Fancy 70.5367 MHz crystal - hard to find. Advice needed.

[aluck]

I have a nice car alarm with a two-way remote. It works at 433.92 MHz and uses Philips/NXP UAA3220TS IC for signal transmission.

As it ages, it started to fail. It used to work at the range up to 1 km, now it's about 30 meters at most. The reason is aging quartz crystal. According to the datasheet http://www.nxp.com/documents/data_sheet/UAA3220TS.pdf it is a 70.5367 MHz quartz (actually, it looks like a standard 32678 Hz clock cylinder-shaped quartz).

I wonder if it is possible to buy a replacement or use some kind of a workaround (except for modifying whole transmission module to use 2nd or 3rd harmonic). Any suggestions?

[grumpydoc]

There are plenty of companies who will do you a crystal cut to any frequency.

Probably not much help though - I don't know what their MOQ is but I doubt that "one" will be what they are thinking of.

I suspect your best bet might be to find something else that uses the UAA3220TS and look for a scrapper.

[aluck]

I suspect your best bet might be to find something else that uses the UAA3220TS and look for a scrapper.

Those crystals age within 3-5 years, so "something else" should be rather new (read: not cheap).

[aluck]

Maybe you could use a capacitor to 'pull' the oscillation frequency one way or another, or even take one of the compensation capacitors off to see what happens.

The hardest part is "to see", it's 433.92 MHz. My scope is up to 100 MHz.

[Smokey]

Depending on how accurate you need the frequency, you can use a DDS to generate any arbitrary frequency.  I'm only familiar with the function gen DDSs but I know they make ones specifically for clock generation for PLL and such. 

http://www.analog.com/en/rfif-components/direct-digital-synthesis-dds/products/index.html

Once again, depending on how much hacking you are willing to do, and how critical the application is, you should do a google search for "crystal penning".  If nothing else, it's a cool concept.

[aluck]

Do you suggest installing DDS into keychain remote?

[Fraser]

Hi,

You state that the UAA3220TS is used for transmission ? Its a receiver IC and not a transmitter 

I am very surprised to hear that the crystal is the cause of this issue. In my experience physical damage to the transmitter fob due to impact and detuning of the simple antenna circuit are the chief culprits for range reduction. Consider your scenario......

On frequency and new = 1km range
Off frequency and old = 30m range..... This does not obey normal laws of simple 433MHz TX/RX systems. If the TX frequency has deviated significantly your receiver should actually ignore the transmission unless we are talking about the signal just entering the skirts of the RX passband filtering and they are pretty broad, if fitted at all. If the receiver frequency has drifted a similar effect would be experienced, not range reduction but rather loss of control. 433MHz receivers are pretty frequency tolerant beasts anyway.

This would be my suggested scenario

On frequency and new = 1km range
On frequency and old but TX antenna circuit detuned or component impact damaged = 30m range

This fits my experiences with such devices, and be aware that dropped keyfobs can suffer cracking of the surface mount components and also dry joints.

You should first consider the ERP of your fob and whether it is still radiating a decent signal. You really need a spectrum analyser to see what is going on here but a simple diode detector type RF sniffer will indicate the effect of changes in the TX aerial circuit resonance.

Don't condemn that crystal just yet.

You should also consider whether changes have occurred in the receiver circuit, such as water ingress effecting the RF front end efficiency or aerial damage/detuning. The receiver IC front end could even be damaged but I would expect total loss of control in that scenario.

Fraser

[aluck]

Thanks for that thorough message, Aurora!

First, it is a two-way fob, both transmitter and receiver. I was incorrect when I named that IC as a transmitter. Nevertheless, as it is a two-way, and it uses KeeLoq crypto, if a receiver on a remote has a high failure rate (i.e. it fails to decode a message 80% of the time), one will have trouble both receiving and sending messages.

By the way, the alarm module in the car uses the same IC.

Second, as I do not have a top-range oscilloscope or spectrum analyzer, I bought an identical fob. The new one worked fine (actually, it still was not a full 1 km range, but remember - I did not replace the module in the car! and it has the same IC and the same crystal). After checking the new fob, I simply started replacing components in the old one - one by one. Then each time I checked whether it works better, worse or about the same. The only significant change was after switching that 70.5367 MHz crystal. Actually, the new fob (with the old crystal) doesn't have a full range now, and the old fob (with the new crystal) works just fine.

Do you see any flaws in my reasoning?..

