Crystal oscillator power requirement help

[4thDoctorWhoFan]

I would like to test this 10Mhz oscillator that was pulled from a Rohde & Schwarz / Polarad Signal Generator, model# 309.
The sig gen was already missing parts when I got it so I cannot power up the sig gen to test the voltages and I cannot find any info on this particular model.

I don't want to damage the oscillator, so I'm hoping somebody is familiar with this particular unit.  It has three power cables and I need to figure out how to power it up properly.  The oscillator is a Vectron 317-8009.  Please see attached photo.

Any help would be appreciated.

[SeanB]

Trace the power rails. One going to the TTL IC will be 5V, and the one lead will be ground. The capacitors by the power input would also give a clue as to voltage rating, picture of the back of the board and the values ( capacitance and voltage) of the 2 capacitors by the power connector will also help.

[ogden]

The capacitors by the power input would also give a clue as to voltage rating

Be careful here! Capacitors can have huge safety margin. You can easily have 25V capacitor on 12V rail which will blow when 25V applied.

[TheUnnamedNewbie]

The capacitors by the power input would also give a clue as to voltage rating

Be careful here! Capacitors can have huge safety margin. You can easily have 25V capacitor on 12V rail which will blow when 25V applied.

Very true. But at the same time, seeing a 12 V capacitor makes means you can be quite sure it's not gonna be a 15 or 24 V rail, and more likely a 12 V or 5 V one.

[4thDoctorWhoFan]

One of the rails is definitely 5V since it goes to the logic circuit.
The question is the other rail which is connected to a 35V cap.  I'm leaning towards 12V.

Also, I have to figure out what the purpose of the other 3 pins of the osc.
None of them are connected directly to the RF out.

[SeanB]

Seeing as there is a 74 series high speed logic IC there, that more than likely is doing a function of high speed output buffer, and driving of the output coax. The IC being there with 2 capacitors suggests it is being used as 2 stages of buffer for the oscillator. Thus the IC will be connected to the oscillator, not the output, and it is giving a 10MHz square wave output, instead of the low drive ability of the oscillator one.

I would lean to the other rail being a 10V one, as this will at least be high enough to power the heater fine, and the oscillator having 4 pins puts one being ground, one 5V supply for the oscillator in the can, one pin probably is connected to a resistor divider providing some trim for the oscillator, provided by the precision resistors on the board, and the last one is a heater supply.

[texaspyro]

Try asking on the time-nuts mailing list.  Lots of knowledgeable people there.   

The heater is probably either 12V or 24V.   If worse comes to worse, try 12V.  If that does not work then 24V.  Or even better, drive it with a lab supply and and increase the voltage in steps. 

The heater needs to heat the crystal to the point where the crystal has the lowest tempco (usually in the 80C range).  The heater is controlled by a thermistor.  Below the proper operating voltage, the oven temp will increase with rising input voltage.  At some point the oven temp will start to stabilize... this will be near the proper operating voltage (which will be a bit higher since the oven needs some headroom).

[ArthurDent]

The pinout on your oscillator may be similar to other models that Vectron made-maybe not. As others have said, a photo of the back side of the board would be a big help in trying to figure it out. Here is a photo of a defective 5-pin different Vectron model that I removed and marked the pins. Check to see if yours has the same pin configuration. You can verify which pin the case is connected to with an ohm meter. The unmarked unused middle pin on my oscillator is used as a voltage reference for the EFC on some models and that pin and the EFC pin may not be used on some models that just have a mechanical adjustment even though they use a common pin arrangement that prevents the oscillator from being placed on the board backwards.

 If the 74S00 on your board is just connected directly to one of the supply lines and that line goes to nothing else, connect a voltmeter to measure pin 14 and adjust a bench supply until the voltage is 5.0 VDC. I can’t tell from the shadows in your photo whether there might be a 78L05 regulator on the board but if there is, to get 5.0 VDC on pin 14 you might have to go higher than 5.0 VDC on that supply lead. If the lead that powers the 74S00 also powers the oscillator then the 3rd lead on the supply could be connected to EFC but you really need to trace out the back side of the board to tell. 

