Hi, I recently bought an Austron 2010B, a disciplined OCXO standard with adjustable disciplining parameters, for use as a clean-up oscillator and a decent fallback to my Ball MRT-H, a rubidium standard. I figured that the quartz standard's 1 Mhz and 5 MHz outputs are fine---the former is exactly what my Truetime 814-149, a time code generator, needs for eventually providing PPS output to my SPARC-based NTP server, and the latter can be doubled, amplified, and distributed, all with a single distribution amplifier unit, for use as a timebase for transmitters and lab instruments. Only later did I realize that for some bizarre reason, all of the outputs are square wave, not sinusoidal! Great. Any ideas on the likely reason that the unit was engineered with only square wave outputs? Obviously this will render division to PPS trivial, but all of the applicable equipment that I've encountered use a sinusoidal reference, not a square one, so it doesn't seem prudent to exclude sine. And naturally, what is the most prudent course of action in this situation? I'd rather use something prebuilt than building my own converter, but all the distribution amplifiers I've looked at lack such a conversion feature, and I'm unsure whether plain square to sine converters are suitable for such time/frequency metrology applications. Thanks in advance, Ruslan

Hi > On Aug 6, 2016, at 3:00 AM, Ruslan Nabioullin <rnabioullin@gmail.com> wrote: > > Hi, I recently bought an Austron 2010B, a disciplined OCXO standard with adjustable disciplining parameters, for use as a clean-up oscillator and a decent fallback to my Ball MRT-H, a rubidium standard. I figured that the quartz standard's 1 Mhz and 5 MHz outputs are fine---the former is exactly what my Truetime 814-149, a time code generator, needs for eventually providing PPS output to my SPARC-based NTP server, and the latter can be doubled, amplified, and distributed, all with a single distribution amplifier unit, for use as a timebase for transmitters and lab instruments. Only later did I realize that for some bizarre reason, all of the outputs are square wave, not sinusoidal! Great. > > Any ideas on the likely reason that the unit was engineered with only square wave outputs? Obviously some system somewhere needed a bunch of square waves. Likely some sort of digital system. Telecom gear comes to mind. Austron would happily sell you one with just about any sort of outputs you wished to pay for. > Obviously this will render division to PPS trivial, but all of the applicable equipment that I've encountered use a sinusoidal reference, not a square one, so it doesn't seem prudent to exclude sine. And naturally, what is the most prudent course of action in this situation? I'd rather use something prebuilt than building my own converter, but all the distribution amplifiers I've looked at lack such a conversion feature, and I'm unsure whether plain square to sine converters are suitable for such time/frequency metrology applications. Provided the phase noise is ok, a square wave works fine. The bigger question is: Do the square wave outputs drive 50 ohms ok? Bob > > Thanks in advance, > Ruslan > _______________________________________________ > 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.

Ruslan wrote: > Only later did I realize that for some bizarre reason, > all of the outputs are square wave, not sinusoidal! > > I'd rather use something prebuilt than building my > own converter A simple Tee network lowpass filter will do the job -- see the attached schematic (NOTE: this circuit is designed for 10MHz -- you would need to scale it for 5MHz and 1MHz). If the input is from 5v CMOS logic (0v low, 5v high) that will put out at least +/- 30mA and you use a 56 ohm input resistor, the circuit will deliver a sine wave output of ~1v RMS (= +13dBm) into a 50 ohm load, which is a common output level for frequency references. Mini-Circuits sells a 5MHz LP filter with BNC connectors (BLP-5+), but it does not include the input resistor or the DC blocking capacitor shown in the attached schematic -- so you would need to add those yourself (again, I'm assuming that the Austron output is a DC-coupled, 0-to-5v square wave). Mini-Circuits does not list a 1MHz filter, and their 1.9MHz filter is not suitable because it wouldn't provide sufficient attenuation of the 3rd harmonic (3MHz). You would put the filter between the Austron output and the 50 ohm input of a distribution amplifier. Best regards, Charles