Toroidal core VS air coil ppm stability in RF
The best core I have found, the T50-6 has a Temperature Stability (ppm /°C) = 35
I was wondering if an air coil can exhibit greater temperature stability.
also, what if the air coil is wound on a pvc form or polyethylene plumber's tube? will it excibit greater ppm stability than the T50-6?
What I mean is that, in terms of temperature stability which is better?
An air coil as such can be expected to have a better stability, but an air coil on a plastic coil former probably hasn't due to it's rather high thermal expansion coefficient. Good air coils are e.g. wound on a ceramic bobbin.
Dear neazoi
Hi
The air core coils have below problems : 1- very low flux permeability . 2- the magnetic flux will go to other sides ( more flux loss ) . 3- if the distance between turns change , the inductance will change . and .....
So , consider that you want 500uH coil , do you know , that what will be the number of turns ?! would you like to calculate it ?! it will be awful . so at high frequencies , we have to use air core . do you know why ?
Ok . i'll tell , you ! because we will need the inductors with very low inductances ( about 1 or more nH !) ( or probably , lower !) . and the other problem , is , that at high frequencies , the ferro and ferrites will going very hot ! it is the basic of induction melt systems . so , if you can , it is better , to prevent from air core coils and transformers .
Good luck
Goldsmith
So say, a silver plated teflon insulated wire, wound on a ceramic former should be expected to have a greater temperature stability than the 35ppm/C of the T50-6? I consider the T50-6 as a reference core, I may be wrong...
---------- Post added at 01:49 ---------- Previous post was at 01:43 ----------
So, for HF work do you think the T50-6 wound coils are a better solution than the air coils, or coils wound on a ceramic former? 35ppm/C is not bad at all to say, I just wondered if it will worth to try an air coil. Inductances in my oscillator are as low as 9uh.
* Core coils are avoided when low loss and high Q is needed though, like in crystal radio, but in my application some loss is not of a problem.
As FvM said , the ceramics are better .
To give a trivial answer, if you look to the micrometals catalog, you find 0 ppm specified for a non-magnetic phenolic resin "core". At first sight, it's obviously better than 35 ppm.
Unfortunately, the specification is only considering the μr temperature coefficient. I guess, you'll get much higher temperature drifts by not fixing the windings thoroughly. Also the thermal expansion of the coil former or even the wire insulation plays a role. Furthermore, the oscillator will be affected by other temperature drift effects. At the end of the day, you have a number of systematical t.c. influences, that can partly compensate when utilized appropriately, accidental effects, e.g. loose windings, and aging.
By using an air core, you can avoid one of many existing temperature drift causes. If you are getting others in return depends on your design.
Yes I am aware of the general issues for a stable oscillator. In fact, one should not only consider the stability of the resonant circuits (crystal, coil, capacitor) but also the stability of other components. NP0 capacitors, low ppm resistors etc. In fact the thermal heating of the oscillator transistor has it's own effect as this may change the internal capacitance.
If the oscillator has to be accurate to the Hz all these issues have to be thought. Temperature compensation may be needed using opposite coefficient capacitors in some places or an oven that encloses all the oscillator circuit. The temperature of the oven depended on the coefficient curve of the crystal used, for maximum accuracy.
And not to mention the operator effect, as you touch or come close to the tuning elements, like var cap, the oscillator may drift, unless you use synthetic shafts.
Also if you use more than one crystals in an analogue synthesis mixing technique each crystal has to have its own oven for maximum stability.
So there are many issues involved in stability as you said.
Of course one could use a dds or pll but there are other issues there, phase noise, jitter etc. A well designed crystal oscillator is by far a superior single HF frequency source than any other method.
I would tend to agree; Nothing beats a simple Hi-Q Quartz crystal resonator to generate low phase and low jitter signals.
There are other options nowadays though, and this is probably overkill for many HF applications requiring low phase noise and low jitter, but this was the solution we utilized in our WiMAX product as the master reference as the "low-phase" noise source for our S-band and C-band LO synthesizers, in fact, it can and was used to 'clean up' GPS-derived 10 MHz source as well as provide "hold-over" when the GPS module reported loss of signal GPS L1 C/A signal:
AD9548 | Quad/Octal Input Network Clock Generator/Synchronizer | Clock Generation and Distribution | Clock and Timing | Analog Devices
Several mentioned applications:
- SONET/SDH clocks up to OC-192, including FEC
- Stratum 2 holdover, jitter cleanup, and phase transient control
- Stratum 3E and Stratum 3 reference clocks
"The digital PLL allows for reduction of input time jitter or phase noise associated with the external references. The AD9548 continuously generates a clean (low jitter), valid output clock even when all references have failed by means of a digitally controlled loop and holdover circuitry."
I had not been aware of such solutions previous to our use of this IC, as stated, it not only provides a 'reference' in lieu of losing an input, but, provides 'clean up' of the phase noise for any reference source used.
Jim