Re: [time-nuts] Double balanced mixer question
I am not quite sure what you want to accomplish but a diode mixer will have a number of undesirable products in its output signal. An analog multiplier such as AD834 should give you a much cleaner output. You still need to filter out the unwanted product. -- Manfred On Fri, Jul 24, 2020 at 5:56 AM wrote: > Hi, > > I'm feeding 5.0 MHZ and 5.01MHz into an HP10514A mixer. > > A buffer and a 12dB attenuator feed each input and a 50 Ohm buffer amp > (10Mhz) is on the output. > > I get a nice sine output but get the 1Hz as amplitude variations. > > Playing with input levels I can minimize the variations but the best I > can get is a 3.2 V P-P with a .4 V P-P amplitude modulation. > > Are there mixer schemes I can use that will eliminate the amplitude > variations? > > Cheers, > > Corby > ___ > time-nuts mailing list -- time-nuts@lists.febo.com > To unsubscribe, go to > http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com > and follow the instructions there. > -- -- Manfred VK3AES ___ time-nuts mailing list -- time-nuts@lists.febo.com To unsubscribe, go to http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com and follow the instructions there.
Re: [time-nuts] Double balanced mixer question
On Fri, 24 Jul 2020 01:37:04 +0200 Attila Kinali wrote: > Due to a diode mixer not being a good multiplier (it only > multiplies the signs and adds up the amplitudes... and has > offset voltages when the switching happens) you will > get quite strong second order components like 2*LO-RF. > The only way to dampen them is to make the mixer as symmetric > as possible or to use a push-pull kind of architecture to > cancel out second order components (e.g., a double-double-balanced > mixer) Blub.. I should not write mails when I'm tired Let me correct this: It would not be 2nd order components, but 3rd order components 3*LO-RF and 2nd order harmonic (2*LO, 2*RF) leakage. While the latter can be reduced by better symmetry, the former is intrinsic in the working of the double balanced diode mixer. I don't know of any way how to reduce that, but I am not a specialist in mixer design, so that doesn't mean anything. Attila Kinali -- Science is made up of so many things that appear obvious after they are explained. -- Pardot Kynes ___ time-nuts mailing list -- time-nuts@lists.febo.com To unsubscribe, go to http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com and follow the instructions there.
Re: [time-nuts] Double balanced mixer question
On Thu, 23 Jul 2020 11:01:28 -0700 wrote: > I'm feeding 5.0 MHZ and 5.01MHz into an HP10514A mixer. > > A buffer and a 12dB attenuator feed each input and a 50 Ohm buffer amp > (10Mhz) is on the output. [...] > Are there mixer schemes I can use that will eliminate the amplitude > variations? I guess you have read Enrico's Mixer Tutorial[1]? If not, I highly recommend it. One very important lesson it contains is that you have to short the signal component you do not want at the mixer. This means you have to offer the mixer a low impedance path at the mixer output to ground. An good inductor with high self-resonance (>20MHz) and inductance higher than ~13µH should do the job. Due to a diode mixer not being a good multiplier (it only multiplies the signs and adds up the amplitudes... and has offset voltages when the switching happens) you will get quite strong second order components like 2*LO-RF. The only way to dampen them is to make the mixer as symmetric as possible or to use a push-pull kind of architecture to cancel out second order components (e.g., a double-double-balanced mixer) As you are using a pretty old mixer, my guess it is using discrete diodes. I would recommend replacing those either by some matched diode array (HSMS-2827 and HSMS-2829 come to mind) or use matched BJT arrays in diode configuration as they are depicted in [2] (e.g. THAT300) to make the switching more symmetric. What you can also try is to drive the two input ports with square wave signals. This should minimize second order harmonics generation within the mixer. But I have not gone through the math for this and thus cannot say how effective it would be. Attila Kinali [1] "Tutorial on the double balanced mixer", by Enrico Rubiola, 2006 http://rubiola.org/pdf-articles/archives/2006-arxiv-0608211v1-mixer-tutorial.pdf [2] "Residual PM Noise Evaluation of Radio Frequency Mixers", by Barnes, Hati, Nelson, Howe, 2011 (available on the NIST document server) -- The bad part of Zurich is where the degenerates throw DARK chocolate at you. ___ time-nuts mailing list -- time-nuts@lists.febo.com To unsubscribe, go to http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com and follow the instructions there.
Re: [time-nuts] Double balanced mixer question
Peter, it* is* too simplistic. It's very likely that the 10 MHz is also being *phase* modulated at 1 Hz, and clipping will not remove that. Dana On Thu, Jul 23, 2020 at 5:23 PM Peter McCollum wrote: > This may be too simplistic, but you could clip it with a limiter to remove > the amplitude variations, then filter it back to a sine. > > Pete > > > On Thu, Jul 23, 2020 at 1:56 PM wrote: > > > Hi, > > > > I'm feeding 5.0 MHZ and 5.01MHz into an HP10514A mixer. > > > > A buffer and a 12dB attenuator feed each input and a 50 Ohm buffer amp > > (10Mhz) is on the output. > > > > I get a nice sine output but get the 1Hz as amplitude variations. > > > > Playing with input levels I can minimize the variations but the best I > > can get is a 3.2 V P-P with a .4 V P-P amplitude modulation. > > > > Are there mixer schemes I can use that will eliminate the amplitude > > variations? > > > > Cheers, > > > > Corby > > ___ > > time-nuts mailing list -- time-nuts@lists.febo.com > > To unsubscribe, go to > > http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com > > and follow the instructions there. > > > ___ > time-nuts mailing list -- time-nuts@lists.febo.com > To unsubscribe, go to > http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com > and follow the instructions there. > ___ time-nuts mailing list -- time-nuts@lists.febo.com To unsubscribe, go to http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com and follow the instructions there.
