Author Topic: Odd frequency crystal  (Read 2998 times)

0 Members and 1 Guest are viewing this topic.

Offline RyanG

  • Regular Contributor
  • *
  • Posts: 213
  • Country: us
Odd frequency crystal
« on: April 20, 2018, 07:27:07 am »
I have a few identical boards I am trying to work on, but the crystal oscillator is broken off of 2 of them. I have tried to find the same parts, but I can't seem to find any with the same frequency. They are 23.6428Mhz. The board does video output for some older black and white CRTs. I forgot to take a picture of the boards before I left work, but I believe the crystal is running into a 7404 hex inverter. Is there something that might be close enough in frequency or some other way to get this to work?
 

Offline drussell

  • Super Contributor
  • ***
  • Posts: 1476
  • Country: ca
  • Hardcore Geek
Re: Odd frequency crystal
« Reply #1 on: April 20, 2018, 08:10:16 am »
Bowling scoring display units, I presume?  (AS-80/90 ??)

If you can't scrounge good crystals from bad boards, really your only options are to call someone like Anderson Electronics that does custom frequency crystals and have some replacements made to order or make a new oscillator circuit / digital programmable oscillator somehow and wire that into the circuit as a replacement.

(Well, that, or replace them with something like the Evolution upgrade boards that do HDMI output instead...)
« Last Edit: April 20, 2018, 08:13:54 am by drussell »
 

Offline RobK_NL

  • Frequent Contributor
  • **
  • Posts: 250
  • Country: nl
Re: Odd frequency crystal
« Reply #2 on: April 20, 2018, 09:15:49 am »
Tell us what problem you want to solve, not what solution you're having problems with
 

Offline ogden

  • Super Contributor
  • ***
  • Posts: 3285
  • Country: lv
Re: Odd frequency crystal
« Reply #3 on: April 20, 2018, 09:22:28 am »
Those Epson programmable oscillators are nice indeed, but is there any other way to avoid buying ~240$ SG-Writer programmator to let them run at desired frequency? Thou still could be cheaper than buying custom-cut crystals :)
 

Offline RobK_NL

  • Frequent Contributor
  • **
  • Posts: 250
  • Country: nl
Re: Odd frequency crystal
« Reply #4 on: April 20, 2018, 09:30:46 am »
I specifically linked there, because you can have these programmed by DigiKey. Other vendors may also offer this service.
Tell us what problem you want to solve, not what solution you're having problems with
 
The following users thanked this post: ogden

Offline coromonadalix

  • Super Contributor
  • ***
  • Posts: 2849
  • Country: ca
Re: Odd frequency crystal
« Reply #5 on: April 20, 2018, 10:31:12 am »
maybe this crystal has an frequency divider at the output ???  you may find another frequency to match ???

Is it for a system like this ??? : http://www.evolutionscoring.com/files/EVOLTUION_AS-90_LOWER_MANUAL.pdf
 

Offline Richard Crowley

  • Super Contributor
  • ***
  • Posts: 4310
  • Country: us
  • KE7GKP
Re: Odd frequency crystal
« Reply #6 on: April 20, 2018, 10:48:53 am »
Maxim DS1077 is an inexpensive programmable oscillator with a frequency range of 8.1kHz to 133MHz
 

Offline RobK_NL

  • Frequent Contributor
  • **
  • Posts: 250
  • Country: nl
Re: Odd frequency crystal
« Reply #7 on: April 20, 2018, 11:02:48 am »
Maxim DS1077 is an inexpensive programmable oscillator with a frequency range of 8.1kHz to 133MHz
But with an initial tolerance of ±0.5% perhaps not the most suitable to replace a crystal.
Tell us what problem you want to solve, not what solution you're having problems with
 

Offline ogden

  • Super Contributor
  • ***
  • Posts: 3285
  • Country: lv
Re: Odd frequency crystal
« Reply #8 on: April 20, 2018, 11:06:31 am »
Maxim DS1077 is an inexpensive programmable oscillator with a frequency range of 8.1kHz to 133MHz

It definitely cannot be used as crystal replacement:

±0.5% variation over temp (+25°C to +70°C)
±0.5% initial tolerance


5000 ppm :) Note that crystals usually are 50ppm.
 

