How do your modern Seiko quartz movements keep time?

[saul]

How do your modern non-HAQ rated Seiko quartz watches run?

I have an Arnie (SNJ025/H851) and a mini-Golden Tuna (SNE498/V157). Both are rated at +/- 15s/month. Both gain steadily. Neither keeps time as well as my 3x7548 or 1x7546 examples or my 1st gen Citizen C023 Aqualand. Even my old 4633 based 4004 beater keeps better time.

I don't recall when last I set their time (haven't bothered with DST on them yet) and both are running about 30s fast. Within spec I suppose but utter fail compared to some 30+ yr old quartz. Is this an Achilles with the Solar quartz families or just bog standard performance these days?

I recently bought a Chinese made Dirty Dozen. Hell of a lot of value for $99 shipped from the other side of the planet. It is running a Seiko VD78. This is a $10 movement.

Have had it for a week and it is running 1 second ahead of atomic reference (of course by publicly stating that I have jinxed it). I would fully expect either of my Solars to be at least +5s after a week, which is rather sad.

[tommie]

I am probably one of the minority on here that pays no attention to seconds gained or lost haha, I see it as a bit of a blessing that it doesn’t concern me tbh, I am very pedantic about most things in life but this isn’t one of them, as long as I’m I’m reading the same time on an analogue watch (give or take a minute) as my phone, I’m happy, I don’t mind setting the time on my watches, I like the interaction with the watch maybe. But…. Saying that, my only true quartz are my tuna (sbbn017, seems spot on) my daily solar citizen, and my g shocks (isn’t digital the most accurate of any watch?)

My latest g is radio controlled so sets itself constantly, compared to my iPhone it is always bang on to the second, I’d highly reccomend it as a daily if accuracy is your thing

[tommie]

Ps.. they also do solar with Bluetooth versions now that sync to your phones clock

[inboost]

saul ,

I think a lot of modern watches are forgoing any kind of tuning capabilities to be able to hone them in and this is where the error would likely come from. With no way to adjust it, the watch is doomed to whatever the crystal oscillator inside is giving for a heartbeat. That can be different watch to watch simply by how it's soldered to the circuit block or maybe how it rests or touches against other parts.

Many upper end modern quartz watches do 'self regulate' in a move called "inhibition". This is where the tick-to-tock quartz step is typically fast but at some kind of interval (often every 10s or 60s) the watch will hang back on the pulse to move the sweep hand a minuscule amount slower and check the fast heartbeat of normal pulses with that slower one. You can't see it with your eye as it's slicing off milliseconds but it works. I believe the watch is watching the step current and timing it's offset from the internal quartz crystal clock beats and makes decisions accordingly. Of course if the internal quartz oscillation frequency isn't spot on (and it won't be) then there can still be error with this adjustment system.

Of course we also have the fact that temperature plays a huge part in quartz oscillator frequency stability. The manufacturers tune watches to keep best time on the wrist at a temperature a good ten or so degrees above storage temp. So if you find a modern quartz watch is gaining or loosing a lot but is never worn, try keeping it on the wrist 24/7 and checking it with something like the 'WatchCheck' phone app and you might see it do better.

At the end of it all we have cheap robotic assembly and fully plastic powertrains so I'm sure these play a part in accuracy as well. There's no human looking over it as it goes together and upon seeing a little bit of mold flash or other abnormality would correct it.

The fact that we are surrounds by GPS disciplined network time clocks (phone, PC, TV. etc) doesn't make a great case for trying to get a watch to that level without just adding those same capabilities. This is why I think we see a lot of GPS and bluetooth connected watches these days.

I'll stick with my older vintage stuff. Some of it can be made amazingly accurate, and has way more charm for me!

Inboost

[saul]

Mar 22, 2024 1:13:40 GMT -8 tommie said:

I am probably one of the minority on here that pays no attention to seconds gained or lost haha, I see it as a bit of a blessing that it doesn’t concern me tbh, I am very pedantic about most things in life but this isn’t one of them, as long as I’m I’m reading the same time on an analogue watch (give or take a minute) as my phone, I’m happy, I don’t mind setting the time on my watches, I like the interaction with the watch maybe. But…. Saying that, my only true quartz are my tuna (sbbn017, seems spot on) my daily solar citizen, and my g shocks (isn’t digital the most accurate of any watch?)

My latest g is radio controlled so sets itself constantly, compared to my iPhone it is always bang on to the second, I’d highly recommend it as a daily if accuracy is your thing

I don't mean to present myself as accuracy obsessed, I am not. Out of 44 watches, 13 are quartz. Most of the mechanical as well as the quartz are older than 25 years, a few from the 1940's and 1950's. If I were obsessed with accuracy I would probably go mad. I only get concerned about a mechanical if it is running wildly out of spec. I don't need functional resolution of time with more precision than to the nearest minute and even then...

I was more curious about other's observations about modern vs vintage quartz. It is totally a "they don't make 'em like they used to" proposition.

I have often wanted to have an atomic time source in the house for easy reference but having purchased and gifted out several radio controlled G's I found there is no place in my house where I can get a signal to sync. My G5600E-1 (3160 module) keeps excellent time and serves as a dresser top reference for setting other watches.

[saul]

Mar 22, 2024 4:25:09 GMT -8 inboost said:

Of course we also have the fact that temperature plays a huge part in quartz oscillator frequency stability. The manufacturers tune watches to keep best time on the wrist at a temperature a good ten or so degrees above storage temp. So if you find a modern quartz watch is gaining or loosing a lot but is never worn, try keeping it on the wrist 24/7 and checking it with something like the 'WatchCheck' phone app and you might see it do better.

 

I only offer up my precious bodily warmth to oiled and jeweled gear trains. 

[wx]

Yea I do certainly agree with your observation on the quality of older quartz movements vs the modern ones out there. I do wonde  if it really boils down to the fact that in the 80s-90s, the “premiumisation” of quartz meant that watch manufacturers paid a lot more attention when constructing those quartz movements, and consequently meant that they were far more accurate and robust than in today’s world where regular quartz movements are made to be easily replaced.

I still remember coming across those old Seiko catalogues where quartz watches were priced even higher than hi-beat movements 😂  

[small]

Heck I have a Sony brand radio alarm clock that runs fast and refuses to DST! I bet it gains 5 minutes over the course of 6 months/ I reset/sync it along with my Casio digitals and none are the same six months later...


Reading the comments by inboost, in my case, most don't get worn all that often, and it might be the temp thing as he said...   

[HiBeat]

Great topic and timely indeed.  In my golf bag I have a lightweight 7N43 all titanium thin beater that I wear on the course. I just found it over the weekend, and it was properly set to DST, so  that means it has been going non-stop for a minimum since we last moved to standard time whenever that was.

I will it 5 months.  It was running 2 minutes fast, but it wasn't set to anything standard but let's just say it was 120 seconds fast over 150 days.   Close enough to show it was under a second per day in the cold garage.

Meanwhile using my QT99 tester I can set a 7546 or 7548 to 0.1 spd

[victor]

I don't have the info open in front of me, but cribbing from a few sources:

Seiko in the '70s was laser focused on achieving accuracy levels that had never before been reached for wristwatches.  Early in the decade they recorded +/- 20 seconds per month with the first high end quartz and by mid decade they achieved +/- 5 sec per month with the 4843 Grand Quartz with single quartz, and by 1978 reached +/- 10 seconds per year with the 9943 Superior, while the 9980 and 9983 achieved +/-5 seconds.

Currently a Spring Drive Grand Seiko can get +/- 15 seconds per month, and a 9F Grand Seiko quartz can get +/- 10 sec per year.  In other words, Seiko achieved accuracy levels nearly 50 years ago that they don't achieve today.  Oh yeah - I believe I saw that there were no pure quartz movements in the Grand Seiko lineup last year, for the first time ever.

Still, the grand prize for quartz accuracy has to go to Citizen - its Calibre 0100 promises +/- 1 second per year.

[inboost]

To bolster what victor has told us, those second per year level high accuracy quartz watches ALL use temperature compensation schemes of one kind or another to get to their level of performance. You know your mechanical watches actually have a similar problem with temperature swings but we don't see it because of their order of magnitude worse day-to-day errors. The little stuff like temperature error just hides in the weeds.

Interesting side note - more precise long case clocks (Grandfather's clocks) have long tubes of mercury filled glass as pendulum bobs. This is useful because when a pendulum gets cold it's pendulum rod shrinks and the clock speeds up. The opposite is true when it gets hot. Now if you introduce some columns of mercury that are very dense and also expand/shrink noticeably with temperature you can offset the effect of temperature. This happens because as the cold pendulum gets colder, it shrinks and that lifts the pendulum bob higher which makes a clock run faster. However the mercury also shrinks in the cold, and therefore its center of mass moves down which counteracts the rod shrinkage going up. Talk about a very elegant solution provided by physics to reduce temperature effects on a precision time piece!

So HiBeat you need to switch your golf beater to a tuned up twin-quartz Seiko and you'll be able to leave it in the garage without much suffering of accuracy. Now, the battery doesn't like low temperatures but that's a different outcome.

[HiBeat]

OK so here is a collection of various old and new Seiko quartzes and yes even a 20+ year old Timex Indiglo (my late father in law's, still my 'going to the movies' watch except I haven't actually gone to a movie since the pandemic started, but I will sooner or later).

Every one was tediously set to perfect sync with Time.is when the clocks changed on March 9 about 10 pm.