Re: [time-nuts] Double balanced mixer question
Hi Corby, On 2020-07-23 20:01, cdel...@juno.com wrote: > Hi, > > I'm feeding 5.0 MHZ and 5.01MHz into an HP10514A mixer. > > A buffer and a 12dB attenuator feed each input and a 50 Ohm buffer amp > (10Mhz) is on the output. > > I get a nice sine output but get the 1Hz as amplitude variations. > > Playing with input levels I can minimize the variations but the best I > can get is a 3.2 V P-P with a .4 V P-P amplitude modulation. > > Are there mixer schemes I can use that will eliminate the amplitude > variations? If you high-pass filte rout the 1 Hz, do you still get variations? Depending on how you detect these, the 1 Hz may confuse you. It can be a bit easier if you have larger frequency difference to start with, since you can analyze the 10 MHz easier with spectrum analyzer or SSB receiver. Cheers, Magnus ___ time-nuts mailing list -- time-nuts@lists.febo.com To unsubscribe, go to http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com and follow the instructions there.
Re: [time-nuts] Double balanced mixer question
This may be too simplistic, but you could clip it with a limiter to remove the amplitude variations, then filter it back to a sine. Pete On Thu, Jul 23, 2020 at 1:56 PM wrote: > Hi, > > I'm feeding 5.0 MHZ and 5.01MHz into an HP10514A mixer. > > A buffer and a 12dB attenuator feed each input and a 50 Ohm buffer amp > (10Mhz) is on the output. > > I get a nice sine output but get the 1Hz as amplitude variations. > > Playing with input levels I can minimize the variations but the best I > can get is a 3.2 V P-P with a .4 V P-P amplitude modulation. > > Are there mixer schemes I can use that will eliminate the amplitude > variations? > > Cheers, > > Corby > ___ > time-nuts mailing list -- time-nuts@lists.febo.com > To unsubscribe, go to > http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com > and follow the instructions there. > ___ time-nuts mailing list -- time-nuts@lists.febo.com To unsubscribe, go to http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com and follow the instructions there.
Re: [time-nuts] Double balanced mixer question
Is the 1 Hz difference signal the one you want, or do you want the sum? The sum will include 10.01 MHz but also 9.99 MHz, and 10.02 and 9.98 and 10.00, and so forth, and a whole lot of other products - basically the sums and differences of every fundamental and harmonic of every frequency involved, at various levels (and phase). For investigating this, it's best to use a spectrum analyzer too. The scope can see the envelope of the modulation, but not the frequency content. The 1 Hz difference signal is easy to LPF out, but the sums have lots of stuff close together, nearly impossible to separate with filters. If you start with a quadrature mixing scheme, you can get much better rejection of what you don't want, but there will still be plenty of spurious stuff, because it can't be done perfectly - as always, it depends on how good it needs to be. I'm guessing the objective is to make a 10 MHz +/- 1 Hz offset reference frequency. Unfortunately, it's very difficult with simple or even fancy up-converting analog mixing and filtering. Better to use a fundamental XO at that frequency, or a DDS, or start with frequencies way higher than the 10 MHz, that can be mixed down to the 10 MHz range, and readily filtered out. Most of the garbage you don't want will be much higher up, and a LPF can be made as good as needed to reject it. Depending on your 5 MHz sources, a good possibility, presuming they can be tuned or PLLed, is instead of adding them, just set one for say 1/2 Hz away from 5 MHz, then double it to 10 MHz +/- 1 Hz, which will be way easier to separate from the fundamental and higher harmonics with a BPF. Ed ___ time-nuts mailing list -- time-nuts@lists.febo.com To unsubscribe, go to http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com and follow the instructions there.
Re: [time-nuts] Double balanced mixer question
I think you basically ended up with an amplitude modulated (AM) signal with 10MHz+1 as a carrier and 1Hz as side-band. I'll be trying something like this soon. If you use low pass filter at near audio frequency, it should eliminate this. --- (Mr.) Taka Kamiya KB4EMF / ex JF2DKG On Thursday, July 23, 2020, 4:50:13 PM EDT, Richard (Rick) Karlquist wrote: Are you filtering out the 10.01 MHz image? Rick N6RK On 7/23/2020 11:01 AM, cdel...@juno.com wrote: > Hi, > > I'm feeding 5.0 MHZ and 5.01MHz into an HP10514A mixer. > > A buffer and a 12dB attenuator feed each input and a 50 Ohm buffer amp > (10Mhz) is on the output. > > I get a nice sine output but get the 1Hz as amplitude variations. > > Playing with input levels I can minimize the variations but the best I > can get is a 3.2 V P-P with a .4 V P-P amplitude modulation. > > Are there mixer schemes I can use that will eliminate the amplitude > variations? > > Cheers, > > Corby > ___ > time-nuts mailing list -- time-nuts@lists.febo.com > To unsubscribe, go to > http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com > and follow the instructions there. > ___ time-nuts mailing list -- time-nuts@lists.febo.com To unsubscribe, go to http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com and follow the instructions there. ___ time-nuts mailing list -- time-nuts@lists.febo.com To unsubscribe, go to http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com and follow the instructions there.
