The SMITHS QUASAR "Constellation"

[abraxas]

The Quasar from The National 15 - The decline of British watchmaking and the role of Smiths*

I am putting this image here first... as the ad on the right always squashes the first post and I wanted you to first see the Quasar "Constellation" in its full glory (below in the next post).



Basically, before I saw these cases in the box here, the only Quasar by Smiths I was aware of, was the TV dialed rendering above from "The National 15 - The decline of British watchmaking and the role of Smiths in a hoped for discovery" by David Read.

*On page 15/88
https://ahsoc.contentfiles.net/media...-_Read_wm6.pdf

[abraxas]

"On visiting a watch mate today got to wear a Smiths quasar prototype. Made in the early 70's it is the only ever English quartz watch ever made. Very high quality and extremely rare."
https://www.imgrumweb.com/post/BViBH7bnq8w

QUASAR "Constellation" by Smiths

Smiths was well advanced in quartz cock movements for aviation, maritime and domestic clocks when Cheltenham watch production ended, but persevered with a quartz wristwatch at their secretive 'Delphina?' research lab in Cheltenham launching their Quasar at the 1973 International watch fair... these are not even signed 'Smiths' but, if my memory serves me well, 'Quasar Time Systems' of whom very little is known. Its was clear at the show that Smiths was wasting its time and money so the Quasar was abandoned shortly after... so the 1974 label is probably one of the last tests.

Whether Smiths sold the 'Quasar' brand/project to Stephen Strauss, who gained Ingersoll, is another unresolved mystery.
[#23 Barry - Brenellic2000]

Links:

Flyaway Smiths at Cotswold auction room today
https://www.mwrforum.net/forums/show...ion-room-today

Smiths' last days in the sun / but will they run ? (Lots of pics)
https://www.mwrforum.net/forums/show...-will-they-run

* * *



MWR reckon that the Smiths case was based on the Omega Constellation 168.0054. But it has a serrated bezel...



On the other hand the Omega Constellation 368.846 has a smooth bezel but is without the DAY complication. Here is a lovely one to study:
https://www.poshtime.com/3028.001/Om...tic-c.1970.php

* * *

So,

1. I love the two crowns of the Quasar. They change the look of the Omega case. By the way, the second crown is for adjusting the Day-Date.

2. What is going on with all the Omega connections? And why Omega?

3. How about a homage? The current and next big-thing is integrated bracelets, as per Naulilus, RO, VC Overseas, etc...


The back of the Quasar.

[size11s]

Thank you for posting this it's nice to see images of the fabled Quasar. That Cotswold auction has yet to release the hoped for information (by me anyway) that seemed to be potentially held in many of the lots.

[Dave+63]

That’s very nice and not a watch I was aware of.
It is very Omegaesque though although that’s no bad thing.

[Mr Curta]

I was just having a bit of fun with the maths to see what was possible with unusual frequencies, but I’ll try to look at the relevant documents at the weekend. I’m using MSBs and LSBs plural to indicate sequential strings of bits from the far left or right of a base-2 number respectively, it‘s a usage that I’m comfortable with and Wikipedia seems to agree (not that that proves anything).

[Rev-O]

Unfortunately (due to the fact the I didn't really go to school in any normal sense of the word combined with my very own -- and vast -- stupidity; I'm not sure any school could have made much of such poor raw material) I am completely lost by this thread.

I have a decent layman's grasp of amplitude and even isochronism but anything involving figures with more than four digits is simply too big a number for me.

But I am always glad that there are people who know and understand these things.

Carry on chaps! As you were!

[abraxas]

If you start working on MWR please give us a link so we can follow the discussion. Thanks.

[M4tt]

Unfortunately (due to the fact the I didn't really go to school in any normal sense of the word combined with my very own -- and vast -- stupidity; I'm not sure any school could have made much of such poor raw material) I am completely lost by this thread.

