Experimenting with a 74ls86 XOR doubler for 5 to 10 Mhz. Typically this would use a 90 degree phase shift to the other gate. The gate acting as a mixer to produce 10 Mhz. The reason to experiment is that I have noticed most of the doubler discussions take a 5 Mhz square wave filter it to a sine wave, feed it to a multiplier scheme and then filter the output. The 7486 method eliminates one of those processes. I have accurate delay lines I can adjust in 2 ns increments (Allen Aviation lump LC). The output is a semi asymmetrical square wave due to some gate timing I need to deal with if possible. Setting the delay taps to 90 degrees produces a 10 MHz output with 5 and 15 Mhz 8-10 db down. Lots of other higher frequency outputs. At this point I have no filtering on the output of the 7486. Purposely mis-adjusting the taps sets either the 5 Mhz or 15 Mhz level higher. Other noise and such are many DB down 50 plus. Why is this a bad method as compared to our typical time-nuts discussions? Regards Paul WB8TSL

Hi Paul, It isn't that it is bad, it is just that 5 and 15MHz products at 8 to 10dB down isn't very encouraging. To make decent use of this technique, I believe that you would have to install 20 to 30dB of 5MHz rejection, and a 10MHz low pass filter in the output circuitry.... And, that is in addition to making a simple very stable 90 degree phase shifter. The 5MHz rejection filter is necessary to prevent phase anomalies from appearing due to the beating of the doubled 5MHz fundamental with the XOR gate created 10MHz signal. Any time you add filters, you are adding temperature dependent phase shifters to your circuit. -Chuck Harris paul swed wrote: > Experimenting with a 74ls86 XOR doubler for 5 to 10 Mhz. Typically this > would use a 90 degree phase shift to the other gate. The gate acting as a > mixer to produce 10 Mhz. > The reason to experiment is that I have noticed most of the doubler > discussions take a 5 Mhz square wave filter it to a sine wave, feed it to a > multiplier scheme and then filter the output. The 7486 method eliminates > one of those processes. > I have accurate delay lines I can adjust in 2 ns increments (Allen Aviation > lump LC). > The output is a semi asymmetrical square wave due to some gate timing I > need to deal with if possible. > Setting the delay taps to 90 degrees produces a 10 MHz output with 5 and 15 > Mhz 8-10 db down. Lots of other higher frequency outputs. At this point I > have no filtering on the output of the 7486. > Purposely mis-adjusting the taps sets either the 5 Mhz or 15 Mhz level > higher. > > Other noise and such are many DB down 50 plus. > Why is this a bad method as compared to our typical time-nuts discussions? > Regards > Paul > WB8TSL > _______________________________________________ > time-nuts mailing list -- time-nuts@febo.com > To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts > and follow the instructions there. >

Chuck Thanks and indeed I do need filters that I have not experimented with and in that respect this would be more like some of the circuits discussed here on time-nuts. I am using nice controlled delay lines and at $66 each thats pretty un-attractive. But hey when you get them for 50 cents at a hamfest you can get crazy. Kind of kicking my self as I think the guy had more. Wasn't sure what to do with them. So the next step would be a 5 Mhz notch. Regards Paul WB8TSL On Fri, Feb 20, 2015 at 11:28 AM, Chuck Harris <cfharris@erols.com> wrote: > Hi Paul, > > It isn't that it is bad, it is just that 5 and 15MHz products at > 8 to 10dB down isn't very encouraging. > > To make decent use of this technique, I believe that you would have > to install 20 to 30dB of 5MHz rejection, and a 10MHz low pass filter > in the output circuitry.... > > And, that is in addition to making a simple very stable 90 degree > phase shifter. > > The 5MHz rejection filter is necessary to prevent phase anomalies > from appearing due to the beating of the doubled 5MHz fundamental > with the XOR gate created 10MHz signal. > > Any time you add filters, you are adding temperature dependent phase > shifters to your circuit. > > -Chuck Harris > > paul swed wrote: > >> Experimenting with a 74ls86 XOR doubler for 5 to 10 Mhz. Typically this >> would use a 90 degree phase shift to the other gate. The gate acting as a >> mixer to produce 10 Mhz. >> The reason to experiment is that I have noticed most of the doubler >> discussions take a 5 Mhz square wave filter it to a sine wave, feed it to >> a >> multiplier scheme and then filter the output. The 7486 method eliminates >> one of those processes. >> I have accurate delay lines I can adjust in 2 ns increments (Allen >> Aviation >> lump LC). >> The output is a semi asymmetrical square wave due to some gate timing I >> need to deal with if possible. >> Setting the delay taps to 90 degrees produces a 10 MHz output with 5 and >> 15 >> Mhz 8-10 db down. Lots of other higher frequency outputs. At this point I >> have no filtering on the output of the 7486. >> Purposely mis-adjusting the taps sets either the 5 Mhz or 15 Mhz level >> higher. >> >> Other noise and such are many DB down 50 plus. >> Why is this a bad method as compared to our typical time-nuts discussions? >> Regards >> Paul >> WB8TSL >> _______________________________________________ >> time-nuts mailing list -- time-nuts@febo.com >> To unsubscribe, go to https://www.febo.com/cgi-bin/ >> mailman/listinfo/time-nuts >> and follow the instructions there. >> >> _______________________________________________ > time-nuts mailing list -- time-nuts@febo.com > To unsubscribe, go to https://www.febo.com/cgi-bin/ > mailman/listinfo/time-nuts > and follow the instructions there. >

Chuck wrote: >To make decent use of this technique, I believe that you would have >to install 20 to 30dB of 5MHz rejection, and a 10MHz low pass filter >in the output circuitry.... * * * >The 5MHz rejection filter is necessary to prevent phase anomalies >from appearing due to the beating of the doubled 5MHz fundamental >with the XOR gate created 10MHz signal. If you start with 5MHz down 10dB in the output, and apply 30dB of filtering, you end up with 5MHz down 40dB. That is nowhere near enough to avoid anomalies in stability plots. You need the 5MHz component in the output closer to -80dBc than to -40dBc to get the anomalies down toward the noise floor. That is a massive amount of filtering very close to the desired 10MHz output, and is why it is important to start with 5MHz below -40dBc straight from the doubler (no filtering). Best regards, Charles