Re: [time-nuts] Double balanced mixer question
Hi Mixing those two frequencies *should* give you 10.01 MHz and 1 Hz at equal levels. Any amplitude inequality will be due to filtering. Feed two tones like that into various amps and you will drive them into compression. You need a good linear amp to handle both at once. Indeed some form of fancy filtering to prevent the 1 Hz from getting to the amp is also an answer. The gotcah is that a narrowband filter likely will give you a bit of temperature dependent delay. Bob > On Jul 23, 2020, at 2:01 PM, cdel...@juno.com wrote: > > Hi, > > I'm feeding 5.0 MHZ and 5.01MHz into an HP10514A mixer. > > A buffer and a 12dB attenuator feed each input and a 50 Ohm buffer amp > (10Mhz) is on the output. > > I get a nice sine output but get the 1Hz as amplitude variations. > > Playing with input levels I can minimize the variations but the best I > can get is a 3.2 V P-P with a .4 V P-P amplitude modulation. > > Are there mixer schemes I can use that will eliminate the amplitude > variations? > > Cheers, > > Corby > ___ > time-nuts mailing list -- time-nuts@lists.febo.com > To unsubscribe, go to > http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com > and follow the instructions there. ___ time-nuts mailing list -- time-nuts@lists.febo.com To unsubscribe, go to http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com and follow the instructions there.
Re: [time-nuts] Double balanced mixer question
Are you filtering out the 10.01 MHz image? Rick N6RK On 7/23/2020 11:01 AM, cdel...@juno.com wrote: Hi, I'm feeding 5.0 MHZ and 5.01MHz into an HP10514A mixer. A buffer and a 12dB attenuator feed each input and a 50 Ohm buffer amp (10Mhz) is on the output. I get a nice sine output but get the 1Hz as amplitude variations. Playing with input levels I can minimize the variations but the best I can get is a 3.2 V P-P with a .4 V P-P amplitude modulation. Are there mixer schemes I can use that will eliminate the amplitude variations? Cheers, Corby ___ time-nuts mailing list -- time-nuts@lists.febo.com To unsubscribe, go to http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com and follow the instructions there. ___ time-nuts mailing list -- time-nuts@lists.febo.com To unsubscribe, go to http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com and follow the instructions there.
Re: [time-nuts] WWVB PM Time Questions
Understood and did plenty of them. That means I am unclear as to whats required. But will for the moment let this drop. Its not my thread. Regards Paul. On Thu, Jul 23, 2020 at 3:56 PM Bob kb8tq wrote: > Hi > > Keep in mind that the EFC range is part of the control loop parameters. > When you change EFC range, you need to recalculate the control loop. > That likely means putting in different components in the control loop. > > Bob > > > On Jul 23, 2020, at 10:22 AM, paul swed wrote: > > > > Bob > > I attempted to use a OCXO a Piezo and could never get it to hold lock. > > Did invert the control and such. Anyhow, just wondering if I need a > simpler > > sloppy VCTXO. > > No time to tinker. But I believe I have some really nice little 10 MHz > > vectron VCTCXOs. > > If that worked then what I am dealing with is an efc range control issue. > > Perhaps the swing is to large... > > Regards > > Paul > > WB8TSL > > > > On Thu, Jul 23, 2020 at 8:58 AM Bob kb8tq wrote: > > > >> Hi > >> > >> 10 MHz *is* a frequency at which you can find a pretty good $3 to $6 > OCXO. > >> I find that a bit amazing, but (at least right now) that’s how it seems > to > >> work. > >> (and yes, they do have EFC). > >> > >> Bob > >> > >> > >>> On Jul 22, 2020, at 10:09 PM, paul swed wrote: > >>> > >>> Looked at the KD2BD schematic and it is a 10 MHz VCTCXO. That seems > >>> reasonable perhaps to find. > >>> Regards > >>> Paul > >>> > >>> On Wed, Jul 22, 2020 at 6:44 PM paul swed wrote: > >>> > I was actually thinking about this quite a bit and perhaps I went the > wrong way. > That is a cheap vcxo class device. His unit was not an oven as I > recall. > Just maybe a bit too good was actually bad. > Hard to say and no time to play right now. > Regards > Paul > WB8TSL > > On Wed, Jul 22, 2020 at 6:06 PM Bob kb8tq wrote: > > > Hi > > > > > > > >> On Jul 22, 2020, at 3:51 PM, paul swed wrote: > >> > >> Ray watch out for my comment on the KD2BD solution. That oscillator > > isn't > >> available and I have not been able to map something else into it. > >> Tried > >> several good grade Oven oscillators. Just be aware of that issue. > >> Would need to do more tinkering and simply don't have that time > right > > now. > >> Also it would be great if the oscillator was something that could be > >> obtained at a reasonable cost. I do not believe at all it has to be > a > >> OCXO as the older true time and spectracoms were not and they locked > > solid. > >> So its a case of getting the control voltages right. > > > > You can buy a surprisingly good little OCXO (for the price) on eBay > >> these > > days in the $3 to $6 range. Find one at 12.6 MHz and it will divide > > nicely to > > 60 KHz. 19.2 MHz looks like a better candidate … Either way I have > not > > (yet) > > seen a 60 KHz multiple show up in the cheap