Oscillator question

Most things expand when they warm up. A nothing-but-copper coil expands too. Copper has a TCE around 16 PPM. The little wound copper inductors that we use in oscillators generally run around +100 PPM. If the copper is wound on plastic, things can get really nasty.

Time does NOT have a temperature coefficient! Well, not a big one. We build delay generators with TC of a few ps/K. A nanosecond is a *long* time.

--

John Larkin         Highland Technology, Inc 
picosecond timing   laser drivers and controllers 

jlarkin att highlandtechnology dott com 
http://www.highlandtechnology.com

The B&W Miniductor design (air core with moulded plastic ribs spacing the turns out can be made with zero TC by choosing the length and diameter correctly. The smaller CTE of copper is balanced by the CTE of plastic times the weaker length dependence of the inductance.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs 
Principal Consultant 
ElectroOptical Innovations LLC 
Optics, Electro-optics, Photonics, Analog Electronics 

160 North State Road #203 
Briarcliff Manor NY 10510 

hobbs at electrooptical dot net 
http://electrooptical.net

How good can that get?

We've had nasty problems with cold flow of plastic bobbins in RF inductors. Baking for a few days helped to relieve the stresses. Longterm drift of inductance is a different problem from prompt TC.

Coaxial ceramic resonators have amazing Q and stability, but start at

600 MHz or so.

There was serious suggestion that an invar (or something) microwave cavity, at millikelvin temperatures, would be as stable a time reference as an atomic clock.

--

John Larkin         Highland Technology, Inc 
picosecond timing   laser drivers and controllers 

jlarkin att highlandtechnology dott com 
http://www.highlandtechnology.com

Even copper on FR4 is pretty stable. The CTE is something like

15ppm/K. Ceramic is more like 7ppm/K, which puts very modest temperature control or compensation within reach. The stable substrate makes the copper behave itself. Plain plastic resin is horrible- tempco and often hygroscopic (nylon** can swell a couple percent from moisture in the air- over hours or days depending on thickness), maybe fiber-loaded engineering resins are better but I'm not sure they use those for commercial inductors (the fibers chew even hardened molds up so there is an incentive not to use them).

We can at see least 0.01ppm in inductance (maybe a lot less depending on bandwidth and certain other things), but the setup for that is "inconvenient".

** famously used for hygrometers and humidistats in the form of a thin filament.

--
Best regards,  
Spehro Pefhany 
Amazon link for AoE 3rd Edition:            http://tinyurl.com/ntrpwu8 
Microchip link for 2015 Masters in Phoenix: http://tinyurl.com/l7g2k48

I generally see 100 PPM/K ballpark in surface-mounted inductors. We generally take out the gross stuff with NTC caps, and fine-tune with a varicap driven from a uP with a temperature sensor and a cal factor. All a huge nuisance. The FR4 doesn't help; terrible mechanical and capacitance TCs, so we try to minimize both.

We build LC oscillators that have to start instantly, in nanoseconds, so we can't just use an XO. After a few 10s of microseconds, we can phaselock them to a good XO, but we need them to be PPM stable fairly long-term.

--

John Larkin         Highland Technology, Inc 
picosecond timing   laser drivers and controllers 

jlarkin att highlandtechnology dott com 
http://www.highlandtechnology.com

It's probably limited by the nonlinearity of the CTE of the plastic, as well as humidity, but I'd expect you could get down into the single-digit PPMs at room temperature if you picked the right stuff.

When I worked for a microwave telecoms manufacturer back in the early '80s, we used fibreglass coil forms.

Maybe our dad's atomic clock, but probably not the newer ones.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs 
Principal Consultant 
ElectroOptical Innovations LLC 
Optics, Electro-optics, Photonics, Analog Electronics 

160 North State Road #203 
Briarcliff Manor NY 10510 

hobbs at electrooptical dot net 
http://electrooptical.net

[snip] [snip]

I know that coaxial resonators exist down to at least the low 300MHz range... I've designed a garage-door opener chip that used them. ...Jim Thompson

-- | James E.Thompson | mens | | Analog Innovations | et | | Analog/Mixed-Signal ASIC's and Discrete Systems | manus | | San Tan Valley, AZ 85142 Skype: skypeanalog | | | Voice:(480)460-2350 Fax: Available upon request | Brass Rat | | E-mail Icon at

| 1962 | I love to cook with wine. Sometimes I even put it in the food.

Ha. Maybe if you made the transistors out of invar (or something). :)

In the olden days before all radios were PLL tuned, people used to use ceramic formers to wind stable inductors on. I think it was usual to use a low-k ceramic so that the stray capacitance is also stable with temperature. There is probably some trick to getting the wire tension right so that the diameter of the turns is not much affected by the copper's CTE. There are a lot of them here, note some seem to be tuned with a metal (conductive) slug to reduce the inductance, rather than a ferrite slug to increase the inductance:

You could probably use quartz tubes also which have very low CTE. If you wanted you could use thick-film or evaporation to make the turns on a ceramic former, though patterning them would call for some ingenuity.

Some modern SMD inductors are available with ceramic formers, e.g.:

I have no idea if they have taken the right precautions to achieve low temperature coefficient of inductance.

Chris

Units below 500 MHz are rare and huge. It's easier for me to buy higher frequency ones and divide down.

They get good in the GHz region: small, cheap, insanely high Qs.

--

John Larkin         Highland Technology, Inc 
picosecond timing   precision measurement  

jlarkin att highlandtechnology dott com 
http://www.highlandtechnology.com