Offline mikeselectricstuff

  • Super Contributor
  • ***
  • Posts: 12199
  • Country: gb
    • Mike's Electric Stuff
Re: Odd frequency crystal
« Reply #9 on: April 20, 2018, 11:27:17 am »
does it need to be that accurate ? maybe try the closest you can find on the shelf
Youtube channel:Taking wierd stuff apart. Very apart.
Mike's Electric Stuff: High voltage, vintage electronics etc.
Day Job: Mostly LEDs
 

Offline vk6zgo

  • Super Contributor
  • ***
  • Posts: 5553
  • Country: au
Re: Odd frequency crystal
« Reply #10 on: April 20, 2018, 12:05:25 pm »
does it need to be that accurate ? maybe try the closest you can find on the shelf
I suppose  it depends on how "fussy" it is.

Years ago, we had a "dumb" terminal which used a  5.2MHz ( or thereabouts) crystal.
We replaced it with a 5.5 MHz one, and everything worked normally.

It looks like a discrete crystal, not an oscillator module, so maybe get a close one, & "pull" the frequency with parallel capacitance, series inductance, or a complex arrangement  of L&C if necessary.

 

Offline CJay

  • Super Contributor
  • ***
  • Posts: 3596
  • Country: gb
Re: Odd frequency crystal
« Reply #11 on: April 20, 2018, 12:12:16 pm »
There are plenty of programmamble MEMs oscillators that will replace that crystal (Silicon Labs hve/had a range of them that are pretty good), most of the 'better' distributors will program them for you for a small price, it'd probably be cheap enough to get a few for the price of one custom crystal.

You might get away with a 24MHz crystal...

It's an odd frequency though, it's not a multiple of the 'common' TV frequencies or the 'standard' frequencies I can think of off the top of my head, wonder what it's being used for, have you traced the output?

 

Offline drussell

  • Super Contributor
  • ***
  • Posts: 1476
  • Country: ca
  • Hardcore Geek
Re: Odd frequency crystal
« Reply #12 on: April 20, 2018, 02:08:13 pm »
Maxim DS1077 is an inexpensive programmable oscillator with a frequency range of 8.1kHz to 133MHz
But with an initial tolerance of ±0.5% perhaps not the most suitable to replace a crystal.

does it need to be that accurate ? maybe try the closest you can find on the shelf
I suppose  it depends on how "fussy" it is.

Years ago, we had a "dumb" terminal which used a  5.2MHz ( or thereabouts) crystal.
We replaced it with a 5.5 MHz one, and everything worked normally.

It looks like a discrete crystal, not an oscillator module, so maybe get a close one, & "pull" the frequency with parallel capacitance, series inductance, or a complex arrangement  of L&C if necessary.

It has to be pretty darn close.  The Brunswick AS-80/90 systems' CPU boards operate at (or are divided from, I presume) one of two different frequencies depending on the model and version of either 23.64280 or 25.65504 Mhz and the display boards must match the CPU in use for them to work.

Even those programmable oscillators like the Epson units mentioned above may not be suitable, from the datasheet, emphasis mine:

Quote
PLL- PLL connection

The SG-8002 series contains a PLL circuit and there are a few cases where the jitter value may be increased when this product is connected to another PLL oscillator (cascading connection). We do not recommend this series for analog video clock use and telecommunication synchronization.  Please check in advance if the SG-8002 series jitter is acceptable to your application.  (Jitter specification of the SG-8002 series is max.250 ps/CL=15 pF)
« Last Edit: April 20, 2018, 02:17:11 pm by drussell »
 

Offline drussell

  • Super Contributor
  • ***
  • Posts: 1476
  • Country: ca
  • Hardcore Geek
Re: Odd frequency crystal
« Reply #13 on: April 20, 2018, 02:29:23 pm »
You might get away with a 24MHz crystal...

Doubtful...  Even a 23.5920 probably won't work, though you might be able to modify the circuit to pull it up enough by loading it weirdly.

Quote
It's an odd frequency though, it's not a multiple of the 'common' TV frequencies or the 'standard' frequencies I can think of off the top of my head, wonder what it's being used for, have you traced the output?