category. > > > > Bob > > > > > >> Regards > >> Paul. > >> > >> On Wed, Jul 22, 2020 at 3:29 PM wrote: > >> > >>> Thanks to all for the suggestions. I dug a ferrite rod AM radio > >> antenna > >>> out of the box this morning. I have a box of 10 left over from the > >> late > >>> 80s. It measures 0.950 mH so I will add some turns to get it to 1.5 > >> mH > >>> which will be easier to resonate at 60 kHz. Then I need to build up > >> the > >>> RF amp and run the output through a 60 kHz crystal filter. I have > to > >>> order the crystals since I don't have any on hand. It will take a > > couple > >>> of weeks to get the front end working. > >>> > >>> I am in Phoenix so the WWVB signal is of decent quality here even > > during > >>> the daytime. My 25 year old cheap Casio watch will sync up within > 2-3 > >>> minutes any time of day or night. It normally syncs at 2 am but I > >> have > >>> done tests to confirm daytime sync works reliably. > >>> > >>> Mark, I know virtually nothing about SDR other than it works. I > >> bought > > a > >>> RTL-SDR.Com module a couple of years ago and played with it a > little > >>> bit. But it quit working and I haven't thought about SDR since > then. > >> My > >>> Icom 7300 is a SDR and it works very well. I think to attempt to > >> design > >>> a SDR would be well over my head. I anticipate this project taking > >> 2-3 > >>> months. If I get it working I will be glad to share everything with > >> the > >>> group. > >>> > >>> Lester and Paul, I will test with the Costas loop as it may be the > >>> easiest way to go. To test the Costas loop I am basically going to > >>> duplicate the KD2BD design. I was reluctant about it in the > beginning > >>> but more reading seems to indicate it will be fine. My AWG allows > me > >> to > >>> set the phase from 0 to 360 degrees independently on each channel > so > >> I >
Re: [time-nuts] WWVB PM Time Questions
Mark you really are in the right place. The fact is Time-nuts has a huge range of skill sets. To your question. Can't resist "No pain no gain". The KB2DB design is a good and very real design. I have built it and it is whats needed when you are receiving 60 KHz covered in all kinds of noise these days. Also look at some of the older designs like spectracoms. You can get those schematics online. My antenna in Boston is a 10' X 10' square loop and preamp some 800' of wire or 400? Been a bit of time. It gives me very nice signals in the day time 100 uv or more level and at night crazy high levels 300-600 to 1000 uv. The loop can be turned to reduce noise. You don't need a loop that big. But I have been very happy with it over the last 7 years. The challenge today is that way back when you could buy real transformers and crystals to make very good TRF radios. Thats all gone so OP-amps it is and a hot soldering iron. Ah the smell of flux in the morning. I did build a radio with the little 60 KHz crystals you can buy for $2 or so from China. A bag of 20. Sorted through them looking for a reasonable unit to use. The bandwidth required is about 2 Hz. You can go wider and a pure opamp solution will be wider. It just allows more noise in. So get that soldering iron out and feel the pain. I for one want to watch the dsp side of what you do on a STM32. Good luck. Regards Paul WB8TSL On Thu, Jul 23, 2020 at 3:56 PM Mark Haun wrote: > On 23-Jul-20 4:35 AM, Detlef Schuecker via time-nuts wrote: > > Hi, > > > >> principles. The STM32L4 series which I often use has a pretty decent > >> ADC: fast (5 MSPS), with about 11 good bits in differential mode, and > >> "proper" hardware downsampling (called the DFSDM in the manual). If the > > Yes, the STM32 series also have built-in OPAmps, so one could hook up a > > loop or a ferrite directly to the uC without too much external > components. > > Sampling at 160ks/s should suffice to get the phase and DCF77 is in > reach > > as well. At this rate you have ~500-1000 processor ticks per sample > which > > should be enough to do real time demodulation. > > Are there any examples (schematics) out there for the front-end > electronics? I haven't found much by googling except the KD2BD design > which is more involved than I would like. The integrated designs used > in "atomic" clocks seem very simple, but I am unsure how to duplicate > them. I would like to use a crystal filer, but I'm at a bit of a loss > to start. For example, would one choose series or parallel resonance? > I believe the impedance of these tuning forks is quite high at series > resonance---tens of kohms. I didn't make much progress the last time I > tried doodling op-amp circuits to use one. > > My general impression of this list is that a lot of folks are pretty > comfortable with analog and discrete digital design, but find DSP and > algorithms "hard." I'm exactly the opposite so maybe I am in the right > place ;) > > Mark > > > ___ > time-nuts mailing list -- time-nuts@lists.febo.com > To unsubscribe, go to > http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com > and follow the instructions there. > ___ time-nuts mailing list -- time-nuts@lists.febo.com To unsubscribe, go to http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com and follow the instructions there.