To save everyone the effort:

Timex Indiglo  circa 1998      3 seconds fast

5C23 circa 1995                  20 seconds fast

Seiko Solar V157                 0 seconds off flat out perfect

Seiko Arnie H851                 2 seconds slow

Vintage

7548                                   3 seconds slow

7C43                                   5 seconds slow

Modern 2022 7C46 (in a vintage 7549 case misleading)    6 seconds slow

7549                                   13 seconds slow 

Now it's been 13 days exactly so these seconds deviation divided by 13 is the seconds per day accuracy

            SPD Accuracy

Timex       +0.23 spd

5C23        +1.54

V157        ZERO SPOT ON PERFECT

H851        -0.15

7548        -0.23

7C43        -0.38

7C46        -0.46

7549        -1.0

Now every one of these is reasonable except my grandfather's old 5C23 which will gain a minute every 6 weeks.

The ancient Timex Indiglo will be 1 1/2 minutes off per year.

Again the simply amazing vintage 7548 (yes I tweaked it on the QT99 a few years ago) still shines like a beacon in performance.

The 7C46 (inside the 7549-7010 tuna case) disappoints at 0.5 spd, the movement is brand new last year straight from Seiko.  I remember camrok had a thread where we all challenged our quartzes and at that time I had a brand new modern 7C46 in an older Golden Tuna 7C46-7009 case and did not perform well then.  It's only 2 movements but neither one of mine is impressing in this WIS regard.

The V157 Seiko Prospex Solar is spot on.   I wear this a lot especially in the summer.  I'll be keeping a 'watch-ful' eye on this one

I am impressed by the performance of the SNJ028 New Arnie, going into hibernation at times in the dark, keeps time inside a minute a year which is all a watch need do, especially when living in a place with the changing of the clocks twice a year like I do.

[meepokta]

Mar 22, 2024 7:09:08 GMT -8 wx said:

Yea I do certainly agree with your observation on the quality of older quartz movements vs the modern ones out there. I do wonde  if it really boils down to the fact that in the 80s-90s, the “premiumisation” of quartz meant that watch manufacturers paid a lot more attention when constructing those quartz movements, and consequently meant that they were far more accurate and robust than in today’s world where regular quartz movements are made to be easily replaced.

I still remember coming across those old Seiko catalogues where quartz watches were priced even higher than hi-beat movements 😂  

I think you may have a point. I have several quartz watches from the 90s using the 8F movement with the perpetual calendar function. I suppose this feature would be considered premium and would add to the manufacturing cost. And perhaps not popular with quartz watch buyers nowadays, hence very few watches offer this. Even the Seiko 9F series doesn’t have perpetual calendar. Only Citizen Chronomaster offers this.

[wx]

Mar 22, 2024 19:19:21 GMT -8 HiBeat said:

OK so here is a collection of various old and new Seiko quartzes and yes even a 20+ year old Timex Indiglo (my late father in law's, still my 'going to the movies' watch except I haven't actually gone to a movie since the pandemic started, but I will sooner or later).

Every one was tediously set to perfect sync with Time.is when the clocks changed on March 9 about 10 pm.

To save everyone the effort:

Timex Indiglo  circa 1998      3 seconds fast

5C23 circa 1995                  20 seconds fast

Seiko Solar V157                 0 seconds off flat out perfect

Seiko Arnie H851                 2 seconds slow

Vintage

7548                                   3 seconds slow

7C43                                   5 seconds slow

Modern 2022 7C46 (in a vintage 7549 case misleading)    6 seconds slow

7549                                   13 seconds slow 

Now it's been 13 days exactly so these seconds deviation divided by 13 is the seconds per day accuracy

            SPD Accuracy

Timex       +0.23 spd

5C23        +1.54

V157        ZERO SPOT ON PERFECT

H851        -0.15

7548        -0.23

7C43        -0.38

7C46        -0.46

7549        -1.0

Now every one of these is reasonable except my grandfather's old 5C23 which will gain a minute every 6 weeks.

The ancient Timex Indiglo will be 1 1/2 minutes off per year.

Again the simply amazing vintage 7548 (yes I tweaked it on the QT99 a few years ago) still shines like a beacon in performance.

The 7C46 (inside the 7549-7010 tuna case) disappoints at 0.5 spd, the movement is brand new last year straight from Seiko.  I remember camrok had a thread where we all challenged our quartzes and at that time I had a brand new modern 7C46 in an older Golden Tuna 7C46-7009 case and did not perform well then.  It's only 2 movements but neither one of mine is impressing in this WIS regard.

The V157 Seiko Prospex Solar is spot on.   I wear this a lot especially in the summer.  I'll be keeping a 'watch-ful' eye on this one

I am impressed by the performance of the SNJ028 New Arnie, going into hibernation at times in the dark, keeps time inside a minute a year which is all a watch need do, especially when living in a place with the changing of the clocks twice a year like I do.

I can’t speak for the other movements, but I’m rather shocked and disappointed with the performance of the modern 7C46. Based on your observations, it would be running at -12 secs/month. And while thats within specs, I would’ve expected better, especially given its a rather new one. Thankfully my 7C46-7009 from 1993 (which I think would likely be a 7C46A) is running pretty well,, I don’t think its been serviced, but doing at about +3 secs/month.

Which also feeds into my assessment that unfortunately they just don’t manufacture quartz movements like they used to anymore..