It's an internal scoring system, not some kind of broadcast thing so it doesn't have to be the same as "proper" TV standards, just enough that the TV will sync to it.   Systems like that were often rather strange...  :)

Here is the installation manual for a series of modern replacements for that kind of systems' display boards that provide an HDMI output for a modern TV...  Perhaps it will give you some insight into the kinds of units the OP is working on.  Page 7-9 talks about the version for the AS-80/90 system that I believe the OP is working on.

edit:  http://www.evolutionscoring.com/files/EVOLUTION_MANUAL.pdf
« Last Edit: April 22, 2018, 08:18:20 am by drussell »
 

Offline CJay

  • Super Contributor
  • ***
  • Posts: 3596
  • Country: gb
Re: Odd frequency crystal
« Reply #14 on: April 20, 2018, 02:49:32 pm »
Well, if it has to be matched to the CPU board then why not change the crystal in both?

Appreciate it makes them non compatible between different units but from what you're saying if there are two versions already then a third, if clearly marked, won't be too much of an issue?

The SiLabs MEMS oscillators were pretty good, low jitter and good stability, good enough that I think in another thread here one was used to fix a Tek 'scope...

If it *absolutely* has to be the right frequency and has to be a crystal there's always http://www.krystaly.cz/en/ who can custom make to order and aren't expensive.

I think your emphasis in the quote about 'video clock generation' is misplaced, it's a TTL Inverter clock signal, it's not going to be great quality and while the jitter may be an issue with high bit rate telecoms or standards compliant video for broadcast, I reckon it'd be fine for a bowling scoreboard, you said it yourself, " just enough that the TV will sync to it"

They're cheap enough to try.
 

Offline drussell

  • Super Contributor
  • ***
  • Posts: 1476
  • Country: ca
  • Hardcore Geek
Re: Odd frequency crystal
« Reply #15 on: April 20, 2018, 03:08:21 pm »
Well, if it has to be matched to the CPU board then why not change the crystal in both?

You would have to change not only the one on the CPU board but also every display board (both upper overhead and lower console) in the system.  Depending on how many lanes the place has, or if these boards need to be able to be used at multiple locations, that could be problematic.  :)

Quote
The SiLabs MEMS oscillators were pretty good, low jitter and good stability, good enough that I think in another thread here one was used to fix a Tek 'scope...

I expect that any of those programmable oscillators would be good enough, it is just something to be aware of in case of potential problems.  :)

Quote
If it *absolutely* has to be the right frequency and has to be a crystal there's always http://www.krystaly.cz/en/ who can custom make to order and aren't expensive.

I agree completely.  There are several good places to buy custom crystals and they aren't necessarily prohibitively expensive.

Quote
I think your emphasis in the quote about 'video clock generation' is misplaced, it's a TTL Inverter clock signal, it's not going to be great quality and while the jitter may be an issue with high bit rate telecoms or standards compliant video for broadcast, I reckon it'd be fine for a bowling scoreboard, you said it yourself, " just enough that the TV will sync to it"

They're cheap enough to try.

Indeed!

I honestly don't know enough about these things to know for sure.  I've never had to work on one personally but helped someone with a similar type of system on another forum solve their problems several years ago.  I don't even recall what system it was, but I ended up doing a fair bit of research and digging into some of their idiosyncrasies at the time...
 

Offline drussell

  • Super Contributor
  • ***
  • Posts: 1476
  • Country: ca
  • Hardcore Geek
Re: Odd frequency crystal
« Reply #16 on: April 20, 2018, 03:36:14 pm »
Actually, on closer inspection of these systems, it seems that each lane pair has a separate CPU board in the lower console (two lanes per CPU) rather than one master unit for a larger bank, so it might be possible to change the frequency slightly on all the "stuff" in one lane pair to another, slightly different frequency, at the expense of parts interchangeability.
« Last Edit: April 20, 2018, 03:38:25 pm by drussell »
 

Online edavid

  • Super Contributor
  • ***
  • Posts: 3012
  • Country: us
Re: Odd frequency crystal
« Reply #17 on: April 20, 2018, 03:56:37 pm »
I also think that 24MHz crystals would be worth a try, but otherwise here are some cheap 23.69MHz crystals: http://halted.com/commerce/ccp12770-crystal-23-690-mhz-28hc-1829-1384.htm