Re: [time-nuts] WWVB PM Time Questions
Hi Keep in mind that the EFC range is part of the control loop parameters. When you change EFC range, you need to recalculate the control loop. That likely means putting in different components in the control loop. Bob > On Jul 23, 2020, at 10:22 AM, paul swed wrote: > > Bob > I attempted to use a OCXO a Piezo and could never get it to hold lock. > Did invert the control and such. Anyhow, just wondering if I need a simpler > sloppy VCTXO. > No time to tinker. But I believe I have some really nice little 10 MHz > vectron VCTCXOs. > If that worked then what I am dealing with is an efc range control issue. > Perhaps the swing is to large... > Regards > Paul > WB8TSL > > On Thu, Jul 23, 2020 at 8:58 AM Bob kb8tq wrote: > >> Hi >> >> 10 MHz *is* a frequency at which you can find a pretty good $3 to $6 OCXO. >> I find that a bit amazing, but (at least right now) that’s how it seems to >> work. >> (and yes, they do have EFC). >> >> Bob >> >> >>> On Jul 22, 2020, at 10:09 PM, paul swed wrote: >>> >>> Looked at the KD2BD schematic and it is a 10 MHz VCTCXO. That seems >>> reasonable perhaps to find. >>> Regards >>> Paul >>> >>> On Wed, Jul 22, 2020 at 6:44 PM paul swed wrote: >>> I was actually thinking about this quite a bit and perhaps I went the wrong way. That is a cheap vcxo class device. His unit was not an oven as I recall. Just maybe a bit too good was actually bad. Hard to say and no time to play right now. Regards Paul WB8TSL On Wed, Jul 22, 2020 at 6:06 PM Bob kb8tq wrote: > Hi > > > >> On Jul 22, 2020, at 3:51 PM, paul swed wrote: >> >> Ray watch out for my comment on the KD2BD solution. That oscillator > isn't >> available and I have not been able to map something else into it. >> Tried >> several good grade Oven oscillators. Just be aware of that issue. >> Would need to do more tinkering and simply don't have that time right > now. >> Also it would be great if the oscillator was something that could be >> obtained at a reasonable cost. I do not believe at all it has to be a >> OCXO as the older true time and spectracoms were not and they locked > solid. >> So its a case of getting the control voltages right. > > You can buy a surprisingly good little OCXO (for the price) on eBay >> these > days in the $3 to $6 range. Find one at 12.6 MHz and it will divide > nicely to > 60 KHz. 19.2 MHz looks like a better candidate … Either way I have not > (yet) > seen a 60 KHz multiple show up in the cheap category. > > Bob > > >> Regards >> Paul. >> >> On Wed, Jul 22, 2020 at 3:29 PM wrote: >> >>> Thanks to all for the suggestions. I dug a ferrite rod AM radio >> antenna >>> out of the box this morning. I have a box of 10 left over from the >> late >>> 80s. It measures 0.950 mH so I will add some turns to get it to 1.5 >> mH >>> which will be easier to resonate at 60 kHz. Then I need to build up >> the >>> RF amp and run the output through a 60 kHz crystal filter. I have to >>> order the crystals since I don't have any on hand. It will take a > couple >>> of weeks to get the front end working. >>> >>> I am in Phoenix so the WWVB signal is of decent quality here even > during >>> the daytime. My 25 year old cheap Casio watch will sync up within 2-3 >>> minutes any time of day or night. It normally syncs at 2 am but I >> have >>> done tests to confirm daytime sync works reliably. >>> >>> Mark, I know virtually nothing about SDR other than it works. I >> bought > a >>> RTL-SDR.Com module a couple of years ago and played with it a little >>> bit. But it quit working and I haven't thought about SDR since then. >> My >>> Icom 7300 is a SDR and it works very well. I think to attempt to >> design >>> a SDR would be well over my head. I anticipate this project taking >> 2-3 >>> months. If I get it working I will be glad to share everything with >> the >>> group. >>> >>> Lester and Paul, I will test with the Costas loop as it may be the >>> easiest way to go. To test the Costas loop I am basically going to >>> duplicate the KD2BD design. I was reluctant about it in the beginning >>> but more reading seems to indicate it will be fine. My AWG allows me >> to >>> set the phase from 0 to 360 degrees independently on each channel so >> I >>> can use that for initial testing. That testing can be done prior to >>> getting the front end working. >>> >>> Richard, is your software posted somewhere? I assume it allows you to >>> specify the time and date you want it to send. That would come in >> handy >>> for writing the code to extract the timer/date data. I have seen >>> examples of the Arduino transmitting the old AM signal but don't >> recall >>> seeing one that
Re: [time-nuts] WWVB PM Time Questions
On 23-Jul-20 4:35 AM, Detlef Schuecker via time-nuts wrote: > Hi, > >> principles. The STM32L4 series which I often use has a pretty decent >> ADC: fast (5 MSPS), with about 11 good bits in differential mode, and >> "proper" hardware downsampling (called the DFSDM in the manual). If the > Yes, the STM32 series also have built-in OPAmps, so one could hook up a > loop or a ferrite directly to the uC without too much external components. > Sampling at 160ks/s should suffice to get the phase and DCF77 is in reach > as well. At this rate you have ~500-1000 processor ticks per sample which > should be enough to do real time demodulation. Are there any examples (schematics) out there for the front-end electronics? I haven't found much by googling except the KD2BD design which is more involved than I would like. The integrated designs used in "atomic" clocks seem very simple, but I am unsure how to duplicate them. I would like to use a crystal filer, but I'm at a bit of a loss to start. For example, would one choose series or parallel resonance? I believe the impedance of these tuning forks is quite high at series resonance---tens of kohms. I didn't make much progress the last time I tried doodling op-amp circuits to use one. My general impression of this list is that a lot of folks are pretty comfortable with analog and discrete digital design, but find DSP and algorithms "hard." I'm exactly the opposite so maybe I am in the right place ;) Mark ___ time-nuts mailing list -- time-nuts@lists.febo.com To unsubscribe, go to http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com and follow the instructions there.
