Understanding My Crystal Oscillator

[EMET1970]

I have an obsolete, proprietary crystal oscillator. It is a 200MHz, 10 pin, SMT component. The number on the unit is 200N1. I cannot find another C.O. like it in size, number of pins or footprint. What I don't understand is 9 of the pins are grounded. Only one pin is used and it obviously puts out the frequency. How does that work? We have 100s of units we repair with this C.O. in it and they are starting to fail due to age and heat.

What we would like to do is find some comparable C.O., mount it on a small PCB, which in turn would be mounted to the footprint of the old C.O.

I am an electro-mechanical engineering technician, not an engineer. Our EE on staff has no clue, either. Any info would be much appreciated!

 

[MacGyverS2000]

Uhm... if it's an oscillator, it had better have a power leg...

4 pins typically have power, ground, out, and N/C. Sounds like your 10-pin has (likely) power, out, and 8 grounds.

Dan - Owner

http://www.Hi-TecDesigns.com

 

[EMET1970]

Trust me, 9 pins go directly to ground. I've been working with (and scratching my head over) these for several years.

 

[EMET1970]

Attached is the oscillator circuit schematic

 

[EMET1970]

Attached is a picture of the PCB where the crystal sits.

 

[zappedagain]

In that circuit I'd say it is a resonator that is tuning the circuit around it. Have you tried measuring the impedance between the 'two' pins?

 

[Comcokid]

Like MacGyverS2000 indicated - unless it has a power pin, it is not a oscillator. Your circuit doesn't show a DC bias being fed in the non-ground pin either. I didn't try to analyze the schematic you posted in terms of the overall circuit (Colpitts, Pierce, etc), but X1 can only be a crystal, or other very-high Q resonant device. The multiple ground pins must be necessary to keep inductance of the device to ground as low as possible, especially needed if it is really a 200 MHz device. Crystals at this high a frequency are uncommon, and usually tempermental. Typically a circuit will use a lower frequency and feed a multiplier circuit from which a higher harmonic is extracted and filtered.

The other possibility is that X1 is a filter. But I'm not aware of filter elements like this applied as a 'shunt-type' arrangement. Filters are usually a in/out type.

 

[EMET1970]

I think it may be an OCXC. There is a heating/cooling system built into this circuit that affects 80% of the PCB. The oscillator's frequency comes up to 200MHz as the board heats up, then stays as the temperature is regulated to remain at a certain point. Any thoughts?

 

[MacGyverS2000]

The heating is likely to ensure a stable frequency over a wide frequency range, but heating in and of itself does not magically make the component work, it's just a stabilizing function. I came across some of your threads on other forums...you cannot simply replace this with a VCXO and expect it to work the same... there's a reason they went with temp-controlled, accuracy.

As has been stated here and on the other forums... this simply cannot be an oscillator if your schematic is correct. It makes no sense to tie a crystal directly to ground, which leads us to think it's is an oscillator... which means your schematic is wrong.

Something needs to get fixed before we can offer any further useful info.

Dan - Owner

http://www.Hi-TecDesigns.com

 

[EMET1970]

Then we're at an impasse. We've used this schematic for years. It's been traced and retraced, corrected - with the exception of the E, B, C pin-outs. It is what it is. All I'm seeing on these forums is the same thing the EEs who've looked at this board in person have said - it doesn't make sense... yet it works. (Again, they weren't questioning the schematics. They were questioning the board design.) Nobody local can pinpoint it so I came here. I know there is someone out there who can look at this type of circuit and CO and know exactly what's going on. It's just a matter of finding that person. I appreciate everyone's time.

 

[zappedagain]

I don't remember my oscillator topologies but how about this for a grounded crystal circuit:

http://hem.passagen.se/communication/txo.html

Z

 

[iop995]

Maybe with same modification there are some chance to use such device:

http://www.maximintegrated.com/datasheet/index.mvp/id/5677/t/al

 

[IRstuff]

The bottom line is that the schematic does not support your device being a full-blown crystal oscillator. If it were somehow a crystal oscillator on its own, the surrounding transistors and components would be unnecessary, since the only possible purpose would be that they serve as buffer amps. The fact that they are not configured as buffer amps says that your device is just a plain a crystal; the only question is whether it's operating in series resonance at 200 MHz.

The snippet of schematic does not have the portion needed assess how a standard crystal oscillator might fit into your circuit. The section around IC5 is needed to do that, or, at least, some information about IC5 and where R60 ends up (kind of looks like it's grounded). In any case, R59 is the only possible way an oscillator signal can get to the rest of the board.

Nonetheless, the circuit doesn't make much sense, particularly regarding Q13, since its collector and emitter are shorted by C59? None of the other transistors, Q14, or Q18, are wired for power output through R59. R59 is essentially connected only to the base of Q18 through a capacitor C72, and so R59 appears to be an input, not an output of this circuit. The purpose of L39, which is tied to Vcc and appears to supply power to the transistors and crystal doesn't make much sense, either.

TTFN
faq731-376

Need help writing a question or understanding a reply? forum1529

 

[ijl]

My guess is that your SMT component is an inverted-mesa crystal. That is, a quarz crystal with thick edges and a thin resonating area in the middle.

My second guess is that Q13 acts as a capacitance diode for fine-tuning the oscillator.

 

[MacGyverS2000]

If this circuit is as old as EMET claims it is, ijl, I highly doubt it's an inverted mesa crystal. Those things were all but theoretical until the late-90's.

Dan - Owner

http://www.Hi-TecDesigns.com

 

[EMET1970]

Q13 is p/n NE85630-T1, if that helps.

IC5 is a mixer p/n ADE 2ASK (another component I'm not familiar with)

More and more, I'm doubting what the manufacturer says about this being an oscillator.

MacGyverS2000 - There are 6 revisions of these units. The oldest dates back 15+ years. The newest I've seen is 2004. They all use the same CO circuitry, though. Yes, I'm sure it's as old as I claim it is...

 

[IRstuff]

Can you post the schematic that shows the mixer?

TTFN
faq731-376

Need help writing a question or understanding a reply? forum1529

 

[EMET1970]

Mixer (R5) circuitry attached.

 

[MagicSmoker]

Any possibility that the crystal is actually a SAW resonator? Granted, SAW resonators don't have nearly the frequency stability of a quartz crystal, but, then again, a 200MHz quartz crystal is going to be a rare bird... Try these guys for the latter: IQD Frequency Products (no affiliation).