10 MHz GPS disciplined oscillator + signal distribution

[TonyAlbus]

Get well soon!

[Zenith]

Nice!, i have good experiance with those
if you like to upgrade to more sattelite standards..
here i made a nice video how to.
your GPSDO can do a lot more then advertised.

https://www.youtube.com/watch?v=eAdJSErqZOk

Good to see you here.

I enjoyed and got a lot from from several of your Youtube videos, particularly the one on the cheap ex cell tower OCXOs.

I have a cheap TM4313 GPSDO, which I assume for my purposes is perfect - good to at least 1 part in 10^9 - and I'm amazed how good those Aliexpress OCXOs are and how quickly they settle down - about five minutes.

On the other hand I have doubts. You may be about to lead me down an expensive rabbit hole leading to having a caesium beam clock in the garage.

[TonyAlbus]

Thanks, great is was helpful, nice to hear.

You will find the 6 port PLL also interesting then.
i use it to generate a 50 MHz reference for some of the Chinesium generators for were i build external ref mods.
10 MHz in , allmost anything out, up to 200 MHz .. all synced to GPSdo

https://www.youtube.com/watch?v=64HS88gCC8I
https://www.youtube.com/watch?v=Iocfuz9ruNU

[Zenith]

It's of interest, but I won't be rushing off my order for one.

My interest in this comes from having three Racal frequency counters and a Marconi synthesised signal generator, all with 30 to 40 year old OCXOs supposed to be good to a few parts in 10^7. The sig gen was low and at the end of its adjustment. Opening up the OCXO and tweaking the internal adjustment (a preset cap) made it possible to adjust it either side of its target frequency. The oldest Racal counter was in great disagreement with the others. They all take hours to stabilise from power on and stabilise at a different point every time. I think that OCXOs have a limited life and they get old and quirky. I had always assumed that replacing them would be very expensive

None of these units had been calibrated for decades and I'm sure sending one to a calibration lab would cost more than I paid for it, then there's the question of how long the calibration would hold. So I bought the GPSDO, which I believe is good to 1 part in 10^10 at least after it's been on for a few hours and set them up against that. What it told me was how quirky these things were, in particular the sig gen and the oldest Racal. So I ordered some of the Aliexpress OCXOs,which is where your video was so useful. Just setting them up on the bench, they all work, they stabilise in a few minutes and don't wander from 10MHz much.

I made up a crude board to replace the OCXO in the oldest Racal. It's sat there for weeks within a few ppb of 10MHz, according to the GPSDO, and it takes a few minutes to stabilise. Now I'm thinking in terms of a better board with a voltage reference to provide the control voltage rather than take it from 5V supply, and making the multi turn adjustment pot less finicky. The Racal OCXOs are 5MHz 1V ptp sine wave and the board delivers a 5MHz TTL square wave. A simple filter should do.

Then there's the sig gen which supplies its OCXO with 5V and 12V and produces a sine wave. That will need some investigation of the existing OCXO.

I don't see any need to feed these units from external references. If they powered up and agreed to within 1 part in 10^8 in a few minutes, I'd think they were behaving perfectly. I also have modern equipment including a DDS function generator and a spectrum analyser, with I think high grade TXCOs. It's nice to know that they are behaving properly and to be able to tweak them back, should the need arise.

I can see the appeal of volt nuttery and time nuttery and how standards are derived from mass, length and time. I have a idle ambition to own a Weston Standard Cell. Mainly I am interested in being confident that the equipment I have performs at or beyond its specification, without sending things to calibration labs.

[tggzzz]

I have a idle ambition to own a Weston Standard Cell.

The saturated ones are usually buggered. I do have one that isn't, as the calibrator inside a DVM.

The unsaturated ones are OK provided they've not been upended/shocked. Good luck with shipping them legally and safely! They make good temperature sensors over a limited range, and a noiseless stimulus when measuring a DVM's noise. I have several.

[Zenith]

We spoke about WSCs and volt nuttery briefly at FRARS last August.