[time-nuts] Double balanced mixer question
Hi, I'm feeding 5.0 MHZ and 5.01MHz into an HP10514A mixer. A buffer and a 12dB attenuator feed each input and a 50 Ohm buffer amp (10Mhz) is on the output. I get a nice sine output but get the 1Hz as amplitude variations. Playing with input levels I can minimize the variations but the best I can get is a 3.2 V P-P with a .4 V P-P amplitude modulation. Are there mixer schemes I can use that will eliminate the amplitude variations? Cheers, Corby ___ time-nuts mailing list -- time-nuts@lists.febo.com To unsubscribe, go to http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com and follow the instructions there.
Re: [time-nuts] WWVB PM Time Questions
Bob I attempted to use a OCXO a Piezo and could never get it to hold lock. Did invert the control and such. Anyhow, just wondering if I need a simpler sloppy VCTXO. No time to tinker. But I believe I have some really nice little 10 MHz vectron VCTCXOs. If that worked then what I am dealing with is an efc range control issue. Perhaps the swing is to large... Regards Paul WB8TSL On Thu, Jul 23, 2020 at 8:58 AM Bob kb8tq wrote: > Hi > > 10 MHz *is* a frequency at which you can find a pretty good $3 to $6 OCXO. > I find that a bit amazing, but (at least right now) that’s how it seems to > work. > (and yes, they do have EFC). > > Bob > > > > On Jul 22, 2020, at 10:09 PM, paul swed wrote: > > > > Looked at the KD2BD schematic and it is a 10 MHz VCTCXO. That seems > > reasonable perhaps to find. > > Regards > > Paul > > > > On Wed, Jul 22, 2020 at 6:44 PM paul swed wrote: > > > >> I was actually thinking about this quite a bit and perhaps I went the > >> wrong way. > >> That is a cheap vcxo class device. His unit was not an oven as I recall. > >> Just maybe a bit too good was actually bad. > >> Hard to say and no time to play right now. > >> Regards > >> Paul > >> WB8TSL > >> > >> On Wed, Jul 22, 2020 at 6:06 PM Bob kb8tq wrote: > >> > >>> Hi > >>> > >>> > >>> > On Jul 22, 2020, at 3:51 PM, paul swed wrote: > > Ray watch out for my comment on the KD2BD solution. That oscillator > >>> isn't > available and I have not been able to map something else into it. > Tried > several good grade Oven oscillators. Just be aware of that issue. > Would need to do more tinkering and simply don't have that time right > >>> now. > Also it would be great if the oscillator was something that could be > obtained at a reasonable cost. I do not believe at all it has to be a > OCXO as the older true time and spectracoms were not and they locked > >>> solid. > So its a case of getting the control voltages right. > >>> > >>> You can buy a surprisingly good little OCXO (for the price) on eBay > these > >>> days in the $3 to $6 range. Find one at 12.6 MHz and it will divide > >>> nicely to > >>> 60 KHz. 19.2 MHz looks like a better candidate … Either way I have not > >>> (yet) > >>> seen a 60 KHz multiple show up in the cheap category. > >>> > >>> Bob > >>> > >>> > Regards > Paul. > > On Wed, Jul 22, 2020 at 3:29 PM wrote: > > > Thanks to all for the suggestions. I dug a ferrite rod AM radio > antenna > > out of the box this morning. I have a box of 10 left over from the > late > > 80s. It measures 0.950 mH so I will add some turns to get it to 1.5 > mH > > which will be easier to resonate at 60 kHz. Then I need to build up > the > > RF amp and run the output through a 60 kHz crystal filter. I have to > > order the crystals since I don't have any on hand. It will take a > >>> couple > > of weeks to get the front end working. > > > > I am in Phoenix so the WWVB signal is of decent quality here even > >>> during > > the daytime. My 25 year old cheap Casio watch will sync up within 2-3 > > minutes any time of day or night. It normally syncs at 2 am but I > have > > done tests to confirm daytime sync works reliably. > > > > Mark, I know virtually nothing about SDR other than it works. I > bought > >>> a > > RTL-SDR.Com module a couple of years ago and played with it a little > > bit. But it quit working and I haven't thought about SDR since then. > My > > Icom 7300 is a SDR and it works very well. I think to attempt to > design > > a SDR would be well over my head. I anticipate this project taking > 2-3 > > months. If I get it working I will be glad to share everything with > the > > group. > > > > Lester and Paul, I will test with the Costas loop as it may be the > > easiest way to go. To test the Costas loop I am basically going to > > duplicate the KD2BD design. I was reluctant about it in the beginning > > but more reading seems to indicate it will be fine. My AWG allows me > to > > set the phase from 0 to 360 degrees independently on each channel so > I > > can use that for initial testing. That testing can be done prior to > > getting the front end working. > > > > Richard, is your software posted somewhere? I assume it allows you to > > specify the time and date you want it to send. That would come in > handy > > for writing the code to extract the timer/date data. I have seen > > examples of the Arduino transmitting the old AM signal but don't > recall > > seeing one that sends the BPSK stream. > > > > I have a GPS time receiver with a 1.2 inch LED display that I built a > > couple of years ago. I have it setup so the UART outputs the gprmc > NEMA > > string in case I want to look at it. I have a couple of spare GPS > > modules lying around and several Arduino Uno modules. I have no > problem >