That's only 0.2% off - hard to believe it wouldn't work.
« Last Edit: April 20, 2018, 10:26:39 pm by edavid »
 

Offline RyanG

  • Regular Contributor
  • *
  • Posts: 213
  • Country: us
Re: Odd frequency crystal
« Reply #18 on: April 20, 2018, 05:13:09 pm »
I am working on an AS90 scoring system as drussell has pointed out. To upgrade the entire system from this old stuff is about $250k to buy brand new stuff, so the boss isn't going to be doing that. Even the upgrade to the Evolution scoring, which still uses a lot of the old stuff is fairly expensive. I did think to try the 24Mhz crystals. I ordered them not long ago, and I am just waiting for them to come in. The problem with the system is that a lot of the stuff is obsolete. It was as if they had things made specifically for their system just to make it harder to work on for people like me. Brunswick has completely dropped support for the system now because it is from the early 90s. In general the system still works well. I got most of our CRT monitors to run great again thanks to james_s here on the forum. Each pair of lanes has 17 pcbs in them for the scoring system, and I have been trying to learn to repair each of them. I will bring one of the bad boards home and take some pictures if that would help get an idea if something could be easily done to modify the boards to work.
 

Offline ogden

  • Super Contributor
  • ***
  • Posts: 3285
  • Country: lv
Re: Odd frequency crystal
« Reply #19 on: April 20, 2018, 09:18:18 pm »
does it need to be that accurate ? maybe try the closest you can find on the shelf

Would you rely on just trying? It easily could be so that when you try - it works, but just because it barely fits into one end of tolerance window. After few months situation can change and you can be screwed with your "I fixed it".
 

Offline drussell

  • Super Contributor
  • ***
  • Posts: 1476
  • Country: ca
  • Hardcore Geek
Re: Odd frequency crystal
« Reply #20 on: April 20, 2018, 09:39:49 pm »
Because all the boards on these systems use that weird clock frequency, I have a funny feeling it isn't going to work correctly, though without more thorough knowledge of the entire system or schematics or something, obviously I can't say for sure....

I just have the impression that it is used to derive all sorts of things including whatever serial communications system these things use to talk between modules and to the front desk display terminal thingy that enables the lanes by game or time or whatever.  I still think that you would probably have to change every crystal on all the boards in the entire system, and even that is only going to work if all the timings in the system are generated off these crystals and don't have to match anything else anywhere in the system.

It's a good question, to be sure!  :)
 

Offline RyanG

  • Regular Contributor
  • *
  • Posts: 213
  • Country: us
Re: Odd frequency crystal
« Reply #21 on: April 20, 2018, 11:21:32 pm »
I'm at work now, and I drew up the relevant part of the schematic because Brunswick does not provide them. They say the schematics are still under NDA. I didn't draw the other connections to the 7404 because I figure they have nothing to do with this circuit, as I assume this circuit just outputs to another pin of the 7404.
 

Offline TheSteve

  • Supporter
  • ****
  • Posts: 3164
  • Country: ca
  • Living the Dream
Re: Odd frequency crystal
« Reply #22 on: April 21, 2018, 03:04:19 am »
There used to be lots of places that would make a crystal to any frequency you wanted. Perhaps most are gone. I know I've had many custom crystals made. If you didn't know the exact characteristics of a crystal you wanted copied you would just send the original to them and they would measure and duplicate. It wasn't too expensive either, less then $20 bucks a crystal. There must still be one or two companies left who can do it.
VE7FM
 
The following users thanked this post: ogden, drussell

Offline drussell

  • Super Contributor
  • ***
  • Posts: 1476
  • Country: ca
  • Hardcore Geek
Re: Odd frequency crystal
« Reply #23 on: April 21, 2018, 09:00:15 pm »
There used to be lots of places that would make a crystal to any frequency you wanted. Perhaps most are gone. I know I've had many custom crystals made.

I have also used custom crystals from a couple different places before with no issues.

Quote
If you didn't know the exact characteristics of a crystal you wanted copied you would just send the original to them and they would measure and duplicate. It wasn't too expensive either, less then $20 bucks a crystal. There must still be one or two companies left who can do it.

Indeed.  There are still several in existence, perhaps some others here can share specific ones in their geographic area or specific experience WRT pricing, lead time, etc.