WSCs were a thing I read about in a physics book when I was ten and I wanted one because they looked cool. Looking back, it's a good thing I didn't get my hands on one. The school had a couple, along with a Fortin barometer, a PYE mirror scale µammeter and an oscilloscope. The scope was a Telequipment S53. We were allowed to play a bit with the scope and take a reading from the barometer, but the µammeter and the WSCs were distinctly out of bounds. I believe the properties of a WSC are permanently changed by having it deliver more than 10µA. This is not a problem with a DVM where even the el cheapos have a resistance of 1MR, but in the days when the AVO ruled supreme (20K/V) attaching your voltmeter was not a thing to be done without thought. More advanced techniques were called for. Valve voltmeters were around

Later on I came by a physics book, 1930s university level, which devoted a chapter on WSCs, their care and keeping, the difference between saturated and unsaturated and how the voltage changed with temperature. They were THE practical voltage standard and I'm sure AVO had a load.

The last one I came across was at L. F. Hanney, (an electronics/junk shop in Bath) when it was closing, over 15 years back. I feel sure you paid it a visit. He had one but said it was no good because his grandson had shaken it up and wrecked it.

The intriguing thing about WSCs is that you could in principle make one yourself. A bit of mercury, a bit of cadmium, cadmium sulphate, all analytical grade, and some distilled water, plus a bit a bit of not particularly hard glass work and a few cm of suitable wire, say platinum. You'd be well away. As an individual it might present difficulties but for a university lab it should be done without breaking step.

[tggzzz]

We spoke about WSCs and volt nuttery briefly at FRARS last August.

We did indeed

Shame the Harwell ARS website only contains the ambiguous statement "the rally will not be taking place in Feb 23".

WSCs were a thing I read about in a physics book when I was ten and I wanted one because they looked cool. Looking back, it's a good thing I didn't get my hands on one. The school had a couple, along with a Fortin barometer, a PYE mirror scale µammeter and an oscilloscope. The scope was a Telequipment S53. We were allowed to play a bit with the scope and take a reading from the barometer, but the µammeter and the WSCs were distinctly out of bounds.

We used the WSC and mirror scale µammeter for specific experiments.

The mirror scale µammeter's main feature was that it was very slow acting, and so could be used to measure (relative) charge. That's the start of a fun discussion with youngsters

The (presumbably saturated) WSC was used only in the way that it is used in professional "DVMs" of the time: with a NiFe cell, a 1m rule, and 1m of resistance wire. We were monitored to ensure the WSC was only connected when the balance point had been roughly located, thus minimising the amount and time of the discharge.

I believe the properties of a WSC are permanently changed by having it deliver more than 10µA. This is not a problem with a DVM where even the el cheapos have a resistance of 1MR, but in the days when the AVO ruled supreme (20K/V) attaching your voltmeter was not a thing to be done without thought. More advanced techniques were called for. Valve voltmeters were around

Later on I came by a physics book, 1930s university level, which devoted a chapter on WSCs, their care and keeping, the difference between saturated and unsaturated and how the voltage changed with temperature. They were THE practical voltage standard and I'm sure AVO had a load.

The last one I came across was at L. F. Hanney, (an electronics/junk shop in Bath) when it was closing, over 15 years back. I feel sure you paid it a visit. He had one but said it was no good because his grandson had shaken it up and wrecked it.

I wouldn't like to quote a specific figure, but an Avo was used to get the balance point about right. Conrad Hoffman discusses it.

IIRC, WSCs were only superseded as the definition in 1990, by the Josephson Junction.

I never visited Hannay

OTOH my Land Rover's fuel gauge (scale: Roentgens/hour) came from this famous shop in Cambridge:

The intriguing thing about WSCs is that you could in principle make one yourself. A bit of mercury, a bit of cadmium, cadmium sulphate, all analytical grade, and some distilled water, plus a bit a bit of not particularly hard glass work and a few cm of suitable wire, say platinum. You'd be well away. As an individual it might present difficulties but for a university lab it should be done without breaking step.

Most "early" technology is that way. Classic example is the things POWs in WW2 used to create radios to pickup Allied transmissions.

As for "practical", getting the chemicals is the major hitch. When I think what we had in our chemistry sets as a kid...

Now, what should I do with my lump of Na, jar of Mg powder, jar of potassium permanganate (and of sodium chlorate?) All from the same source as my WSCs.

[Zenith]

Shame the Harwell ARS website only contains the ambiguous statement "the rally will not be taking place in Feb 23".