[time-nuts] "Best Practices" for cascading L1 GPS Antenna Distribution Amplifiers
I am wondering if anyone either knows (or can point me to some pertinent information regarding) the "best practices" for feeding L1 to two HP/Symmetricom 58517A (8 port) Distribution Amplifiers? 1. Can I simply choose one of the first amplifier’s ports and use that to feed the L1 RF to the input of the second unit? 2. What about the DC needs of the second amplifier? 3. Or, should I use, say, a 58512A (2 port) unit and distribute its’ 2 outputs, one to each of the inputs of the two 58517A units? 4. Or, perhaps it doesn’t really matter, as the 58512A outputs are ultimately the same thing as using one of the first 58517A outputs to feed the second 58517A? Thoughts appreciated. Jim Robbins N1JR ___ time-nuts mailing list -- time-nuts@lists.febo.com To unsubscribe, go to http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com and follow the instructions there.
Re: [time-nuts] WWVB PM Time Questions
Hi 10 MHz *is* a frequency at which you can find a pretty good $3 to $6 OCXO. I find that a bit amazing, but (at least right now) that’s how it seems to work. (and yes, they do have EFC). Bob > On Jul 22, 2020, at 10:09 PM, paul swed wrote: > > Looked at the KD2BD schematic and it is a 10 MHz VCTCXO. That seems > reasonable perhaps to find. > Regards > Paul > > On Wed, Jul 22, 2020 at 6:44 PM paul swed wrote: > >> I was actually thinking about this quite a bit and perhaps I went the >> wrong way. >> That is a cheap vcxo class device. His unit was not an oven as I recall. >> Just maybe a bit too good was actually bad. >> Hard to say and no time to play right now. >> Regards >> Paul >> WB8TSL >> >> On Wed, Jul 22, 2020 at 6:06 PM Bob kb8tq wrote: >> >>> Hi >>> >>> >>> On Jul 22, 2020, at 3:51 PM, paul swed wrote: Ray watch out for my comment on the KD2BD solution. That oscillator >>> isn't available and I have not been able to map something else into it. Tried several good grade Oven oscillators. Just be aware of that issue. Would need to do more tinkering and simply don't have that time right >>> now. Also it would be great if the oscillator was something that could be obtained at a reasonable cost. I do not believe at all it has to be a OCXO as the older true time and spectracoms were not and they locked >>> solid. So its a case of getting the control voltages right. >>> >>> You can buy a surprisingly good little OCXO (for the price) on eBay these >>> days in the $3 to $6 range. Find one at 12.6 MHz and it will divide >>> nicely to >>> 60 KHz. 19.2 MHz looks like a better candidate … Either way I have not >>> (yet) >>> seen a 60 KHz multiple show up in the cheap category. >>> >>> Bob >>> >>> Regards Paul. On Wed, Jul 22, 2020 at 3:29 PM wrote: > Thanks to all for the suggestions. I dug a ferrite rod AM radio antenna > out of the box this morning. I have a box of 10 left over from the late > 80s. It measures 0.950 mH so I will add some turns to get it to 1.5 mH > which will be easier to resonate at 60 kHz. Then I need to build up the > RF amp and run the output through a 60 kHz crystal filter. I have to > order the crystals since I don't have any on hand. It will take a >>> couple > of weeks to get the front end working. > > I am in Phoenix so the WWVB signal is of decent quality here even >>> during > the daytime. My 25 year old cheap Casio watch will sync up within 2-3 > minutes any time of day or night. It normally syncs at 2 am but I have > done tests to confirm daytime sync works reliably. > > Mark, I know virtually nothing about SDR other than it works. I bought >>> a > RTL-SDR.Com module a couple of years ago and played with it a little > bit. But it quit working and I haven't thought about SDR since then. My > Icom 7300 is a SDR and it works very well. I think to attempt to design > a SDR would be well over my head. I anticipate this project taking 2-3 > months. If I get it working I will be glad to share everything with the > group. > > Lester and Paul, I will test with the Costas loop as it may be the > easiest way to go. To test the Costas loop I am basically going to > duplicate the KD2BD design. I was reluctant about it in the beginning > but more reading seems to indicate it will be fine. My AWG allows me to > set the phase from 0 to 360 degrees independently on each channel so I > can use that for initial testing. That testing can be done prior to > getting the front end working. > > Richard, is your software posted somewhere? I assume it allows you to > specify the time and date you want it to send. That would come in handy > for writing the code to extract the timer/date data. I have seen > examples of the Arduino transmitting the old AM signal but don't recall > seeing one that sends the BPSK stream. > > I have a GPS time receiver with a 1.2 inch LED display that I built a > couple of years ago. I have it setup so the UART outputs the gprmc NEMA > string in case I want to look at it. I have a couple of spare GPS > modules lying around and several Arduino Uno modules. I have no problem > picking up the GPS satellites using only the patch antenna that is >>> built > into the modules. > > Ray, > AB7HE > > > ___ > time-nuts mailing list -- time-nuts@lists.febo.com > To unsubscribe, go to > http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com > and follow the instructions there. > ___ time-nuts mailing list -- time-nuts@lists.febo.com To unsubscribe, go to >>> http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com and follow the instructions there. >>>
