Frequency Standards

[RobinBirch]

Hi,
I am toying with the idea of building a frequency standard tied to the Radio 4 droitwich signal.

As I understand it, most of these devices consist of a 198 kHz receiver and then a set of phase locked loops using the receiver as the base clock. Each of the loops gives out the reference frequency you want, say 1 MHz, 10 MHz and so on.

I am assuming that the divide or multiply circuits that are used in the PLLs should be whole numbers so that you are matching the base frequency as closely as possible. Given that 198 kHz is a mathematically inconvenient number what is a reaonable set of multipliers?

Also, has anyone got a circuit for such a thing that I could have a look at?

Cheers

Robin

[Station X]

There's a design for one of these in "Test Equipment For The Radio Amateur". The design is for 200 KHz, but when Droitwich moved to 198 KHz an update was published in Radcom.

If you don't mind waiting until after Christmas I can scan the original article for you, but the update may be more difficult to locate.

[RobinBirch]

Graham,
That would be brilliant. Send it after Christmas by all means. If someone can find the update that would be beautiful.

Robin

[Station X]

I undoubtedly have the update. It's just a case of going through the Radcom indexes.

When did Droitwich go to 198 KHz? The update must post date that.

[RobinBirch]

Dunno, an allied question is why did they go to 198 when 200 is such a nice round number?

Robin

[Station X]

That was to comply with a new band plan where the channels were spaced at 9 KHz intervals.

[kalee20]

It's because 198kHz is a multiple of 9kHz, and 200kHz isn't! To cram as many stations as possible in the AM bands, they are as close together as possible in frequency terms, and 9kHz has been internationally agreed as the minimum spacing (this also imposes a limit of very little programme content above 4.5kHz to avoid spilling over into an adjacent channel).

It also makes things easier for synthesized tuning in receivers - you have a master clock at 9kHz and just use multiples of this (with an appropriate intermediate frequency also a multiple of 9kHz, e.g. 468kHz).

I'm also looking forward to the answer as to when it was changed - I do agree, 200kHz is nice... as is 1500m...

[mickjjo]

R4 changed from 200kHz to 198kHz in 1989....

http://en.wikipedia.org/wiki/Droitwi...itting_station

Regards, Mick.

[Ian - G4JQT]

I'm also looking forward to the answer as to when it was changed...

See http://mysite.wanadoo-members.co.uk/...ans_hist2.html

Ian

[Top Cap]

There are probably many ways of doing it, maybe you could use a CD40102 or CD40105 to divide the 198kHz down to 1 kHz then phase compare it to a 1kHz signal derived from your Oscillator/divider chain in order to lock the Oscillator.
I believe that modulation on Radio 4 can give phase shifting errors and the usual trick is to feed the signal through a 198kHz quartz crystal to strip off the modulation prior to dividing down. I often wonder though if you only have to do this if you are after real precision. Not quite sure either if a symmetrical signal is required for the phase detector so you may have to divide by 99 then use a divide by two to get 1kHz 50:50 square wave.
If I can, I will attach a divider I made using 74930 dual decade dividers. By using the divide by 5 halves first at each stage I ensured that every decade provided a 50:50 square wave. The LED at the end just flashed at 1Hz as a check that the divide chain was working ok. The 20MHz driver is a crystal in an oven and adjusted for longest zero beat with WWV, about the best way available here. I did have a chart somewhere that enabled you to work out accuracy by using an oscilloscope displaying WWV on one channel and your standard on the other channel. You just had to count the number of times the waveforms 'slipped' past each other, allied to the timebase setting and scope graticule. Les

[RobinBirch]

Hi there,
Curiously enough I have a couple of 198 kHz crystals in the depths of my spares bin. These were left over from an abandoned project at work to bild a clock system based on Droitwich using the 50 Hz FM signal that is buried in the carrier.

My intention is to make up a simple 198 kHz receiver and run the RF through one of the crystals to give a clean 198 kHz signal, square that up and then divide down to 2 kHz.

Then make a Phase locked loop system to create 10 MHz from the 2kHz.

My main inellectual issue with this is that I am creating a sine wave from a PLL driven by square waves that are created from a sine wave (the original carrier). Although the frequency should be fine how much jitter and other noise will I get in this and what should I do to ensure that it is a good clean signal?

Robin

[KeithsTV]

There was an off air frequency standard design using the Droitwich 198kHz in the November 2006 Practical Wireless. It uses a 10MHz crystal oscillator divided down to 200kHz and 2kHz. The 200kHz signal is mixed with the incoming 198kHz to generate a 2kHz IF. This 2kHz IF signal feeds a phase detector and is compared with the 2kHz output from the divider. The output from the phase detector controls the 10MHz oscillator.

One other thing to be aware of is that the Droitwich signal also carries data in the form of phase modulation of the carrier. This is used to control switching equipment for night storage heaters. The modulation is such that averaged over a second there is no net phase shift in the carrier. Any phase locked frequency standard therefore should have filter time constant greater than a second.

Keith

[RobinBirch]

That sounds interesting, how was the crystal oscillator controlled?

Or do you have a copy you could scan for me?

Hopefully with two or three examples (see station X's offer earlier) I can create what I want.

