Hello All, This is for the medium-precision nuts amongst us. I have invested £14.50 in one of these Vanguard 0.1ppm TCXO's to pimp my Siglent FG which has a footprint for it. It is its internal ref at 25 MHz. The swapping-in was simple to do and the FG works fine with it. Since it has a counter mode I did a quick check to show whether it was worthwile: I made it count the 10 MHz from my Proteus GPSDO. The Siglent reads to 1 Hz and straight after turn-on it went straight to 10.000000 MHz. Over about 1 day I never saw it more than +/-1 Hz off, and this involved a deliberate temperature change of just over 5 deg C, basically by having the heating in the room off and on. For most of the time the display sat solidly at 10.000000. So, to summarise, in a rough sort of way the thing lives up to its 0.1ppm spec, at least around the 20C temperature mark. I bought it from a Hong Kong seller on Ebay - naturally there is always a chance that other devices sold with the same description/label might not perform as well. I'm quite aware that the generator (DDS) suffers from other sources of error, which won't be improved by the clock being better, but at least the nominal frequencies it outputs are now going to be very close to the mark without the need for an external ref. Kind regards, Thomas.

What is the phase noise like on the cheap TCXO? Being on frequency is far from the only characteristic that matters. I did some research on cheap TCXOs for Kenwood radios and found truly awful phase noise and lots of spurs above and below the oscillator frequency. I don't know if these characteristics are important for your unit. One eBay source says -125 dBc/1kHz. (Should be stated as -125 dBc/Hz @ 1 kHz). That is not very good. I found the cheap TCXOs I tested actually had worse phase noise at 15 - 20 kHz offset. Mark W7MLG On Feb 8, 2018 4:24 AM, "Thomas Allgeier" <th.allgeier@gmail.com> wrote: > Hello All, > > This is for the medium-precision nuts amongst us. I have invested £14.50 > in one of these Vanguard 0.1ppm TCXO's to pimp my Siglent FG which has a > footprint for it. It is its internal ref at 25 MHz. The swapping-in was > simple to do and the FG works fine with it. Since it has a counter mode I > did a quick check to show whether it was worthwile: I made it count the 10 > MHz from my Proteus GPSDO. The Siglent reads to 1 Hz and straight after > turn-on it went straight to 10.000000 MHz. Over about 1 day I never saw it > more than +/-1 Hz off, and this involved a deliberate temperature change of > just over 5 deg C, basically by having the heating in the room off and on. > For most of the time the display sat solidly at 10.000000. > > So, to summarise, in a rough sort of way the thing lives up to its 0.1ppm > spec, at least around the 20C temperature mark. I bought it from a Hong > Kong seller on Ebay - naturally there is always a chance that other devices > sold with the same description/label might not perform as well. > > I'm quite aware that the generator (DDS) suffers from other sources of > error, which won't be improved by the clock being better, but at least the > nominal frequencies it outputs are now going to be very close to the mark > without the need for an external ref. > > Kind regards, > > Thomas. > > _______________________________________________ > time-nuts mailing list -- time-nuts@febo.com > To unsubscribe, go to https://www.febo.com/cgi-bin/m > ailman/listinfo/time-nuts > and follow the instructions there. >

On 2/8/18 5:34 AM, Mark Goldberg wrote: > What is the phase noise like on the cheap TCXO? Being on frequency is far > from the only characteristic that matters. I did some research on cheap > TCXOs for Kenwood radios and found truly awful phase noise and lots of > spurs above and below the oscillator frequency. I don't know if these > characteristics are important for your unit. > > One eBay source says -125 dBc/1kHz. (Should be stated as -125 dBc/Hz @ 1 > kHz). That is not very good. I found the cheap TCXOs I tested actually had > worse phase noise at 15 - 20 kHz offset. > Vectron VT-702 (the first one in the list online) -99 at 10Hz -123 at 100Hz -143 at 1kHz going to Mouser and looking for the cheapest 10MHz TCXO FOX924 (about $2) - no data SiT5000 (about ) - -140 dBc@1kHz, -150 dBcfrom 10k to 100k, -160 @ 1M (no price) ECS -TXO-3225-100-TR ($2.71 each) - -135dBc @ 1 kHz ASTX-H11 ($3.16) -130dBc @ 1 kHz, -158dBc @100k FOX922CE at 16.369 MHz, -145 @ 10kHz

On Thu, Feb 8, 2018 at 10:51 AM, jimlux <jimlux@earthlink.net> wrote: > On 2/8/18 5:34 AM, Mark Goldberg wrote: > >> What is the phase noise like on the cheap TCXO?Vectron >> > > > VT-702 (the first one in the list online) > -99 at 10Hz > -123 at 100Hz > -143 at 1kHz > > going to Mouser and looking for the cheapest 10MHz TCXO > FOX924 (about $2) - no data > SiT5000 (about ) - -140 dBc@1kHz, -150 dBcfrom 10k to 100k, -160 @ 1M (no > price) > ECS -TXO-3225-100-TR ($2.71 each) - -135dBc @ 1 kHz > ASTX-H11 ($3.16) -130dBc @ 1 kHz, -158dBc @100k > > FOX922CE at 16.369 MHz, -145 @ 10kHz > > > Can you provide a link to the "list online"? I went to eBay and searched for "Vanguard TCXO" and got a list of them with various frequencies. They all specify -125 dBc/Hz at 1 khz, which is concerning of itself, as phase noise generally is higher with higher frequency. They are golden however, so there's that! Vectron and others make decent TCXOs, better than -135 dBc/Hz at 1 kHz, which is what I have measured. I believe none of them are what the original poster is talking about. For a comparison, a Wenzel OCXO I am using for a reference is in the range of -155 dBc/Hz at 1 kHz. I am always wary when no or few specifications are provided. If you look at manufacturer;s like Vectron, they provide lots of data, phase noise at many frequencies, aging, etc. 73, Mark W7MLG

On 2/8/18 11:49 AM, Mark Goldberg wrote: > On Thu, Feb 8, 2018 at 10:51 AM, jimlux <jimlux@earthlink.net> wrote: > >> On 2/8/18 5:34 AM, Mark Goldberg wrote: >> >>> What is the phase noise like on the cheap TCXO?Vectron >>> >> >> >> VT-702 (the first one in the list online) >> -99 at 10Hz >> -123 at 100Hz >> -143 at 1kHz >> >> going to Mouser and looking for the cheapest 10MHz TCXO >> FOX924 (about $2) - no data >> SiT5000 (about ) - -140 dBc@1kHz, -150 dBcfrom 10k to 100k, -160 @ 1M (no >> price) >> ECS -TXO-3225-100-TR ($2.71 each) - -135dBc @ 1 kHz >> ASTX-H11 ($3.16) -130dBc @ 1 kHz, -158dBc @100k >> >> FOX922CE at 16.369 MHz, -145 @ 10kHz >> >> >> > Can you provide a link to the "list online"? I just went to digikey and searched for TCXO and 10 MHz > I am always wary when no or few specifications are provided. If you look at > manufacturer;s like Vectron, they provide lots of data, phase noise at many > frequencies, aging, etc. > it's what you pay for - if the spec says -135 at 10kHz, and that's all, then that's what they test. - that's cheaper. A $100 oscillator will tend to have a lot more data than a $2 one <grin> At some point, you just BUY a batch of oscillators and test them yourself.