I have a decent layman's grasp of amplitude and even isochronism but anything involving figures with more than four digits is simply too big a number for me.

But I am always glad that there are people who know and understand these things.

Carry on chaps! As you were!

An electronic overflow integrated circuit isn't hard to visualise, just think water:

Imagine a tap running water into a pair of cups.

Every time both of the cups (2 in total) are full, they empty into a second set of cups.

Every time both of these cups (4 in total now) are full they empty into a third set of cups.

Every time both of these cups (8 in total now) are full they empty into a fourth set of cups.

Every time both of these cups (16 in total now) are full they empty into fifth set of cups.

Every time both of these cups (32 in total now) are full they empty into a sixth set of cups.

Every time both of these cups (64 in total now) are full they empty into a seventh set of cups.

Every time both of these cups (128 in total now) are full they empty into a eighth set of cups.

Every time both of these cups (256 in total now) are full they empty into a ninth set of cups.

Every time both of these cups (512 in total now) are full they empty into a tenth set of cups.

Every time both of these cups (1024 in total now) are full they empty into an eleventh set of cups.

Every time both of these cups (2048 in total now) are full they empty into twelfth set of cups.

Every time both of these cups (4096 in total now) are full they empty into a thirteenth set of cups.

Every time both of these cups (8192 in total now) are full they empty into a fourteenth set of cups. (First in the Beta 21 or 35SQ)

Every time both of these cups (16384 in total now) are full they empty into an eleventh set of cups. (First in the Seiko Cal.3832)

Every time both of these cups (32768 in total now) are full they sound a gong! (First in the Girard Perregaux Cal.352)

And in this way, by simply repeating a simple process fifteen times you can divide a 32768hz vibration down to a 1hz pulse suitable for driving a stepper motor.

Marie Curie discovered that you could vibrate quartz, but making a watch out of it had to wait for the germanium and then silicon integrated circuit. Like most computing and calculating, it's not hard, it's just incredibly repetitive.

[alfat33]

Marie Curie discovered that you could vibrate quartz,.

Pedantically, I think that was her husband and his brother.

[M4tt]

I was just having a bit of fun with the maths to see what was possible with unusual frequencies, but I’ll try to look at the relevant documents at the weekend. I’m using MSBs and LSBs plural to indicate sequential strings of bits from the far left or right of a base-2 number respectively, it‘s a usage that I’m comfortable with and Wikipedia seems to agree (not that that proves anything).

Ooh, that's lovely.

I don't think that we are in any danger of misunderstanding each other and as you were explaining something I didn't fully grasp, it's churlish to labour the point, so I'll stop. I'll be really interested to see what you make of it all. I'll try to dig up a patent from Smiths closer to the exact watch.

[HookedSeven]

Hey... my Quasars !

Edit to add:

I still have pretty much everything over on the MWR thread and am happy to answer any questions anyone might have. They’re great cases and all very interesting but getting them running is an uphill struggle. Smiths obviously found the same !

[abraxas]

I would really like to know the reason behind Smiths using the Omega case/bracelet design. Also, were they actually intending to sell the watch with this case?

[BillN]

The more you get into "watch appreciation" the more you realise how many different watches there are - millions upon millions must have been produced.

Thanks for starting the thread - always something new to learn ........ as they say "everyday is a school day"

[HookedSeven]

I would really like to know the reason behind Smiths using the Omega case/bracelet design. Also, were they actually intending to sell the watch with this case?

Agree this is an interesting question, but no idea how to research further. I scoured all the other lots in the auction looking for documentation, or for a lead as to who the items belonged too, but no luck. Also the auction house couldn’t give me any details of the consignor.

[size11s]

Agree this is an interesting question, but no idea how to research further. I scoured all the other lots in the auction looking for documentation, or for a lead as to who the items belonged too, but no luck. Also the auction house couldn’t give me any details of the consignor.

Did you get a good look at the lot with all the original drawings/specs in it? Maybe there was something in there.

