Indeed. Eevn with a perfect adapter, there's always an opportunity for jitter introduction towards the DAC itself. Jitter reduction should be done right at the DAC chip, otherwise it doesn't make a lot of sense.
jan didden
The DAC receiver will deal with any high frequency jitter from the connection. That leaves low frequency jitter from the clock in the transport. To reduce this you either need a better clock, or a crystal-controlled PLL which goes down to low frequencies - I assume this is what your jitter reducer is. I think you are solving a non-problem: the jitter reducer deals with LF jitter from the transport and the DAC deals with HF jitter from the cable. I don't see any reason to worry about the cable.
Jitter has to be minimised by the last stage. Two ways to do this: maintain low jitter throughout (not fully possible), or correct for jitter using a PLL. A decent crystal oscillator will have low jitter, although it could get worse at lower frequencies. A cable can introduce high frequency jitter, which is why S/PDIF receivers have a PLL to reduce HF jitter. They can't reduce LF jitter because at lower frequencies the PLL simply follows the signal - it has to do this in order to lock on to the incoming data stream. The PLL will of course introduce its own jitter, but provided it reduces the incoming jitter by a sufficient amount then you are still better off with a PLL than without it.
Or TNC connectors we use them for radio altimeters ( 4.3 GHZ ) 50 ohms are good to at least 11GHZ, 75 are only good to about 1 GHZ. http://www.spectrum-et.org/new_web2/adapters/pdf/BetweenSeries/BetweenSeries-TNC.pdf
Last edited:
The75 ohm version has the same dimensions as the 50 ohm connectors and some compromises where made to the insulator size in the 75 ohm version. In other words they would have to have a different inside diameter to outside diameter ratio to be optimized for 75 ohms. From the Wiki. Most TNC connectors are 50-ohm type even when used with coaxial cable of other impedances,[citation needed] but a 75-ohm series is also available, providing a good SWR to about 1 GHz.[2] These can be recognized by a reduced amount of dielectric in the mating ends. They are intermatable with standard types. They seem to have an air gap in addition to the insulation.
Last edited:
No, It's not the diameters of the cables but the ratio of the inside diameter to the outside diameter and the dielectric constant of the insulator. This is the problem, if you keep the physical ratio of the inside and outside conductors the same you must change the dielectric constant of the insulator to match the impedance to 75 ohms. What they did to compensate for the change needed in the dielectric constant was to add an air gap which reduced the high frequency response. 75 ohm tnc. Amazon.com: 75 Ohm TNC Female Crimp for RG59 Cable: Electronics
Last edited:
I'm not saying it is the cables, I was just using that as an example of geometric change which apparently preserves impedance yet still creates reflections. I think it is unlikely that an airgap will reduce HF response as air is a fairly good dielectric, unless it is just the discontinuity which is the problem.
Let's take a 50ohm M & F connector and a pair of 50ohm cable ends designed to be compatible with the connector.
They will have a bandwidth that works within a tolerance.
Now replace all those 50ohm components with 75ohm components, again with compatibility between connector and cable.
Why does the 75ohm connection version lose 10GHz of bandwidth?
They will have a bandwidth that works within a tolerance.
Now replace all those 50ohm components with 75ohm components, again with compatibility between connector and cable.
Why does the 75ohm connection version lose 10GHz of bandwidth?
Changing the dielectric changes the impedance. That was explained a little earlier.
Introducing a thinner dielectric and using air for the remainder also changes the impedance.
That introduction of the air to replace the missing dielectric does not help explain the missing 10GHz.
There's something else you are not telling me.
Introducing a thinner dielectric and using air for the remainder also changes the impedance.
That introduction of the air to replace the missing dielectric does not help explain the missing 10GHz.
There's something else you are not telling me.
Go back and read carefully. RMJ1 explained that all TNC connectors are 50 ohm, even when used with 75 ohm cable in a 75 ohm system. The impedance mismatch causes issues (poor SWR) above about 1 GHz. Below that point, the mismatch occurs over a small enough length of the transmission line that there is little effect. At higher frequencies, the physical length of the impedance-mismatched portion (the connectors) is bigger relative to the wavelength of the signal in the transmission line, so it has more impact.
Getting back to BNC, these are commonly available in 75 ohm, but even then they perform worse than the 50 ohm versions. This is due to an air gap in the dielectric which is required to provide the 75 ohm impedance from a connector whose dimensions were originally chosen to provide 50 ohm impedance. BNC is also generally not used above a few GHz anyway since the slots can start to radiate signal. The TNC is mostly identical but is threaded so it does not have this problem. Regardless, this is SPDIF and the driver circuits simply do not have the slew rate to provide GHz range signal components.
Any 75 ohm BNC which isn't defective will not degrade an spdif signal (the GHz discussions were academic, not practical- these are hundreds to thousands of times higher than spdif frequencies). If you stick to a name brand (e.g., Amphenol) purchased from a reputable distributor (e.g., Mouser), the signal transmission will far and away not be limited by the connector.
I agree with SY, BNC is an excellent interconnect and should work very well for this purpose. Make sure to stick with 75 ohm, and don't bother too much with teflon or such. The 75 ohm ones can be identified visually, so look into that (try here). Don't be too surprised if the BNC input or output jacks on your gear are not 75 ohm!
I also agree with sticking to name brand stuff, I bought some cheap BNC T's off eBay and they are garbage.
I also agree with sticking to name brand stuff, I bought some cheap BNC T's off eBay and they are garbage.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.