Re: [time-nuts] WWVB PM Time Questions
Hi, > principles. The STM32L4 series which I often use has a pretty decent > ADC: fast (5 MSPS), with about 11 good bits in differential mode, and > "proper" hardware downsampling (called the DFSDM in the manual). If the Yes, the STM32 series also have built-in OPAmps, so one could hook up a loop or a ferrite directly to the uC without too much external components. Sampling at 160ks/s should suffice to get the phase and DCF77 is in reach as well. At this rate you have ~500-1000 processor ticks per sample which should be enough to do real time demodulation. Very nice project. Cheers Detlef Schücker "time-nuts" schrieb am 22.07.2020 18:44:08: > Von: "Mark Haun" > An: time-nuts@lists.febo.com > Datum: 22.07.2020 19:29 > Betreff: Re: [time-nuts] WWVB PM Time Questions > Gesendet von: "time-nuts" > > Hi Ray, > > A similar project has been on my to-do list for a couple of years now > (so don't hold your breath!). It should be possible to do this using > the built-in ADC for a really nice, minimalist demonstration of SDR > principles. The STM32L4 series which I often use has a pretty decent > ADC: fast (5 MSPS), with about 11 good bits in differential mode, and > "proper" hardware downsampling (called the DFSDM in the manual). If the > noise is white it should be easy to get > 16 good bits into the > demodulator. My plan was to clock the MCU from a cheap OCXO and make it > part of the carrier tracking loop, for a simple WWVB disciplined oscillator. > > I also have some questions about the NIST document, but in general, I > think you need to assume that your receiver is already locked, in which > case you know when to expect the transitions. Getting to this point > (acquisition) is another story, but there are all sorts of > correlation-based tricks that you can use, similar to GPS code > acquisition. As others have noted, carrier phase lock can be aided by > squaring. > > My problem is that the SDR stuff is all pretty straightforward for me, > but the analog electronics leading up to the ADC are black magic. I > have a preamp circuit modeled on John Magliacane's design > (http://www.arrl.org/files/file/QEX_Next_Issue/2015/Nov-Dec_2015/ > Magliacane.pdf), > fed from a ~ 60-turn tuned loop made from old ribbon cable, but I am > having problems getting anything "reasonable" looking on the scope. I > don't know if the problem is my suburban location or if you really can't > see the signal until you get the bandwidth << 1 kHz. (I am using > RC-tuned op-amp stages so the -3 dB response is necessarily a few kHz.) > Eventually I would like a small loopstick design with a crystal filter > like the off-the-shelf clocks use, but I haven't a clue how to get > there. Maybe this project would benefit from a collaboration :) > > Regards, > Mark > > On 21-Jul-20 4:37 PM, rcb...@atcelectronics.com wrote: > > I want to decode the WWVB time information using the BPSK information > > that is broadcast. I will use a STM32 to do the actual decoding of the > > bit stream. This is just an exercise in "can I do it?" as I know I can > > buy clocks for $30 that use the BPSK method. At one time you could buy > > an IC that output the data stream but I believe there is nothing > > currently available to do that. > > > > I read the document "Enhanced WWVB Broadcast Format" by John Lowe from > > NIST. One thing that is confusing to me is this paragraph: "Although the > > phase representing the information in each symbol is shown to be > > available before the amplitude in it transitions from VH to VL, it is > > recommended that receivers extract it only from the high amplitude > > portion of the symbol. This is not only because of the higher power > > there, allowing for more robust phase demodulation, but also because the > > low amplitude portion may be used in the future for additional (higher > > rate) phase modulation." > > > > How would you detect the phase had changed if you don't detect it when > > it changes at 100 msec after the carrier level drops? After the signal > > reverses phase wouldn't any reference you are using then be in lock with > > the current phase of the signal after it has changed? > > > > Or would you use a locally generated 60 kHz ultra stable signal as the > > phase reference? If so, how would you keep your local source locked to > > the 0 degree phase signal of WWVB and have it ignore the 180 degree > > phase shift? > > > > I have done a lot of searching and reading from various sources. But I > > haven't really found a good explanation of the hardware that would be > > used to detect the phase changes. Is there a block diagram somewhere > > that would illustrate the steps needed to detect the change during the > > high power portion of the WWVB signal? > > > > I am going to put together a 60 kHz amplifier using a couple of > > FET/transistors and a couple of high speed opamps. Then I can observe > > the signal on my scope. Using the output of the amp I can then try to > >