Next connect the coax to one channel on a scope and the output pin from the oscillator to a second scope input. Connect another adjustable supply to the second supply lead and increase the voltage slowly until you see a waveform on the scope. What you should see is the output level from the oscillator pin increase and at some level it will be high enough so you will see an output square wave on the coax. As you increase the supply voltage you’ll probably see the 10 Mhz level on the oscillator pin stop increasing because most good oscillators have an internal regulator for the oscillator to help isolate it from the oven circuitry even though they may have only one supply pin. The correct supply voltage for that lead will be slightly higher than what is required for the output to level off.

Most OXCOs operate from common voltages so 5 , 12, 15, or 24 are likely voltages to be the correct one but the voltage rating on the caps divided by 2 is probably the highest I’d dare try (I'm betting on 12).  I’ve used this method on unmarked oscillators I’ve pick up at flea markets but I make no claim that this method could work in your case with so little information so you’re on your own. Good luck.

[4thDoctorWhoFan]

I have attached a pix of the rear of the pcb.

layout

Thought I would also include a closeup of the components on the pcb.

20180111_195507

A member from another board found the layout.

2x2 OSC Pinout

[4thDoctorWhoFan]

I have had it powered on now for about an hour using 5V & 12V and I'm getting the waveform shown in the picture directly off pin 3.
Is that a typical looking sine wave for one of these osc?

On initial power up, the 12V line was pulling .22A and after a few minutes dropped to .13A.
Pin 2, the no connect line, has been at 5V since initial turn on and has not changed.
Pin 1 has remained at 0V.

scope

[texaspyro]

Try terminating the output with 50 ohms.

[mmagin]

I have seen similar waveforms for oven oscillators that I have.  It might be better with termination.

[ArthurDent]

The pin out drawing I sent earlier isn't different than the one you received later except that I mentioned that rather than be NC, pin 2 can have a reference voltage that can be used with EFC. It appears that pin 2 of your oscillator has the reference voltage and Q1 is just used as a switch to connect pin 2 of the oscillator to pin 14 of the 74S00 to turn the output on the coax on/off. Check the voltages on pin 2 and pin 14 to see if pin 2 is about 1/2 a volt higher than pin 14 or you can disconnect the external 5 VDC to see if pin 2 still reads 5 VDC. If the 5 VDC is present with the external 5 VDC disconnected, you can run the board just from 12 VDC if you connect the lower end of the 10K resistor to pin 2 on the oscillator and ignore the 5 VDC lead.

If the EFC (pin 1) is also connected internally you can probably connect a 10-turn pot from Vref (pin 2) to ground (pin 4) and the wiper to pin 1 to trim the output frequency.     

[4thDoctorWhoFan]

The pin out drawing I sent earlier isn't different than the one you received later except that I mentioned that rather than be NC, pin 2 can have a reference voltage that can be used with EFC. It appears that pin 2 of your oscillator has the reference voltage and Q1 is just used as a switch to connect pin 2 of the oscillator to pin 14 of the 74S00 to turn the output on the coax on/off. Check the voltages on pin 2 and pin 14 to see if pin 2 is about 1/2 a volt higher than pin 14 or you can disconnect the external 5 VDC to see if pin 2 still reads 5 VDC. If the 5 VDC is present with the external 5 VDC disconnected, you can run the board just from 12 VDC if you connect the lower end of the 10K resistor to pin 2 on the oscillator and ignore the 5 VDC lead.

If the EFC (pin 1) is also connected internally you can probably connect a 10-turn pot from Vref (pin 2) to ground (pin 4) and the wiper to pin 1 to trim the output frequency.     

Thanks for the great info.
I'll look into it.