I thought it was off for this year when it hadn't appeared on the lists by early January. As I recall it's around the second Sunday in February. Realistically it takes a month for it to be noticed and stallholders to book and arrange their plans, so I assumed it wasn't going to happen.

We used the WSC and mirror scale µammeter for specific experiments.

The mirror scale µammeter's main feature was that it was very slow acting, and so could be used to measure (relative) charge. That's the start of a fun discussion with youngsters

The (presumbably saturated) WSC was used only in the way that it is used in professional "DVMs" of the time: with a NiFe cell, a 1m rule, and 1m of resistance wire. We were monitored to ensure the WSC was only connected when the balance point had been roughly located, thus minimising the amount and time of the discharge.

Our mirror scale µammeter was a PYE. I saw a physics master use it once and optical levers were covered in text books. He sent it off the end of the scale by pushing a magnet into a coil of wire too quickly. I saw one at a rally a time back, but it looked sad and tatty. It's not something I'd particularly want, even if slightly used and in its box.

I used a meter bridge to find the value of the mystery resistor. I recall the meter was a 100µA-0-100µA number desensitised to home in with a 33K resistor. In the 60s any halfway decent used panel meter would cost a quid - this is when a Mars bar cost 6d (2.5p).

I wouldn't like to quote a specific figure, but an Avo was used to get the balance point about right. Conrad Hoffman discusses it.

IIRC, WSCs were only superseded as the definition in 1990, by the Josephson Junction.

They had a long reign. I'm pretty sure they were used in some form in early bench DMMs.

I never visited Hannay

OTOH my Land Rover's fuel gauge (scale: Roentgens/hour) came from this famous shop in Cambridge:

I only went to L F Hanney three times. The prices were a bit steep. Parking was hard round there and you didn't know what he had in stock and had to ask for everything, which I found annoying. I don't think you missed very much. I went to Centre Electronics in Birmingham on several occasions and bought a few items from there.

It's sad there are few if any of those places left.

If I'm ever in Cambridge I shall look that place up.

The intriguing thing about WSCs is that you could in principle make one yourself. A bit of mercury, a bit of cadmium, cadmium sulphate, all analytical grade, and some distilled water, plus a bit a bit of not particularly hard glass work and a few cm of suitable wire, say platinum. You'd be well away. As an individual it might present difficulties but for a university lab it should be done without breaking step.

Most "early" technology is that way. Classic example is the things POWs in WW2 used to create radios to pickup Allied transmissions.

As for "practical", getting the chemicals is the major hitch. When I think what we had in our chemistry sets as a kid...

Now, what should I do with my lump of Na, jar of Mg powder, jar of potassium permanganate (and of sodium chlorate?) All from the same source as my WSCs.

Na, good as a conversation starter at pool parties if they start to stagnate.

If you fish, potassium permanganate is good for catching worms. You water a couple of square meters of the lawn with a solution and the worms emerge. There are other uses.

Mg powder can be mixed with some oxide; copper, iron, chromium etc and you have thermite, alternatively you can mix it with the sodium chlorate and you have flashlight powder. Both of these afford hours of innocent amusement.

I believe you can buy metallic cadmium but with secure delivery the cost is eye watering. You can probably buy the other stuff too. All rather expensive and a lot of trouble. I believe Weston had to do a lot of experimenting to be able to produce consistent cells.

[Specmaster]

OTOH my Land Rover's fuel gauge (scale: Roentgens/hour) came from this famous shop in Cambridge:

I think I've been in that shop before myself, was it in Mills Road, along with the other shop I used to frequent, Cambridge Recycle I think it was called, also in Mills Road, towards the end of the retail section on the right.

[Zenith]

The intriguing thing about WSCs is that you could in principle make one yourself. A bit of mercury, a bit of cadmium, cadmium sulphate, all analytical grade, and some distilled water, plus a bit a bit of not particularly hard glass work and a few cm of suitable wire, say platinum. You'd be well away. As an individual it might present difficulties but for a university lab it should be done without breaking step.

Most "early" technology is that way. Classic example is the things POWs in WW2 used to create radios to pickup Allied transmissions.

As for "practical", getting the chemicals is the major hitch. When I think what we had in our chemistry sets as a kid...

I recall my chemistry set was a bit disappointing.

POWs making radio receivers was a bit different. It was the highly ingenious application of technology they knew very well, in very difficult circumstances.