[T4P]

http://www.soasystem.com/eng/rig433/ Rough measurement

[Fraser]

Aluck,

Ah the full diagnostic story explains all 

I see no failure in your logic as complete component exchange is a pretty certain method of finding a faulty component or PCB. This is the first time I have heard of a crystal causing this sort of issue in a fob though. Normally they are shattered after a drop and no longer oscillate.

Sorry I can't help with a new crystal but if they are used in keyfob transceivers I would hope that they are a COTS product rather than a custom component. As others have said, if you can find a keyfob board that uses the same Philips IC there is a good chance that the same crystal will be used. I recently bought a whole load of new fob transmitters with the same encryption as your unit. I will see if the same IC and crystal is used.

Regards

Fraser

[aluck]

Aurora, a lot of people, including several of my friends, are experiencing the same problem with this brand of alarm. They decay over time (possibly due to vibration and temperature - it's not uncommon to have -36C in cabin at winter and +60C at summer).

Please let us know if you have the same crystal in your fobs.

[codeboy2k]

30m range seems like good enough anyways for a car alarm. 

Why would you want to unlock your car from 1km if you can't see who might be standing beside it at that moment, hear the chirp, look around see no one nearby and steal your car

 

[sorin]

http://img01.taobaocdn.com/bao/uploaded/i1/T14dDBXdNiXXboS2gZ_031926.jpg_310x310.jpg
you can buy it on taobao
http://item.taobao.com/item.htm?spm=a230r.1.10.1&id=15571322035
its chip

[aluck]

Why would you want to unlock your car from 1km if you can't see who might be standing beside it at that moment, hear the chirp, look around see no one nearby and steal your car

1) 1 km in a field means about 200 m in a town
2) Sometimes I use it to check the temperature in the cabin. Sometimes to crank up the engine in winter.

[Fraser]

aluck,

I have checked the remote key fobs that I have and none contain that crystal as they use a custom encoder IC and the encrypter chip. Same idea, different chip set.

Sorry I can't be of more help on this.

Fraser

[aluck]

Aurora, Sorin, thanks!

I tried to contact that guy from TaoBao to check if he is able to put it on ebay. Will update you on the progress.

[aluck]

Update: that guy from taobao does not respond on e-mails.

Yesterday a friend of mine suggested replacing that crystal with a programmable oscillator like those: http://www.ecliptek.com/oscillators/default.aspx?FilterDeviceType=MO

Not exactly those, as I need 1.5 VCC, but something like that.

He said they are kind of cheap ($2/piece), but they need to be programmed, and each brand has it's own unique programmer. And those programmers are rather expensive.

Do you, guys, have any experience with that programmable thing?

[NiHaoMike]

30m range seems like good enough anyways for a car alarm. 

Why would you want to unlock your car from 1km if you can't see who might be standing beside it at that moment, hear the chirp, look around see no one nearby and steal your car

 

You could rig it to a relay to blast the fan on full in fresh air mode for a few minutes. Great for the summer...

[krenzo]

Do you, guys, have any experience with that programmable thing?

I've ordered programmable oscillators from DigiKey before.  You pick out the type you want (case, freq range, etc), and you tell them what frequency you want it programmed to which they will do before shipping.

[amyk]

Update: that guy from taobao does not respond on e-mails.

They do not respond unless you are asking a direct question about the product, and asking them to sell it somewhere else is definitely going to get ignored. Most of the sellers there are like that.

[aluck]

They do not respond unless you are asking a direct question about the product, and asking them to sell it somewhere else is definitely going to get ignored. Most of the sellers there are like that.

Personally, I don't care for sellers that are hard to contact. Don't worth the effort.

[PA0PBZ]

I think you should find out what is wrong with the crystal: is it off frequency or did it stop oscillating at all?
I have no experience with crystals going off frequency that much, but then I live in a climate that is always between -15 and +30C.
So get hold of a frequency counter or just a scanner/receiver and find out what it is doing.
It's perfectly possible that you will be able to get it back on the right frequency with a coil/capacitor and if you're lucky it might even have  stopped aging...

[aluck]

I think you should find out what is wrong with the crystal: is it off frequency or did it stop oscillating at all?

Because it does work sometimes, it is definitely oscillates once in a while.

It's rather hard to measure it's frequency. That crystal is coupled with an inductance in sequence, several capacitors and a resistor. Adding probe's capacitance messes things up. Couldn't see anything else than some harmonic around 300 kHz even on a good crystal. It's a typical schematic from this appnote: http://www.nxp.com/documents/application_note/AN00036.pdf at page 8. Any ideas on how to measure this guy's frequency?