[HookedSeven]

Did you get a good look at the lot with all the original drawings/specs in it? Maybe there was something in there.

You mean from the original Cotswold auction ? These Quasars came later from an auction in the SW. I had high hopes for the original auction too but every man and his dog turned up !

[snowman]

Made in the early 70's it is the only ever English quartz watch ever made.

Doesn't the Sinclair watch count?

M

[HookedSeven]

Doesn't the Sinclair watch count?

M

No idea !

Was it entirely UK built, or did the chip, LED module, or other bits come from abroad. Didn’t they blow up or stop working after a week ? If so that’s better than Smiths managed ;-)

[snowman]

No idea !

Was it entirely UK built, or did the chip, LED module, or other bits come from abroad. Didn’t they blow up or stop working after a week ? If so that’s better than Smiths managed ;-)

No idea either, which is why I asked.

M

[HookedSeven]

No idea either, which is why I asked.

M

Oh if I’d known you had no idea I’d have just made something up :-)

[snowman]

Oh if I’d known you had no idea I’d have just made something up :-)

I remember it being 'the next big thing' when I was a kid and one of my friends got one for Christmas, but I could never understand the appeal of something you had to press the button to see the time, to be honest.

I recall it being touted (by Sir Clive) as a 'Great British product', but I don't know how much was made here...

M

[alfat33]

Really interesting to see Ollie and thanks for the writeup.

Designing your own quartz movement at the time must have been a brilliant engineering challenge. I don’t blame them at all for wanting to have a go.

Looking forward to seeing some more pics and technical details when it arrives.

[Rev-O]

Will do!

Thanks for your interest -- these are of limited appeal but fascinating to me.

[Rev-O]

BTW, I know very little about quartz watches.

I understand that 14 jewels is a lot for a quartz but have no idea about the 1.5MHz thing. Is that "high frequency"? I'm guessing it's probably prehistoric by todays standards, but maybe not?

[Daddelvirks]

Thanks for a very interesting read.

I really didn't know Smiths were even involved in quartz at all.
And with its original box, wow, a real gem added to your Smiths collection.

[HookedSeven]

BTW, I know very little about quartz watches.

I understand that 14 jewels is a lot for a quartz but have no idea about the 1.5MHz thing. Is that "high frequency"? I'm guessing it's probably prehistoric by todays standards, but maybe not?

It seems that the standard for quartz watches is 32768Hz (32.768kHz), which is 2 to the power of 15.

I’m not sure if there are higher frequencies in different types, but if the Smiths is running at 1.5MHz then that’s a heck of a lot faster. There’s even a 4 MHz labeled example amongst my Quasar bits so they were obviously playing with frequency. But maybe it’s wrong to assume that a lower frequency is prehistoric. It might be easier to get better stability from crystals with lower frequencies. I’m sure somebody knows...

[alfat33]

BTW, I know very little about quartz watches.

I understand that 14 jewels is a lot for a quartz but have no idea about the 1.5MHz thing. Is that "high frequency"? I'm guessing it's probably prehistoric by todays standards, but maybe not?

I’d wondered about that. I’m way out of date on this stuff. But it stood out for me so I did a bit of Googling. Most modern quartz watches are 32kHz which I guess everyone knows. 262kHz is considered Ultra High Frequency for watches. Therefore 1.5MHz = 1500kHz must be extremely high frequency.

Multiple Megahertz quartz oscillators these days are more likely to be used in different applications like communications etc.

Another interesting angle is that a designer would cut a quite different shape quartz crystal for a low frequency application (eg 32kHz) vs high frequency (1MHz +). The physical vibration modes are different. There are also different basic circuit designs depending on how your crystal is cut and how you want it to perform.