Most of these kinds of places, if you contact them and say you need a general purpose, low cost, non-ultra-precision-spacecraft-grade crystal, can make you a few for a reasonable price, given the custom nature of the job, though it probably varies significantly between places depending on what their major market currently is so it probably pays to shop around.  :)

To start:

Lap-Tech  (Ontario, Canada)  http://www.laptech.com  (289)481-2019
(also available from TFC in the UK  www.tfc.co.uk, they're part of Lap-Tech)

Anderson Electronics (Pennsylvania, USA)  http://www.aextal.com  (814)695-4428

FMI (California, USA)  http://www.frequencymanagement.com  (714)373-8100

Transko  (California, USA)  https://www.transko.com/  (714)528-8000

FOX (Florida, USA)  https://www.foxonline.com/

EuroQuartz  (UK)  http://www.euroquartz.co.uk/  +44(0)1460230000

QuartzLab (UK)  http://www.quartslab.com/

Krystaly (Czech Republic)  http://www.krystaly.cz/en/
 

Offline evb149

  • Super Contributor
  • ***
  • Posts: 1666
  • Country: us
Re: Odd frequency crystal
« Reply #24 on: April 22, 2018, 12:13:09 am »
I would make sure it is the crystal that is broken as opposed to the oscillator circuit that is not oscillating.  The latter could be caused by gain changes, IC failure, PSU voltage changes, temperature changes, ESD, et. al. even with a good crystal.
It isn't uncommon for poorly designed crystal oscillator circuits to sometimes work and sometimes fail to work just due to changes due to component aging, temperature, et. al.  Generally the capacitive load and drive level vs loss may need to be adjusted to optimize the reliability.

Crystals usually don't often fail though they can over time particularly if abused by excess mechanical abuse, excessive drive levels, et. al.

My ideas to fix  the circuit in the problem units would generally flow like:

A: Get an exact replacement crystal, if able to be sourced economically either standard or custom.

B: Try a programmable oscillator that is close and see if it works.

C: It sounds like there is worry about changing to an off frequency crystal because the system has multiple units that must be synchronized to a given frequency.  That's a disadvantage for substituting an inexactly matched crystal like a programmable oscillator.  But you could turn it into an advantage by realizing that you HAVE one or more other working oscillators with exactly the right frequencies "nearby" in the same system.  So you could use a couple buffers / drivers, pick off the master clock from one good unit, buffer it by a video or LVDS driver, feed it down some CAT-5 or coax cable or fiber or something suitable, receive it with the appropriate line receiver, add a 74LS04 or whatever suitable buffer at the end (and probably a little ESD protection on both ends), and there you go.  Exact synchronization with no hard to find parts.

D: Rumor has it there are cheap DDS and PLL modules out there as ready made module level boards.  So get one that is crystal controlled, program it to output your desired frequency, feed it through a TTL buffer into the unit, wire it together to the defective unit for power and signal connections, done, pretty stable, pretty exact frequency.  Probably not cheaper that one would HOPE to be able to source a replacement crystal for, but cheaper than you MAY be able to source one, and pretty easy.  I imagine there are some cheap FPGA or clock synthesizer IC evaluation boards that could also be programmed trivially to synthesize and output this frequency.  You'd probably do as well or better with a programmable oscillator component but if those don't seem to have the frequency / temperature tolerances you want then I guess a board / module level clock synthesizer could have better results in frequency tolerance and stability though not ideal jitter (which probably doesn't really matter here much) compared to a true crystal.

E: It probably wouldn't be hard to build or buy some little huff puff type VFO and set it to the correct frequency.  Someone is probably selling a little board / kit / something.  I'd think a clock synthesizer IC board would be more readily available and directly applicable though.

Most old cheap crystals and rudimentary oscillators aren't going to be reliably accurate to more than 20ppm so at 23MHz that's 460Hz or so.  So if you get anything that's approximately that close in frequency and reliable to stay approximately in that range it ought to work.  Even twice that tolerance maybe 50ppm wouldn't be surprising for old cheap consumer electronics to accept as component manufacturing tolerances...

If you're inclined to mess with PLLs, it is 400 * 59107 Hz... not that hard to generate any number of ways these days.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf