and 40MHz and will investigate their performance
as references. As modern Fract-N synths prefer a higher Fref than 10MHz, wondering if these will do, and they can always be locked to 10MHz separately.
--
Robin Szemeti - G1YFG
--
Robin Szemeti - G1YFG
|
|
|
Hi Robin,
Just on my way out......... yes, the 101MHz example has better stability and accuracy than the cheap Czech xtal in the transverter which is a bit touchy getting onto frequency and I can’t fault the 25MHz one for stability.
Regards
Colin.
Sent from
Mail for Windows 10
toggle quoted messageShow quoted text
From: Robin Szemeti - G1YFG
Sent: 06 September 2020 11:38
To: UKMicrowaves@groups.io
Subject: Re: [UKMicrowaves] TCXO Behaviour
Interesting ... I have an almost identical device in my 817 as a "high precision" reference. It is the second one I have had as the first died after a couple of years. I can't say the one in the 817 is particularly stable. Would you say
yours is significantly better in terms of freq stability than a standard crystal?
On Sun, 6 Sep 2020 at 10:46, Colin Ranson <g8lbs@...> wrote:
Hi Robin and others, just for info I’m using a series of these in various applications. (e-bay number
132652769518) The 25mhz one I’m using (via an LC filter) to stabilise my QO-100 LNB, it is 10hz (as measured by G4DDK) out but very stable so 10Hz x 390 means my RX
is 3.9KHz out..... but hey-ho, beggars and all that.
Also use a 101MHz example to lock my 70cm ICENI (The xtal I used is not particularly good) and a 116MHz example in my ongoing 2m transverter project.
BTW I run the 25MHz ref and the 595MHz osc in my QO-100 downconverter 24/7/365.
Enjoy this lovely sunny Sunday, my missus says I have to put the soldering iron down and take her out !
Regards
Colin G8LBS.
Sent from
Mail for Windows 10
Out of curiosity, did you test the higher frequency TCXOs and come to any conclusions?
in my 3.4GHz transverter and for a relative low cost frequency source, it's excellent. But I can't work out why
there is a steady frequency drift from turn-on. It's only about 0.1ppm so is within spec, and occurs over the first 20 - 30 minutes of operation.
Isn't the whole point of a TCXO that it's temperature compensated ? Perhaps it's just compensation loop-gain is too low.
Guess I'm expecting too much :-)
I've just bought a batch of each of the low cost TCXOs that Farnell & RS have available in several different frequencies, 19.2, 26 and 40MHz and will investigate their performance as references.
As modern Fract-N synths prefer a higher Fref than 10MHz, wondering if these will do, and they can always be locked to 10MHz separately.
--
Robin Szemeti - G1YFG
--
Robin Szemeti - G1YFG
|
|
|
Fair enough, that looks like the easy option then.
Pity the output is on tag strip and not an SMA, but, cheap enough, can't complain I guess!
Thanks for the recommendation, I'll get one ordered!
On Sun, 6 Sep 2020 at 12:06, Colin Ranson < g8lbs@...> wrote:
Hi Robin,
Just on my way out......... yes, the 101MHz example has better stability and accuracy than the cheap Czech xtal in the transverter which is a bit touchy getting onto frequency and I can’t fault the 25MHz one for stability.
Regards
Colin.
Sent from
Mail for Windows 10
Interesting ... I have an almost identical device in my 817 as a "high precision" reference. It is the second one I have had as the first died after a couple of years. I can't say the one in the 817 is particularly stable. Would you say
yours is significantly better in terms of freq stability than a standard crystal?
On Sun, 6 Sep 2020 at 10:46, Colin Ranson <g8lbs@...> wrote:
Hi Robin and others, just for info I’m using a series of these in various applications. (e-bay number
132652769518) The 25mhz one I’m using (via an LC filter) to stabilise my QO-100 LNB, it is 10hz (as measured by G4DDK) out but very stable so 10Hz x 390 means my RX
is 3.9KHz out..... but hey-ho, beggars and all that.
Also use a 101MHz example to lock my 70cm ICENI (The xtal I used is not particularly good) and a 116MHz example in my ongoing 2m transverter project.
BTW I run the 25MHz ref and the 595MHz osc in my QO-100 downconverter 24/7/365.
Enjoy this lovely sunny Sunday, my missus says I have to put the soldering iron down and take her out !
Regards
Colin G8LBS.
Sent from
Mail for Windows 10
Out of curiosity, did you test the higher frequency TCXOs and come to any conclusions?
in my 3.4GHz transverter and for a relative low cost frequency source, it's excellent. But I can't work out why
there is a steady frequency drift from turn-on. It's only about 0.1ppm so is within spec, and occurs over the first 20 - 30 minutes of operation.
Isn't the whole point of a TCXO that it's temperature compensated ? Perhaps it's just compensation loop-gain is too low.
Guess I'm expecting too much :-)
I've just bought a batch of each of the low cost TCXOs that Farnell & RS have available in several different frequencies, 19.2, 26 and 40MHz and will investigate their performance as references.
As modern Fract-N synths prefer a higher Fref than 10MHz, wondering if these will do, and they can always be locked to 10MHz separately.
--
Robin Szemeti - G1YFG
--
Robin Szemeti - G1YFG
--
Robin Szemeti - G1YFG
|
|
|
Andy G4JNT
I only did a quick test on each. All seem to hold their frequency quite well to a fraction of a PPM, far better then a straight crystal would. But some seem to have a long time constant in their compensation. If you waft air over them the frequency changes rapidly, only by a fraction of a PPM, but sudile (at uWaves) , enough to be annoying. So if using for GHz, they do need to be inside a closed container in still air - and definitely away from any cooling fans.
BTW, I encountered one, once that was a frequency hopper. Sounded like JT4 when multiplied up - those are designed / touted to overcome EMC regs by spreading interference. Or it may have been digital temperature compensation that was applied via a 4 bit D/A and PRN mod. If that was the case, it would have been a pretty clever bit of design.
On Sun, 6 Sep 2020 at 00:35, Robin Szemeti - G1YFG < robin@...> wrote: Out of curiosity, did you test the higher frequency TCXOs and come to any conclusions?
I'm using this TCXO in my 3.4GHz transverter and for a relative low cost frequency source, it's excellent. But I can't work out why there is a steady frequency drift from turn-on. It's only about 0.1ppm so is within spec, and occurs over the first 20 - 30 minutes of operation.
Isn't the whole point of a TCXO that it's temperature compensated ? Perhaps it's just compensation loop-gain is too low.
Guess I'm expecting too much :-)
I've just bought a batch of each of the low cost TCXOs that Farnell & RS have available in several different frequencies, 19.2, 26 and 40MHz and will investigate their performance as references. As modern Fract-N synths prefer a higher Fref than 10MHz, wondering if these will do, and they can always be locked to 10MHz separately.
--
Robin Szemeti - G1YFG
|
|
|
Noted, thanks for the update. Just need something to tame this very wobbly crystal in the PLL ... I suspect it may be made from cheese. I'll try one with a block of "something" around it to shield it ... polystyrene or similar.
I only did a quick test on each. All seem to hold their frequency quite well to a fraction of a PPM, far better then a straight crystal would. But some seem to have a long time constant in their compensation. If you waft air over them the frequency changes rapidly, only by a fraction of a PPM, but sudile (at uWaves) , enough to be annoying. So if using for GHz, they do need to be inside a closed container in still air - and definitely away from any cooling fans.
BTW, I encountered one, once that was a frequency hopper. Sounded like JT4 when multiplied up - those are designed / touted to overcome EMC regs by spreading interference. Or it may have been digital temperature compensation that was applied via a 4 bit D/A and PRN mod. If that was the case, it would have been a pretty clever bit of design.
On Sun, 6 Sep 2020 at 00:35, Robin Szemeti - G1YFG < robin@...> wrote: Out of curiosity, did you test the higher frequency TCXOs and come to any conclusions?
I'm using this TCXO in my 3.4GHz transverter and for a relative low cost frequency source, it's excellent. But I can't work out why there is a steady frequency drift from turn-on. It's only about 0.1ppm so is within spec, and occurs over the first 20 - 30 minutes of operation.
Isn't the whole point of a TCXO that it's temperature compensated ? Perhaps it's just compensation loop-gain is too low.
Guess I'm expecting too much :-)
I've just bought a batch of each of the low cost TCXOs that Farnell & RS have available in several different frequencies, 19.2, 26 and 40MHz and will investigate their performance as references. As modern Fract-N synths prefer a higher Fref than 10MHz, wondering if these will do, and they can always be locked to 10MHz separately.
--
Robin Szemeti - G1YFG
--
Robin Szemeti - G1YFG
|
|
|
And as an update, I fitted the "golden" TCXO to the 13cm transverter and it is indeed very stable as described, can't fault it. I tuned in the Telford beacon when the shed was nice and warm one evening left it running. It was still the same tone the next morning when the shed was somewhat cooler. It appears to be excellent, blowing some warm air over it has minimal effect, certainly for SSB purposes, it is more than adequate.
Noted, thanks for the update. Just need something to tame this very wobbly crystal in the PLL ... I suspect it may be made from cheese. I'll try one with a block of "something" around it to shield it ... polystyrene or similar.