I found this article which explains quite a bit of it if you can navigate the physics.

https://apps.dtic.mil/dtic/tr/fulltext/u2/a256373.pdf

I deduce from this it is quite likely that the Smiths engineers may have made a unique design of crystal and circuit, no doubt to the highest purist principles, while the rest of the world was gravitating towards the easy-to-mass produce ‘tuning fork’ shaped low frequency crystals and associated circuits.

That said I may have misinterpreted this; I am a bit rusty and some of the forum quartz experts may be able to clarify.

[M4tt]

I’d wondered about that. I’m way out of date on this stuff. But it stood out for me so I did a bit of Googling. Most modern quartz watches are 32kHz which I guess everyone knows. 262kHz is considered Ultra High Frequency for watches. Therefore 1.5MHz = 1500kHz must be extremely high frequency.

Multiple Megahertz quartz oscillators these days are more likely to be used in different applications like communications etc.

Another interesting angle is that a designer would cut a quite different shape quartz crystal for a low frequency application (eg 32kHz) vs high frequency (1MHz +). The physical vibration modes are different. There are also different basic circuit designs depending on how your crystal is cut and how you want it to perform.

I found this article which explains quite a bit of it if you can navigate the physics.

https://apps.dtic.mil/dtic/tr/fulltext/u2/a256373.pdf

I deduce from this it is quite likely that the Smiths engineers may have made a unique design of crystal and circuit, no doubt to the highest purist principles, while the rest of the world was gravitating towards the easy-to-mass produce ‘tuning fork’ shaped low frequency crystals and associated circuits.

That said I may have misinterpreted this; I am a bit rusty and some of the forum quartz experts may be able to clarify.

For the period it wasn’t that high. Omega produced a 2.4 MHz Marine Chronometer, while Junghans, Citizen and even Casio produced a range of watches at 4.19Mhz, all with lenticular crystals rather than tuning forks.

I’d presume first that the Smiths is sporting a lenticular crystal and second that the excessive jewels and odd frequency (for base 2) suggests that it wasn’t entirely relying on an IC for stepping down to 1 Hz. Does anyone know whether it used a stepper motor, a small escapement or a small tuning fork to transition from solid state to mechanical? Does it tick or is it watch smooth or tuning fork smooth?

If there are parts, has anyone seen the tooth counts on the wheels and cogs of the train? Did Smiths buy in their IC or did they develop their own? Smiths instruments would have had good reason to develop a UK IC for military uses, but would that have been passed to a dying Smiths? More to the point, does anyone have the 1973 issue of the HJ in which the watch is described?

Whatever the answers, it’s a fine looking watch both outside and, especially, inside. Congratulations.

[Mr Curta]

What are the chances of a YouTube vid rev'd

[Mr Curta]

We need to see one running.

[M4tt]

We need to see one running.

Here’s a few of my Omega ticking from 1hz to 720hz

https://m.youtube.com/watch?v=Mj1GOACnbJc

Here’s a couple more:

https://m.youtube.com/watch?v=smfx2F61cSg

https://m.youtube.com/watch?v=hKEYK4GhsBo

[size11s]

It looks spectacular Ollie, very many congratulations. The box is a fabulous extra bonus too even with its ('just got out of bed') bed sheet surround . I remember John Senior saying that he thought the movement was very good quality after the Cotswold auction.

[Mr Curta]

I've only just got round to properly looking at the other articles. The second one says that IC reduces f to 0.5Hz.

[M4tt]

I've only just got round to properly looking at the other articles. The second one says that IC reduces f to 0.5Hz.

Indeed, but with two discrete stators, and four rotor magnets, that offers a far greater range of possibilities than the standard dipole stepper motor. However, as you know, magnetic effects drop off sharply and so such a large and crude motor is liable not to be doing anything too clever. My bet would be both acting in concert, with the leftmost pulling the incoming magnet, while pushing the middle magnet to its partner which in turn would be pulling that magnet while pushing the outgoing magnet. This would generate the most torque of any setup I can think of.

With four magnets, it also means the motor would rotate half a turn in two increments every second. The idea that we have a four pole, two stator, stepper motor set up that acts half like a conventional dipole two stator motor turning the rotor by half a turn every half a second seems silly.

The moment you start thinking about motion rather than discrete steps, the possibilities open up a bit - and the power drain drops right off as you are just maintaining momentum...

(I confess I'm in the middle of a migraine and so this may not make sense. Sorry)

[Rev-O]

From the mwr thread:


- it is a one second jump with a merry little bounce (a mechanically poor design)
- it needs to be set by TIM on the second pip, not third as there is a one second delay
- (mine) eats batteries... 9 months is about right
- there is a serious design flaw in the age-hardening flexible pcb. It is repairable by skilled engineers with ingenuity
- it's a lovely watch which excellent time keeping
- they're b***y heavy - Charles Atlas course recommended (or 3 Weetabix)!
- there were a nurse's fob watch and pocket-watch prototypes... and doubtless several others in the pipeline
- this was a 'sideline' to Smiths' advanced aviation, motoring and marine quartz clock technology... see 'The Golden Years' -parts 1 and 2.

Enjoy!

[Mr Curta]

HookedSeven's collection of circuit boards and development models must have some stories to tell. One of them is labelled as 4MHz.

[HookedSeven]

Can I help with lending any bits ?

[M4tt]

Can I help with lending any bits ?

There’s a whole lot of expertise here, so possibly the most useful thing would be decent resolution images of all the loose bits, especially the bits that are a bit mysterious. This way everyone can work towards a consensus and give the wisdom of crowds a chance.

This has to have some sort of ‘motor’, some sort of IC and so on. With fifteen jewels, it obviously has a considerable train as well. Ultimately both the IC and train are doing maths and clearly more of that maths is being done by the train as 1.5khz isn’t on a base two progression.

Does anyone know how long the battery lasted, I bet it’s far less than a year.

[Mr Curta]

You mean MHz, and it's a power of two geometric progression (albeit utilised in base-2) that usually determines the oscillator frequency that quartz timing circuits employ to derive a pulse, but I understand your thinking.

However it doesn't have to be powers of two, as MSBs can effectively be ignored by generating a pulse using a binary digital counter that has fewer bits than the frequency. Rather neatly the 16 LSBs of a binary counter driven at 1.507328MHz lend themselves very well to counting in seconds, half seconds, quarter seconds and so on, depending on the number of bits employed by the counter.

1507328₍₁₀₎ = 101110000000000000000₍₂₎

[Rev-O]

OK! I've got the relevent issues of HJ from 1974 -- will post scans later

May I suggest we start a new thread either over at Watch Talk OR (better) at the specialist Smiths subforum on mwr?

That would give better exposure, access to experts and stop us from cluttering up the Time Factors forum!

I'm a mod at mwr so could (I think) move the contents of this thread over there.

With the info from HJ and access to some parts (or at least pics of parts) we'd have a very good start.

[Rev-O]

All material copyright BHI / HJ

[Mr Curta]

Really appreciated, John. It's great to see one running.

[M4tt]

You mean MHz, and it's a power of two geometric progression (albeit utilised in base-2) that usually determines the oscillator frequency that quartz timing circuits employ to derive a pulse, but I understand your thinking.

However it doesn't have to be powers of two, as MSBs can effectively be ignored by generating a pulse using a binary digital counter that has fewer bits than the frequency. Rather neatly the 16 LSBs of a binary counter driven at 1.507328MHz lend themselves very well to counting in seconds, half seconds, quarter seconds and so on, depending on the number of bits employed by the counter.

1507328₍₁₀₎ = 101110000000000000000₍₂₎

Ok, I think I see the logic, even though I was always taught that the LSB and MSB were singular, I see what you mean. but I'm unclear how that would be instantiated in the sort of simple IC overflow counter that was available at the time. I'm not disagreeing, only asking for an explanation.

Sorry about the daft mistakes, I'd just struggled through an RSC production of Taming of The Shrew with all the sexes reversed, which was interesting, but surprisingly hard going. I was probably unwise to post before flopping into bed.

[M4tt]

So it's an AT cut lenticular crystal. The same as the Omega 2.4. Hardly suspicious at all!

[M4tt]

Ok, if my French is up to it, I've got the seminal patent:

https://worldwide.espacenet.com/publ...A1&KC=A1&ND=4#

I also notice that Smiths patented a bloody tuning fork too - now there's a Smiths to be looking out for!

[Mr Curta]

Ok, if my French is up to it, I've got the seminal patent:

https://worldwide.espacenet.com/publ...A1&KC=A1&ND=4#

I also notice that Smiths patented a bloody tuning fork too - now there's a Smiths to be looking out for!

That's interesting, I know about the Smiths tuning fork clock movement but not one for a watch.

[bobbee]

Ok, if my French is up to it, I've got the seminal patent:

https://worldwide.espacenet.com/publ...A1&KC=A1&ND=4#

I also notice that Smiths patented a bloody tuning fork too - now there's a Smiths to be looking out for!

Hi Matt, a translating option is available on the Espacenet site:

http://translationportal.epo.org/emt...=fr&TRGLANG=en

[alfat33]

I also notice that Smiths patented a bloody tuning fork too - now there's a Smiths to be looking out for!

I sincerely hope that wasn’t this that I forgot to bid on last month:



https://rover.ebay.com/rover/0/0/0?m...2F254481545146

[Mr Curta]

Ok, I think I see the logic, even though I was always taught that the LSB and MSB were singular, I see what you mean. but I'm unclear how that would be instantiated in the sort of simple IC overflow counter that was available at the time. I'm not disagreeing, only asking for an explanation.

Most significant bits (plural) are the bits closest to, and including, the MSB. Similarly for LSBs. A simple 13 bit binary counter will overflow exactly 189 times per second at 1.548288MHz (every time all 13 of those trailing zeros appear simultaneously) . That sounds like an odd number but it factorises into usable numbers for gearing.

It may be easier to visualise with lower numbers. 40Hz (not a power of 2) will overflow a 3 bit counter exactly five times a second - every time the counter passes 111₂. Converted to base-10 that happens at counts of 8, 16, 24, 32 and 40.

This could all be absolute nonsense and it's achieved in a completely different manner, but that way seems fairly straightforward.

[M4tt]

Most significant bits (plural) are the bits closest to, and including, the MSB. Similarly for LSBs. A simple 13 bit binary counter will overflow exactly 189 times per second at 1.548288MHz (every time all 13 of those trailing zeros appear simultaneously) . That sounds like an odd number but it factorises into usable numbers for gearing.

It may be easier to visualise with lower numbers. 40Hz (not a power of 2) will overflow a 3 bit counter exactly five times a second - every time the counter passes 111₂. Converted to base-10 that happens at counts of 8, 16, 24, 32 and 40.

This could all be absolute nonsense and it's achieved in a completely different manner, but that way seems fairly straightforward.

It's a minor thing, but are you sure about the MSB thing? I'm pretty sure that the MSB is one end, the LSB the other and all the others are merely SBs. The reasoning I was taught for the names was that flipping the MSB gave the largest change in the number while flipping the LSB gave the least change. Perhaps it's just a matter of taste, but how would you distinguish between the MSBs and LSBs when they meet in the middle of the string? It's not a big deal, just terminology, but it confused me.

Either way, understanding dawns, that's really cool. Thanks.

Have you looked at the patent and the circuit diagram? I'm pretty sure it's an earlier paper that marked a significant point in development, but it looks to be well before the final version. That at least looks to be an iterated binary overflow. However, the main article talks of binary being old hat and the Quasar dividing by 'higher factors' (note the plural!) and outputting not 1 hz but .5

And, by way of ante: