Honeywell quartz clock component

[sofTest]

Re: Honeywell quartz clock component

You're right. With the kind of inaccuracy you're getting, it's way beyond the fine tuning of that trimmer.

I am wondering about the supply to the chip on that board. I assume there is a zener diode somewhere on the board, probably between the unknown component, and (what looks like) the battery. If it is a rechargable NiCad and is no longer holding a charge, then there's a good chance that one or more cells have shorted. This may be pulling the supply volts down, causing the chip to operate erratically. Just a thought.

EDIT...btw, as Krankshaft has pointed out, this clock uses a transformerless mains supply. Be very careful if you take any readings while it is powered up!

You are right, the diode (tests OK) is between the battery and the brown capacitor. Would it harm just removing the battery? Or is any other modifications needed then?

[sofTest]

Re: Honeywell quartz clock component

variable capacitor is open circuit and yes that can change the tuning so much. Replace it with similar from a old clock. Not rare.

The type of components design tells me this is far older than 20. Looks like 30 years or older. Very early quartz clock.

Cheers, Wizard

I've got no old clocks around, and none of the boards I've got with crystals on them got a trimmer. If I were to buy a new one, what capacitance range do you recommend? Does this have anything to do with load capacitance of the crystal?

A new clock costs USD 200+ (probably more here in Norway, haven't received the invoice yet), so a couple of dollars for components for testing will be nothing.

Yes, you are probably right about the age.

[Wizard]

Re: Honeywell quartz clock component

Try these you find with close same diameter. Should work. Not terribly critical.

Cheers, Wizard

[kc8adu]

Re: Honeywell quartz clock component

the glue on the trimmer is likely the problem.the stuff goes conductive and corrosive.
yes i do rf work.no big deal to me.have 2 tm-742's apart in front of me now.i overhaul a lot of radio gear.

[Pyr0Beast]

Re: Honeywell quartz clock component

NiCad battery might act the same as a Zener diode. 2 in series do give a good 2.5V reference . Might be replaced with NiMh, however, they don't give as stable reference.

Clock drifting behind might be also due to insufficient power to drive the clock-gear-motor, or dust in the gear. Some 'water-type' oil would do good to the gears after such a long time.
You can see that if the pointer for seconds sometimes just twitches back and forward.

[Wizard]

Re: Honeywell quartz clock component

Always disassemble, clean out and reassemble, oil. Can't just oil and forget, the gum is still there.

Cheers, Wizard

[Radio Fox]

Re: Honeywell quartz clock component

You are right, the diode (tests OK) is between the battery and the brown capacitor. Would it harm just removing the battery? Or is any other modifications needed then?

Can you read the markings on the zener diode? Then we will know what voltage the chip is working at. I would leave the battery for now.

I was trying to read the markings on the I.C. Does it say 1115A 158S. The photo isn't too clear, but that's what it looks like. I tried Googling it, but could not find anything.

In the UK, Maplin sell those type of trimmers, but I don't think Digikey sell them. The green body one is 2 - 22pF, but it depends on the diameter as well.
http://www.maplin.co.uk/module.aspx?moduleno=460

[Wizard]

Re: Honeywell quartz clock component

Zener is last thing you want in a battery powered clock as regulator, I think zener is for protection or to cut spikes from pulsing the coil for the motor.

Cheers, Wizard

[sofTest]

Re: Honeywell quartz clock component

NiCad battery might act the same as a Zener diode. 2 in series do give a good 2.5V reference . Might be replaced with NiMh, however, they don't give as stable reference.

Clock drifting behind might be also due to insufficient power to drive the clock-gear-motor, or dust in the gear. Some 'water-type' oil would do good to the gears after such a long time.
You can see that if the pointer for seconds sometimes just twitches back and forward.

Always disassemble, clean out and reassemble, oil. Can't just oil and forget, the gum is still there.

Cheers, Wizard

The gear is encapsuled, and clean. The clock doesn't have a seconds pointer either.

Can you read the markings on the zener diode? Then we will know what voltage the chip is working at. I would leave the battery for now.

I was trying to read the markings on the I.C. Does it say 1115A 158S. The photo isn't too clear, but that's what it looks like. I tried Googling it, but could not find anything.

In the UK, Maplin sell those type of trimmers, but I don't think Digikey sell them. The green body one is 2 - 22pF, but it depends on the diameter as well.
http://www.maplin.co.uk/module.aspx?moduleno=460

The zener is axial, yellow casing with the number 2 on it, and the manufacturer ITT. If there are other markings, it's obscured by the battery.

The IC reads 1115A 1585, and the logo is an e. I've tried to Google it before I started this thread, but nothing came up. Probably to old.

Actually, DigiKey does carry trimmers. Put in trimmer in the search field, and variable capacitors comes up as one option. The one on the board is 12 mm, but that is not available. I guess a 10 mm could be accommodated, so SG3007-ND would be a close match. I would need bottom adjustment also.

Also, I've reading up on crystal accuracy. The replacements I've found on DigiKey got a frequency tolerance of +/- 30 ppm or +/- 50 ppm. According to the formula found on this page, that would mean a max drift of 78 or 130 seconds a month. For the duty this clock is doing, neither would be a problem. I've also found a local source carrying a +/- 20 ppm crystal. So I'm thinking, why not put in a new crystal, and do away with the trimmer?

This clock is probably around 30 years old, so the quality of crystal manufacturing might not have been as good as today (I'm guessing). So that might have necessitated a trimmer. I've been looking at various boards of today with crystals, and found none with trimmers. Due to better crystals?

What are your thoughts? If you agree, how would you bypass the trimmer? I'm charging my Sony compact camera, so I can provide better macro pictures than my mobile phone camera later, if you like.

Thanks.

[Radio Fox]

Re: Honeywell quartz clock component

...The zener is axial, yellow casing with the number 2 on it, and the manufacturer ITT. If there are other markings, it's obscured by the battery...

There should be other markings to indicate it's voltage. The reason I asked is that I would be tempted to replace the electrolytic you had previously removed, & power it up again to measure the supply to pin 1 of the chip.

...The IC reads 1115A 1585, and the logo is an e. I've tried to Google it before I started this thread, but nothing came up. Probably to old...

Yep, I couldn't find anything either.

...Actually, DigiKey does carry trimmers. Put in trimmer in the search field, and variable capacitors comes up as one option. The one on the board is 12 mm, but that is not available. I guess a 10 mm could be accommodated, so SG3007-ND would be a close match. I would need bottom adjustment also...

Ahh..that's why I didn't see them. When I searched for them I filtered for "through hole" mounting, and "top adjustment". All the ones I have ever bought have been top adjustment only. The digikey ones are also bottom adjusting.

...Also, I've reading up on crystal accuracy. The replacements I've found on DigiKey got a frequency tolerance of +/- 30 ppm or +/- 50 ppm. According to the formula found on this page, that would mean a max drift of 78 or 130 seconds a month. For the duty this clock is doing, neither would be a problem. I've also found a local source carrying a +/- 20 ppm crystal. So I'm thinking, why not put in a new crystal, and do away with the trimmer?

This clock is probably around 30 years old, so the quality of crystal manufacturing might not have been as good as today (I'm guessing). So that might have necessitated a trimmer. I've been looking at various boards of today with crystals, and found none with trimmers. Due to better crystals?

What are your thoughts? If you agree, how would you bypass the trimmer? I'm charging my Sony compact camera, so I can provide better macro pictures than my mobile phone camera later, if you like.

Thanks.

The problem with removing the trimmer from the circuit, is that the quoted ppm figures only apply if the correct load capacitance is present on the circuit the crystal is connected to. There is no way to determine this exactly, hence the trimmer. If you did want to get rid of the trimmer, you would have to replace it with a low value fixed capacitor (say 5p6 - 10p).

Zener is last thing you want in a battery powered clock as regulator, I think zener is for protection or to cut spikes from pulsing the coil for the motor...

Well it looks like the battery is only for backup, the usual supply is from the transformerless mains circuit which needs the zener to regulate the supply to pin 1 of the chip.

[sofTest]

Re: Honeywell quartz clock component

Updates (been busy elsewhere):

The IC is an Eurosil 1115A with properties (full datasheet not found):
http://www.mikrocontroller.net/topic/60893 -> Google translation

Also, I've found two threads in German (this is a German made clock), with same or similar design:
http://www.mikrocontroller.net/topic/153596 -> Google translation
http://www.mikrocontroller.net/topic/87453 -> Google translation

Both concluding with the most common failure is a bad battery. Since the battery on mine is dead, the battery issue need to be resolved before any more testing can be done.

I've got three options:
1. Get hold of an 1.2V 14x17 mm NiCd battery with solder pads (not easy, but possible, in Norway (any more))
2. Use an 1.2V NiMH battery. Easily obtainable.
3. In the German threads, there is a reference to use a Goldcap (I'm guessing a Panasonic Gold Cap backup capacitor), with a LED as a limiter. I can't easily get hold of backup capacitors of a specification of a few uFs, with the space restrictions of my clock, so not very feasible.

So my question is, what downside would there be to use a NiMH battery?

Thanks for your patience.

[Krankshaft]

Re: Honeywell quartz clock component

Technically those are called super capacitors gold cap is just a brand name.

Super caps last longer then batteries, don't contain toxic metals and charge rapidly. Unlike standard lytics they are in the farad range while filters are in the uF range.

They're usually used for their long life for example one of my digital thermostats has a supercap in it to retain the clock and programmable settings during a power failure.

To use a Nimh you would need to assess it's output voltage and current. Is it printed somewhere on the top or bottom? If it is everything will be fine.

[sofTest]

Re: Honeywell quartz clock component

To use a Nimh you would need to assess it's output voltage and current. Is it printed somewhere on the top or bottom? If it is everything will be fine.

Do you mean the new or old one? Of the old one, nothing is known, but from the links I found, it's 1,2V NiCd, probably 150-180 mAh. If new, two NiMH candidates that fits physically are:
http://www1.elfa.se/elfa~no_no/go.js...no&i=69-313-07
http://www1.elfa.se/elfa~no_no/go.js...no&i=69-313-15

Full data sheet found under the tab Dokumenter.

I'm also wondering about Pyr0Beast comment:

Might be replaced with NiMh, however, they don't give as stable reference.

[Krankshaft]

Re: Honeywell quartz clock component

I don't know what he meant by stable reference but we're dealing with 30 year old tech here if that's what was there then that's what can go back without any problems. I know it's not Ni Mh because they didn't even exist in consumer products till 89.

If you have a datasheet for the IC I'm sure it will specify it's supply voltage an extra volt probably won't matter. As for current you'll have to defer to the posts in those other forums or the datasheet of the IC for that.

[stj]

Re: Honeywell quartz clock component

supercap's can leak - they always do in the xbox.

[Pyr0Beast]

Re: Honeywell quartz clock component

Goldcap would be a good choice. I got a hold of 10F 2.5V ones. They work excellent and are a good replacement for batteries. However, they don't come even close with their capacitance. Also, they do not provide any voltage limiting and any excess voltage would lead to their internal faliure (open circuit). Also, they do not 'waste' any energy so current regulating circuit would probably produce higher than normal voltage.
Supercaps can be killed, esp. in VCR's.

You could zap the current battery (to burn any filaments inside which are causing the short) and charge it with a couple of mA untill it is full. Unsolder it first of course. NiCad's can last 'forever' if they are of good quality.

I'm also wondering about Pyr0Beast comment:
More of a warning. NiMh would work (theoretically), but has half the life of NiCd

supercap's can leak - they always do in the xbox.
Those are the ordinary caps.

[stj]

Re: Honeywell quartz clock component

check the end of the nicad to see if it's flowered - white powder escaping.

that's alkaline and very corosive, so if it's powdered you need to bin it fast.

[Pyr0Beast]

Re: Honeywell quartz clock component

KOH - alkaline is corrosive. However once it has contact with air it will absorb any CO2 in it and neutralize itself into white powder which is harmless.

[kc8adu]

Re: Honeywell quartz clock component

i have seen them leak fast enough to eat up a board.
had a journey video game where the battery leak got under and into a 12 pin molex that carried among other things the remote sense lines for the power supply.
the guy i got it from replaced every part of the power supply.+5 would still drift.the game would run fine if you chased it with a vom and screwdriver.
desoldered the molex and cleaned up the mess once i traced it down to a rs issue.
and hardly a week goes by where i dont see a newall dp7 with among other issues a leaking nicad.they can make quite a corroded mess.

[Radio Fox]

Re: Honeywell quartz clock component

Looks like the guys on the German forums came to the same conclusion as my earlier post. A transformerless PSU cannot supply much current, so it doesn't take much for the supply volts to drop if a component goes faulty. Although I initially assumed that the battery consisted of a series of button cells, it looks like it is just a single 1/3 AA cell. In which case, the cell cannot be short circuit, as it would have stopped the clock completely.

I dont think a supercap is suitable for this situation. Firstly the capacity of a supercap is only a fraction of that of a battery. A 10H supercap @ 1.2V has an energy capacity of just 7.2 Joules. A 250mAH battery has an energy capacity of 1080 Joules. Looking at a few dimensions, it looks like a 10H supercap might be physically too big anyway. A supercap of similar dimensions to the existing battery would be about 3.3H (just 2.52 Joules).

Also, a supercap would discharge at an exponential rate. As soon as the circuit started to draw current, the voltage would drop quite quickly, and continue to drop. As we don't know the supply limits for the chip, it could stop working very quickly once it started to discharge the supercap.

I would opt for the NiMH battery, that's if you actually need a backup battery. I cant see why the circuit shouldn't work without it, as the zener is directly across the supply pins of the chip.


btw...why has my sig stopped showing?