Above: The nameplate. Model OSC 73-52, Freq 10.0 MHz, SER. NO. 20129-37, A-0960-0465-1

Above: The whole unit

Above: The oven unit, all disassembled

The KIDDE OVENAIRE DIV Precision Crystal Oscillator

(Click images for bigger)

On March 7th, 2008 (3/7/2009) I went with friends to the yearly Mike and Key club ham fest swap meet at the Puyallup, Washington state fairgrounds where there's a huge two story building - and both floors were filled with rows of tables with people selling all sorts of different stuff!

One of the gizmos I picked up for $12 was a "10.0 MHz Kidde Ovenaire Div Precision Crystal Oscillator, Charlottesville Virginia" installed on an HP circuit board of model "03325-66509" which provides a fine-adjust pot as well as a 14.1v voltage regulator.

It was supposed to be good, but when I tested it, it would put out 10MHz for about 15 seconds then get erratic and stop. But my amp meter told me that it's internal heater was working because it started out drawing about 0.39 amps but after a while it went on down to about 0.2 amps, and also it was slightly warm to the touch. Upon letting it cool and trying again, the same thing happened - it worked at first, then it stopped working. Well, it's a heated device so this doesn't take a super-sleuth (Good thing for me) to figure out that it's got a heat-related mechanical problem.
So I took it apart to fix it, and found some interesting things.



Above: The quartz crystal with an entirely glass (or fused quartz) case


The two input diodes

My Kenwood R-1000 at 10Mhz

The quartz crystal wafer enclosed in an entirely glass case

I think the most interesting and surprising thing was the quartz crystal -- specifically it's housing was entirely made of either white glass (for the base) or clear glass (for the dome/top)!


After reassembling it, I tested it and it seems to work fine. It starts out about 400 hertz too high, but after about 10 or 15 minutes of warm up time, it seems to settle to within about a fourth of a hertz of WWV on 10MHz - so I guess I can't really complain!

To test it, I tuned my Kenwood R-1000 to the 10MHz WWV time standards station in AM. I then turned on the crystal oven unit and found that the low power 10MHz RF that it transmitted from my attached clip lead caused a beat on the radio receiver -- in other words, sort of a warble of a few hundred hertz. But as the crystal oven unit warmed, the beat slowed down until it was about 5 hertz, and I could actually see my meter needle bobbing up and down. As the crystal unit finished coming up to temperature, the beat reduced finally until the rolling in and out of phase took around 4 seconds.

(For those who don't know, a crystal oven is a little insulated compartment with a few-watt heating element. Inside the inner chamber is a crystal oscillator which produces a frequency (10MHz in my case). There is also a temperature sensor which monitors the temperature of the inside chamber, and keeps it at a constant temperature so that the frequency generated will be always the same. This is needed for very high accuracy frequency applications because crystals do change their output frequency slightly as a result of temperature change. )

Specifications (For the oven module)


I don't know what the specifications from the manufacturer are, but these seem to be be pretty close:

  • Frequency: 10.0MHz
  • Accuracy: At least 1 part in 10 million
  • Voltage: Probably 12-14 volts.*
  • Current: 0.39A at 14v when cold, about 0.25 when hot.
  • Warmup time: Probably around 10-15 minutes.**

    Notes:
    *The built in regulator on the carrier board that came with mine limits the voltage feeding the unit to 14.11 volts. It also seems to run fine at around 12 volts but the accuracy may only be within about 1Hz.
    **The manufacturer probably suggests half an hours warm up but I have no idea.

    Specifications (for the whole HP PCB unit)


  • Voltage: 15-18 seems to work fine. The voltage regulator puts out 14.1V and has about 0.9 to 1v dropout - so to get 14.1v out, at least 15v input is required. The unit ran fine without overheating for an hour long test at 18v.
  • Number of power inputs: 2. Use one or other or both. Diode ored.
  • Output is capacitively coupled.

    Pinout (HP board)


    On the 3 pin plug, center is ground and the outside two pins are both power, diode ored together. You can connect them up to two separate power sources and it'll draw current from the higher voltage source because it has two diodes so power cannot feed back out one of the inputs.

    Of course the coaxial jack is the capacitively coupled 10MHz output. The jack is basically an RCA jack, except that the center pin on a standard RCA cable is too long and bottoms out. Maybe it's low-profile RCA or something.

    Pinout (Oven module)


    Looking at the bottom, with the pins in a row running left to right near the top edge, the pins will be nearer to the right side. From left to right, the pin functions are as follows: Output (DC Coupled!), Adjust, Ground, Power.

    The fine-adjust seems to adjust between 0v and 4.6v. mine is adjusted all the way to 0v.


    Webpage published 3/13/2009
    Last updated 3/13/2009