The whole idea is to provide a calibration standard for some test gear. I have a comms tester that I use for my own ends - works well and is very accurate. however it is long out of calibration and I don't fancy stumping up the cash to get it professionally done. One of the things that i use it for is checking the frequency of tranceivers in gliders and up to now this state of affairs has been fine. Unfortunately they (the powers that be) are changing the regulatory environment under which we look after such things and so I will have to be able to state the calibration of the kit that I use. My intention is to say that I have checked the comms tester against droitwich, the error is x and so the results measured are good to y.

Robin

[XTC]

Originally Posted by RobinBirch

I am toying with the idea of building a frequency standard tied to the Radio 4 droitwich signal.

Have you had a look on the second hand market for a commercial off-air standard? If you like building test equipment fine, but considering the time and possible frustration, you're often better off buying a ready-made unit.

Does it have to be based on the Droitwich 198KHz signal? GPS based time standards look interesting.

http://www.rt66.com/~shera/index_fs.htm

Pete.

[KeithsTV]

Robin

If you PM me I can send you a copy.

Incidently back in the 60s I used to live about 8 miles from Droitwich and with 100ft of wire as an aerial and a good earth I could get about 1v dc out of a crystal set tuned to 200kc/s (before Hertz came in).

I also visited the transmitter with the radio club from school. The LW transmitter comprised two 200kW transmitters in parallel ( A single output valve in each transmitter with 14kv HT at 14A anode current). The engineer showing us round said that when they wanted to do maintainance on one of the transmitters they just switched it off. I asked don't you lose some listeners? He replied "only a few fringe area listeners and we're not too bothered about them"

Keith

[RobinBirch]

Originally Posted by XTC

Have you had a look on the second hand market for a commercial off-air standard? If you like building test equipment fine, but considering the time and possible frustration, you're often better off buying a ready-made unit.

Does it have to be based on the Droitwich 198KHz signal? GPS based time standards look interesting.

http://www.rt66.com/~shera/index_fs.htm

Pete.

All fair comment. I must admit that I prefer building stuff as I tend to understand it better.

Robin

[XTC]

I mention buying s/h rather than building, because I went through a phase of building low distortion audio oscillators. They performed, but considering what you can find on ebay etc., for £50 or less - nicely packaged, beautiful attenuator, digital readout - it was fun, but not the most effective way to get a low distortion oscillator.


It's worth checking with the powers that be that calibrating your own equipment with an off-air standard is acceptable. I'm sure it would be accurate, but whether it's acceptable is another matter. It could just be that you can have your own caesium beam derived standard, but if it doesn't have a current calibration ticket from an approved lab, it doesn't count.

Pete.

[kalee20]

I'm a bit puzzled here! If you build a caesium beam frequency standard, why does it need a calibration certificate? The whole idea of SI units is that they are defined so that they can be re-created anywhere. And the second is defined in terms of the frequency emitted from an excited caesium atom.

So, a frequency standard derived from a caesium beam, is by definition correct.

[RobinBirch]

And that I think is the point.

If you have a primary standard, which something like a cesium standard is (I think) then you don't need a calibration ticket. Secondary standards which are derived need calibration tickets.

I think that if I can show that I am deriving my reference from a maintained standard (Droitwich in this case) and document that fact as a "before using the test gear it is checked against......." then I am ok.

An example is the manometer that we use to check airspeed indicators. This is a water column, mercury for high speed ASIs, and the effective speed is given by the air pressure that supports the water to a given height. Gravity is always switched on and to all intents and purposes doesn't change. The scale marking distances don't change. It has water in it and water is water for the purposes of the test. This is a primary standard as it is a direct representation of the physical quantity that you are measuring.

An example of the sort that requires a ticket is the altimeter that is used to check other altimeters. This will require regular recalibrating as the mechanism can wear or start building up a bit of friction. This is a secondary standard and requires a regular ticket. Fortunately it is one that is fairly simple to do.

Robin

[XTC]

Originally Posted by kalee20

I'm a bit puzzled here! If you build a caesium beam frequency standard, why does it need a calibration certificate?

I'd like to shake the hand of the private individual who builds a caesium beam standard from scratch and offer him a bottle of Champagne. The offer is limited to the first who presents himself in the UK, but the Champagne would be a good one (IMHO).

Caesium beam standards can either be primary, or secondary like the el-cheapo commercial units which sell for 20 grand.

The point here is that Robin is dealing with officialdom, which could be reasonable and assume that as a qualified engineer who has been maintaining glider tranceivers for years, he knows what he's doing and using an off-air standard for calibrating his equipment is perfectly good. OTOH, it could prescribe that all frequency standards used for such work shall bear a valid test certificate issued by a lab qualified to standard XYZ. That would, ridiculously, include a caesium beam standard, unless he had the time and patience to negotiate an exception.

Given an off-air standard was acceptable, even that might have to be checked periodically by a test house against their, err, off air-standard. Having them verify that a Quartzlock or Advance was behaving itself would be one thing. Having them verify a homebrew job, could involve negotiated prices.

The powers that be might allow that as a qualified engineer, experienced in the field, he could calibrate other test gear against his accepted standard. They might insist that it all had calibration certificates and unbroken, valid seals.

All I'm saying is that in his position, I'd be very careful to investigate the requirements of the powers that be and get it in writing that what I was doing was acceptable.

Pete.