I only did a quick test on each. All seem to hold their frequency quite well to a fraction of a PPM, far better then a straight crystal would. But some seem to have a long time constant in their compensation. If you waft air over them the frequency changes rapidly, only by a fraction of a PPM, but sudile (at uWaves) , enough to be annoying. So if using for GHz, they do need to be inside a closed container in still air - and definitely away from any cooling fans.
BTW, I encountered one, once that was a frequency hopper. Sounded like JT4 when multiplied up - those are designed / touted to overcome EMC regs by spreading interference. Or it may have been digital temperature compensation that was applied via a 4 bit D/A and PRN mod. If that was the case, it would have been a pretty clever bit of design.
On Sun, 6 Sep 2020 at 00:35, Robin Szemeti - G1YFG < robin@...> wrote: Out of curiosity, did you test the higher frequency TCXOs and come to any conclusions?
I'm using this TCXO in my 3.4GHz transverter and for a relative low cost frequency source, it's excellent. But I can't work out why there is a steady frequency drift from turn-on. It's only about 0.1ppm so is within spec, and occurs over the first 20 - 30 minutes of operation.
Isn't the whole point of a TCXO that it's temperature compensated ? Perhaps it's just compensation loop-gain is too low.
Guess I'm expecting too much :-)
I've just bought a batch of each of the low cost TCXOs that Farnell & RS have available in several different frequencies, 19.2, 26 and 40MHz and will investigate their performance as references. As modern Fract-N synths prefer a higher Fref than 10MHz, wondering if these will do, and they can always be locked to 10MHz separately.
--
Robin Szemeti - G1YFG
--
Robin Szemeti - G1YFG
--
Robin Szemeti - G1YFG
|
|
|
Robin,
Pleased you are happy with your ‘Golden’ TCXO, as I said before, can’t fault them for the price. I run my 101MHz (70cm) and 25MHz (QO-100) examples 24/7/365 as is the 116MHz one that will run in an eventual 2m transverter. I would
like them to ‘cut’ me a 25.788++++++++++MHz for QO-100 432MHz IF from an LNB but I dare not ask the price for a ‘one off’ !!
Good luck
Regards
Colin de G8LBS.
Sent from
Mail for Windows 10
toggle quoted messageShow quoted text
From: Robin Szemeti - G1YFG
Sent: 26 September 2020 00:33
To: UKMicrowaves@groups.io
Subject: Re: [UKMicrowaves] TCXO Behaviour
And as an update, I fitted the "golden" TCXO to the 13cm transverter and it is indeed very stable as described, can't fault it. I tuned in the Telford beacon when the shed was nice and warm one evening left it running. It was still the
same tone the next morning when the shed was somewhat cooler. It appears to be excellent, blowing some warm air over it has minimal effect, certainly for SSB purposes, it is more than adequate.
Noted, thanks for the update. Just need something to tame this very wobbly crystal in the PLL ... I suspect it may be made from cheese. I'll try one with a block of "something" around it to shield it ... polystyrene or similar.
I only did a quick test on each. All seem to hold their frequency quite well to a fraction of a PPM, far better then a straight crystal would. But some seem to have a long
time constant in their compensation. If you waft air over them the frequency changes rapidly, only by a fraction of a PPM, but sudile (at uWaves) , enough to be annoying. So if using for GHz, they do need to be inside a closed container in still air
- and definitely away from any cooling fans.
BTW, I encountered one, once that was a frequency hopper. Sounded like JT4 when multiplied up - those are designed / touted to overcome EMC regs by spreading interference.
Or it may have been digital temperature compensation that was applied via a 4 bit D/A and PRN mod. If that was the case, it would have been a pretty clever bit of design.
On Sun, 6 Sep 2020 at 00:35, Robin Szemeti - G1YFG <robin@...> wrote:
Out of curiosity, did you test the higher frequency TCXOs and come to any conclusions?
in my 3.4GHz transverter and for a relative low cost frequency source, it's excellent. But I can't work out why there is a steady frequency drift from turn-on.
It's only about 0.1ppm so is within spec, and occurs over the first 20 - 30 minutes of operation.
Isn't the whole point of a TCXO that it's temperature compensated ? Perhaps it's just compensation loop-gain is too low.
Guess I'm expecting too much :-)
I've just bought a batch of each of the low cost TCXOs that Farnell & RS have available in several different frequencies, 19.2, 26 and 40MHz and will investigate their performance
as references. As modern Fract-N synths prefer a higher Fref than 10MHz, wondering if these will do, and they can always be locked to 10MHz separately.
--
Robin Szemeti - G1YFG
--
Robin Szemeti - G1YFG
--
Robin Szemeti - G1YFG
|
|
|
1