dot matrix

[clockace]

continuing the saga. i have now installed the trimmer, uninstalled the trimmer, changed the qc with out much change in accuracy from my last post of dec22. on the 23rd i reinstalled the trimmer and now it is running dead on with my atomic clock! this one has been a real challenge, but also the results are now worth it. it is a 4 reed mkI/II pcb with dot matrix display. the speed is adj by a combo of the sliders and rear mag. i think the speed is adj as slow as it will go now because the little bar between the leap yr[4] and the secs display goes away if i slow it down one more notch. i am not sure this is a mkII+ or what. howard said it may be something his father was working on just before the mkII. i am hoping it will continue to keep the correct time or at least run slow over time and then i can adj it with the trimmer. it has now become my favorite led wearer. peter

[Synchroserious]

Peter, This confirms something i was not sure of till now.That is that some MK II,s had the dot matrix display.If you were able to adjust the rate with combo of switches it is a MK II from late 1975 through 1976.This probably means the trimmer is obsolete and it is causing your range problem.I take it the module had no trimmer from the start? Did you try the switch combo scheme before installing the Trimmer?

In any event great work and i am sure it took lots of patience.Congrats!At this point it is running too good to attempt any modifications.Another way of slowing it down would be to not wear it 7 days a week.

Being a Dot Matrix Mk II it represents the earliest form of the Synchronar and it's long run as the most accurate Quartz watch available.Today it's rank(MK IV 1981-present) is comparably equaled only by a few very expensive thermo compensated movements.Nevermind about the by chance atomic signal watches.

[clockace]

hi howard, when you sent the pcb it only had the dot display attached. no sc, qc, trimmer altho' the solder pads are there for it. i tried it both with and without the trimmer. without, the closest i could get it was +2secs/day. the adj with the sliders and rear mag would not allow it to go any slower. there must be a min/max range on the earliest ones like that. adj slower with these right now does not do anything. if i go one notch slower the divideing bar between the leap yr[4] and the secs display goes away. is there a limit on the mkIII/IV's? i am very happy with its performance as it is now. this one has given me the most satisfaction of all those i got from you that i have running. only one of them was a failure, but it too contributed to the repair of one of the others. it is the one i unsoldered the chip from and installed it on one that had some of the contact legs corroded off. peter

[Sully008]

Peter,

Congrats on getting your dot matrix more accurate. Is there anywhere that shows the combinations of sliders/magnets needed to get to the speed adjustments on a MkI/II? I've never had to adjust either of the modules I have, but I've never checked accuracy that much . They are in sync with the clock I used to adjust them to.

[Synchroserious]

Peter,maybe it is a MK I/II hybrid.Was the QC you used an old one or a new tested one?.On almost all occasions when there is little range it is a sign of a QC that is out of range.At least that has been the case on the 500 Mk IV,s i have made.

If the range is like yours something is not quite right.However with it's present performance you might as well leave it be.The MK III,s and MK IV,s also have a wide range of adjustment on the frequency synthesizer circuit that are never supposed to max out one way or the other.

Remember that 3 out of 10 new $2.50 QC,s are either slightly off or require more voltage to start than the best 7.

[clockace]

i agree that something may not be quite right. the qc's i have used on all of mine that are replacements are from the same source. zantech, inc. they are not expensive, but all the others have been well within range. the fact that the 2nd one i installed did not make that much difference should verify at least consistancy there. like you, howard, since i got it as close as it is, i am not willing to mess with it more. if i fix it enough i will probably break it! the qc's are rectangular shaped like the originals but smaller. mike at zantech said no matter the size they are all 32,768 freq. + or - a small percentage. it is possible that there is a problem in the chip or something else i guess, since the pcb was more or less a discard when you sent it to me. i really like the dot matrix. i lites up bright and the digits are larger than the bar segs. the fact that it is an early version also adds to the desirability as far as i am concerned. i am wearing it constantly now and have for the time being put the rwr aside . will try to give them all good wrist time tho'. peter

[clockace]

Posted: Tue Dec 30, 2008 1:01 pm Post subject: : dot matrix

--------------------------------------------------------------------------------
sully wrote:
Peter,

Congrats on getting your dot matrix more accurate. Is there anywhere that shows the combinations of sliders/magnets needed to get to the speed adjustments on a MkI/II? I've never had to adjust either of the modules I have, but I've never checked accuracy that much . They are in sync with the clock I used to adjust them to.************* sully, in reply: i am not good at putting inserts in my posts
this is the only dot matrix i have that adjusts that way. i have 2 others w/mkI/II pcb's that have bar segs. they adj with the slider rear mag combo as well. they are well within range as far as accuracy tho'. they do not however show the speed # on the display. each notch also changes speed by 16 sec/yr instead of 8. i have printed some directions for this setting that i got from one of the view topics on the forum. peter

[clockace]

to speed up: turn on and hold the sec/yr switch. turn on and hold the date switch. slide the rear mag left for each step faster. to slow down: turn on and hold the sec/yr switch. turn on and hold the time switch. slide the rear mag to the left once for each step slower. this will not work on the 6 reed modules. you only have the trimmer on those. also w/o the display of speed you can only make an adj, reset the time and watch the results. according to the directions i have each step is 16sec/yr, so it is quite sensitive. if howard can shed more light on this it would be appreciated, but he told me that during the early mkI/II era he was just a small boy and didn't get that much experience on these models. his is primarily on the III/IV's of which i guess he has built 100's. right howard? peter

[retroleds]

I guess I'll chime in. DOn't drive yourself too crazy over it Pete - I have two atomic clocks in my office, and I notice that they get about 4-5 seconds apart during the day, more in summer, and then are virtually dead on to each other when I trndle in here around 6 am most mornings. My theory - they don't really keep good time, they just get reset every night. I'd agree with those that suspect your crystal might be running slow. Why use an old crystal? I'm thinking they were made from natural crystal and were not terribly consist ant compared to a modern crystal with lab grown quartz in it. Mouser sells then with a variety of aging tolerances, the lower the better. I personally burn in a new crystal every few days using a cheap LCD module that has some clips hooked to it - whenever I remember I unclip the current one and clip in another, putting the one I ran into a jar for the next repair. Don't know if it helps but I swear they are more likely to spark up a repair piece. Some brands(Frontier's and early Pulsars in particular) seem to be able to sit for hours before they push the crystal enough juice to fire up.

This has been about the hottest area of the forum lately - carry on!

[clockace]

ed, an atomic clock or watch is nothing more than a quartz timepiece that resets itself using either the signal from boulder, colo. or satellite. so the accuracy during the time between resets can vary altho' minimally as a rule. they also have an inherent error of +/- 1 sec because of the stepper motor that drives the hands. the digital ones are also that way. the display cannot keep up with the accuracy of the qc. i have had customers try to return these things because of this. they really need to get a life! i would think tho' if the watch is fast the qc must be fast. it is the chip that divides the 32768 freq by 2 numerous times until you get a 1 sec pulse. at least that is my understanding of it. omega watch had a qc model that had a crystal oscillator that vibrated at almost 2.5 million in the late 70's or early 80's. it was called a megaquartz. that watch earned a rating of "marine chronometer. wish i had one of those. peter p/s: i think this is the hottest area of the forum because the synchronar has a unique, geeky following.

[ronaldheld]

I own a MC from Omega.

[bucko170]

Peter there is one on ebay right now offered for sale with a 'buy it now' price!

http://cgi.ebay.com/ws/eBayISAPI.dll?Vi ... %26fvi%3D1

Nice watch, nice price also

[retroleds]

Peter, I realize the atomic clock is just a cheap quartz clock...I guess I didn't underscore that as their basic weakness.

Those faster oscillation crystals - yep,probably why so many people consider a common Hughes module one of the more accurate. 786,432 hz vs. 32,768. Requires a few more divides, but that doesn't seem to be an issue. The trimmer comes into the picture too - most trimmers for watches are 16-18 pf, the crystal adds 12.5pf(be sure to buy the correct ones, for a total of 30 pf. Admittedly, I have no idea if that holds true for the Synchronar - hard to draw an inference from the few that have a trimmer in the circuit. Maybe Howard can add something to that line of inquiry(understandably, those were only on the earliest models).

[rewolf]

....The trimmer comes into the picture too - most trimmers for watches are 16-18 pf, the crystal adds 12.5pf(be sure to buy the correct ones, for a total of 30 pf....

Actually, it's a bit different: the crystal does not ADD to the capacitance (in fact, its own [shunt] capacitance is less than 2pF), but it REQUIRES an external LOAD capacitance of typically 12.5pF to be added.
This load capacitance is made of two capacitors, the trimmer being one part of it.
The actual load capacitance is the SERIES combination of both load capacitors - so you need 2 capacitors of 25pF each to get 12.5pF load capacitance (CL = 1/(1/C1+1/C2)). Accounting for stray capacitance (3-10pF on each side) leads to 15-22pF load capacitors.

This holds for a "Pierce" oscillator with "parallel" crystal.

A good description is here (PDF) or here (Wikipedia).

[clockace]

update on dot matrix: last week i had a small disaster with the module. the display went dead. checked the cells, they had 2.6v. what now? decided to change the qc. fired right up. now i am back to where it was. runs 6-8 secs/day fast. trying to slow w/trimmer. so far no change, but i think i may have started with the trimmer turned to the dead zone, so i am in the process of turning it a tiny amt/day to see if it changes at all during a 24 hr period. hopefully there will be some change thru the 360 degree adjustment. i think i was hammering too hard on a clock part i was repairing and ruined the qc that was in there. too bad because that one was keeping near perfect time. peter

[Led-Time]

Have you tried a frequency counter to see what its running at.

[clockace]

that kind of stuff is beyond the equipment i have. what kind of instrument is a frequency counter and where would i find such? or could some electronics shop have that piece of equip.? can a qc be tested before it is installed and powered up? might be too many questions for a single thread, but i am kind of electronically challenged. peter

[retroleds]

This load capacitance is made of two capacitors, the trimmer being one part of it.
The actual load capacitance is the SERIES combination of both load capacitors - so you need 2 capacitors of 25pF each to get 12.5pF load capacitance (CL = 1/(1/C1+1/C2)). Accounting for stray capacitance (3-10pF on each side) leads to 15-22pF load capacitors.

Or in plain english - purchase crystals with a 12.5 pf designation unless you know the circuit to have other specs. in combination with the rest of the circuit.