Looking at the Conrad Hoffman WWW pages
https://www.conradhoffman.com/stdcell.htm

maybe it isn't something you could tinker together with a bit of cadmium, a bit of mercury and all the rest, even though Weston started off that way. You'd produce something but likely not satisfactory. The purity of the materials was essential. This involved distilling mercury and distilling sulphuric acid. Did they distill the cadmium? Then having produced mercurous sulphate and cadmium sulphate they had to be purified by successive re-crystallizations. Distilling the sulphuric acid seems a very heavy number, as for the rest, it's off the scale.

[tggzzz]

OTOH my Land Rover's fuel gauge (scale: Roentgens/hour) came from this famous shop in Cambridge:

I think I've been in that shop before myself, was it in Mills Road, along with the other shop I used to frequent, Cambridge Recycle I think it was called, also in Mills Road, towards the end of the retail section on the right.

That's the one.

I heard it closed maybe a decade ago, and Google streetview shows it boarded up.

[tggzzz]

IIRC, WSCs were only superseded as the definition in 1990, by the Josephson Junction.

They had a long reign. I'm pretty sure they were used in some form in early bench DMMs.

They were. In my Solartron 1420 it is used, via a trimpot, to calibrate the internal voltage reference - a neon bulb.

I presume they are in some Fluke "DVMs" from the 1960s, in particular those that have knobs for internal KVDs.

It's sad there are few if any of those places left.

If I'm ever in Cambridge I shall look that place up.

They were of their time.

Gees is long gone, but fondly remembered.

Mg powder can be mixed with some oxide; copper, iron, chromium etc and you have thermite, alternatively you can mix it with the sodium chlorate and you have flashlight powder. Both of these afford hours of innocent amusement.

I believe you can buy metallic cadmium but with secure delivery the cost is eye watering. You can probably buy the other stuff too. All rather expensive and a lot of trouble. I believe Weston had to do a lot of experimenting to be able to produce consistent cells.

I must get around to playing with thermite. My memory, probably faulty, was that iron and aluminium oxides were sufficient - provided you could get them up to a suitable temperature. I guess the magnesium does that.

[Zenith]

Usually aluminium powder but magnesium powder works well, and some oxide of a metal which has less affinity for oxygen than aluminium. Iron is fine. The way the demonstration was done in school labs was to start the reaction with a bit of magnesium ribbon. When I saw it done it was with aluminium powder and chromium oxide and it left a nugget of chromium in the bottom of the crucible. There are some oxides you absolutely wouldn't use for obvious reasons, such as lead, cadmium, mercury etc. My own experiment involved black copper oxide from a chemistry set and aluminium filings, half filling a spent .303 case, which was put on the coal fire. After a couple of minutes there was a bright light. The top half of the case was neatly melted off and could be fished out with a poker.

It has long had peaceful uses in welding tram lines and lately railway lines, where the purpose is to create molten steel of a given composition, so iron oxide, manganese oxide and maybe chromium oxide in the right proportions. I don't know how they get the amount of carbon right.

[mnementh]

mnem
*undisciplined dwagon*

[mansaxel]

Of course these antennas have a 3/4" NP (don't know if taper or straight) thread combined with a 1" marine antenna mounting thread. I have a source (Völkel at least, but perhaps also Walter) for 3/4" NP thread dies, and European 3/4" steel pipe is about 0,4mm off in diameter from US 3/4 steel pipe, meaning I probably can use European pipe and cut NP threads, but there is a delay to that procedure and I have a nice REMS EVA BSPT set... Has anybody tried BSPT in NPT and how bad was it? Pressure tightness is not necessary here.

11 TPI vs 11.5 TPI. They should wind together far enough to provide a reasonable mechanical join.
Can you get good spanners/wrenches onto both ?

It works very well as is! Now that the antennas have arrived, I finally remembered to bring a G3/4" (ie straight thread, not taper) nipple to work, and it slides on perfectly, if a bit loose for pressure tightness. Maybe pipe compound and flax fibre would help. I then have reason to believe that taper thread will provide a stout fit, especially if mounted using flax and compound.

I will keep you posted; for this is time nuttery on an enterprise scale.

[tautech]

Of course these antennas have a 3/4" NP (don't know if taper or straight) thread combined with a 1" marine antenna mounting thread. I have a source (Völkel at least, but perhaps also Walter) for 3/4" NP thread dies, and European 3/4" steel pipe is about 0,4mm off in diameter from US 3/4 steel pipe, meaning I probably can use European pipe and cut NP threads, but there is a delay to that procedure and I have a nice REMS EVA BSPT set... Has anybody tried BSPT in NPT and how bad was it? Pressure tightness is not necessary here.

11 TPI vs 11.5 TPI. They should wind together far enough to provide a reasonable mechanical join.
Can you get good spanners/wrenches onto both ?

It works very well as is! Now that the antennas have arrived, I finally remembered to bring a G3/4" (ie straight thread, not taper) nipple to work, and it slides on perfectly, if a bit loose for pressure tightness. Maybe pipe compound and flax fibre would help. I then have reason to believe that taper thread will provide a stout fit, especially if mounted using flax and compound.

I will keep you posted; for this is time nuttery on an enterprise scale.

Flax fibre here is known as plumbers hemp with which any lube, graphite grease or anti seize compound would work just fine.
For the likes of a crox join on copper an old plumber taught me how to spin hemp to wrap for the crox seal and to keep it there with just a wipe of spit or more politically correct wipe of graphite grease.

[25 CPS]

I finally got some time on the bench until the Elenco smoked and one of the other things I was working on was getting the GPSDO and the two distribution amplifiers set up more permanently.

Top centre seemed to work best so that none of the test equipment that needs the 10 MHz feed is a worst case scenario distance from the D/As. I would've preferred status lamps on the front panel and power and signal connectors on the back to avoid having the status lamps not visible if the connectors were facing towards the rear to have easy cable runs to the equipment back panels. I picked the other option to have the status lamps facing forward so they're visible from the front of the workbench although this is going to cause some clumsy cable runs to wire everything up where it's going to have to be guided 180 degrees from the front towards the back of the shelf before going to where each one needs to be.

The frequency being displayed on the counter is out because I disconnected the GPSDO while moving everything around and it hadn't locked and stabilized yet. I've got the counter on it to confirm good 10 MHz out. I've also got a couple of other things powered up so I can see what their internal oscillators are doing and compare with and without external reference. This and the other things are going to wait until the basement airs out a bit.

[wkb]

As for distribution amps in my ham shack I got lazy and bought two Kramer video distribution amplifiers (total cost incl shipping from Germany €15), see the picts.

Did I really need them? Well.. I have a home built GPSDO based on a Samsung UCCM module which has its own distrib amp. And my Efratom HLN Rb reference also has its own distrib amp. But my second Samsung UCCM GPSDO has been boxed in a small case to allow for /P operation and does not have a distrib amp. Ah, reason for buying the Kramer's

69752C00-E7B8-4CB2-AD8A-14BE4C56E183.jpeg

CE5B5A28-572C-4462-9FAD-E51A9A8E206D.jpeg

Wilko

[Cubdriver]

'Want' and the ability to pay is sufficient justification in my book.

-Pat

[25 CPS]

Continuing with the 10 MHz project, I pulled two BNC cables off the back of my HP 3325A function generator and the 5334B frequency counter just out of curiosity to see what their internal oscillators were generating.

The 3325A's internal reference is a 1 MHz square wave and can accept a 1 or 10 MHz reference in. The internal is not too far off frequency:





Cramming the 3325A, counter, and scope into the same shot.



Using an analog scope, line the trace up with the graticule divisions, count them off, multiply by the time per division, then take the reciprocal using an appropriate calculator would be one way of doing it.



Or, mark it off with cursors and let the scope's math functions take care of measuring the delta-T and calculating the reciprocal.



The internal reference of the 5334B is a 10 MHz sine that's quite a bit larger in amplitude compared to the 3325A's square wave and not too far off frequency either but the earlier testing makes me wonder how stable over time it is. Anyhow, that's a glance at what's being replaced with external reference on the first two machines to get cut over.



On the right, a happy, glowing external reference indicator light.



And the 1 KHz sine out of the function generator into the counter is pretty much bang on since they're both marching to the same timebase reference now. The 3324B doesn't have an external reference indicator but the 3325A does so that should come on when I cut it over to the GPSDO.