Sunday, July 22, 2018

Open Source DIY Agilent 53132A 53131A 53181A OCXO Ultra High Stability Timebase Option Oven Upgrade

While i was working with my own GPSDO project. i need to have a frequency counter with descent stability so purchased my self a Agilent 53132A which is a 12 digit frequency counter , big brother to 53131A 10 Digit Counter. Both are really nice units.
But they unusable standard Timebase. So optional oven oscillator time base need to purchase. but 53132A and 53131A both unit are no longer available for sale. and neither of the Time base upgrades.

At first i decided to look around for time bases on online retail sites as already knew other people are selling them.
But i had few doubts , how well the whole thing is put together. what kind of parts they have used? How good is oven they are using? and Are they using same parts as per the original Agilent time base?
As it is quite normal for stuff listed on common online retail sites, It are quite cheaply made. and i found not even a single Time base who provides exact Bill of martial and give any specification details about the OCXO they are having. ofcourse OCXO is quite expensive so they all were using old OCXO which is totally fine but some data should be available.

3Ghz Upgrade 

Overall i was not quite satisfied with the quality of the module available on online. I did find few people who designed there own board but non of them were willing to give source of the PCB away and they were using electrolytic capacitors instead of tantalum, which i find a little odd in a part which get hot.
As 53132A/53131A component level specification is already in public domain so schematic is already available in document.
so i decided to make 53132A OCXO Ultra stability Time base my self. most important part completely open source.

Schematic , Download high resolution pdf from my github

As you can see there is not much in schematic , just few component , DAC to generate control voltage for OCXO, Reference for DAC, and a differential compactor to provide differential 10M output, few passive elements.
i have made this board to be able to fit different type of oscillators, big double oven in size of 51mm x51mm (MV89A type)and small ~36mm x 27mm (CTS970 or  BLILEY NVG47A1282 type)

As differnt oscillator may need different control voltage range . so it is possible to configure connection of internal register in DAC AD7243AR to provide output in various ranges , 0 to 5V what i need for MV89 , 0 to 10 and -5 to +5V. you can configure this according to your oscillator using J2 J3 and J4,

as small 36mm x 27mm oscillator may need 12V or 5V power supply so you can switch power voltage using J6 and J5.

i also have lad pattern for trip pot which can be used to tune OCXO or to fine tune reference chip.
so if you are not planing to use automatic calibration , you can skip many of the parts from the board and use pot to calibrate. J9 and J8 is used to switch between OCXO ctrl votage and reference chip fine tuning.
when DAC and every thing else is populated one can use POT  to fine tune reference chip. because with 12Bit dac you do not get much resolution  in control voltage.

why i chose MV89A is because this is one of the few Oscillators widely available on ebay and also have data sheet and specification available. Morion MV89 has a catalog in public domain which give some idea of specifications but does not tell exactly about the parts commonly available on ebay so i contacted Morion and ask them exact specification they were nice enough to replay exact specification for the part i have XO00281M-CT-MV89

Here the specification of both of my oscillators compared to Agilent Time base options , Agilent  have three(medium , high , ultra high stability oven) different type available all of them are listed here. small little Bliley and CTS970 outperform medium oven and MV89A is similar or better to Ultra high Oven

Medium Oven CTS970/Bliley NVG47A1282 High Oven Ultra High Oven MV89A (XO00281M-CT-MV89)
Temp Stability : 200PPB 5PPB 2.5PPB 2.5PPB .25PPB
Daily : 40PPB1PPB 40PPB 0.5PPB 0.1PPB
Monthly: 200PPBNA 15PPB 3PPB NA
Yearly: NA300PPB NA 20PPB 20PPB

Here how PCB Design Looks. Source files for PCB , Bill of material and Gerber are available in github,

Buy Blank PCB , HASL Finish, Free International Shipping


Calibration and Testing 
I have installed ocxo in in my 53132A and used my own GPSDO to calibrate it. here the viedo explanation.

 Here is my GPSDO project 

Mounting Dimensions 


  1. can I use this module with 53181A?

  2. Hello Gaurav! This is Pedro Silva, from Portugal. As i am not a very fond of soldering SMD components, i would like to get one ready-to-use board, maybe without the OCXO itself installed, which i could buy and install it myself. Do you know if someone sells these boards including components, in Europe? Thanks!

    1. Yes it is possible to buy this board assembled , Send me an email , my email you can find on my github or somewhere on this blog .


  3. Hello Gaurav. I've just recently bought a 53131A and am now looking into upgrading it with an OCXO. I have a question about the DAC: the AD7243AR has a built-in 5V reference, was there a particular reason to use an external MAX6350 reference instead of the built-in one?


    1. Original Agilent ocxo module also used external chip as refernce. I kept its as is. In my view use of external at bare minimum, can be justified by poor 30PPM Temperature Coefficient of on chip reference. While MAX6350 offers 0.5PPM Temperature Coefficient. Regards

    2. OK, thanks a lot. I've meanwhile built one of your PCBs, got a used MV89A from ebay and it seems to work. I had a bit of an oh-oh moment when the instrument failed the power-on self test ("GPIB: FAIL") but I found that I had mounted R3 and the trigger level for the comparator was 2.5V when it should have been 0V. After removing R3 the counter boots and reports "UNCALIBRATED", so that's the next step on the list.

    3. Alright, timebase calibration went with no problem, I have two questions about the schematic: What's the function of JP1, I don't see it explained anywhere but in one of the photos you posted it is soldered closed. Also, why are L2 and L3 of different value? That doesn't seem logical for a differential driver.

    4. This board clone of Agilent High and Ultra High stability time base. As per Agilent low level specification JP1 is there to select between Ultra High Stability time base and High Stability time base , set JP1 to sort to ground. L2 L3 have different value on original Agilent board. Maybe they drive differently on main board.

  4. Hello Mr Graurav, greetings ... this week i bought a pcb from you and i will test it in some ocxo that i have here ... just a doubt: i can use it in an ocxo MV102, similar to the MV89 but with square wave output .. . grateful

    1. As long as it fits , Square wave will be handled by the comparator and apropriate signal will be generated. So Square wave oscillator is not a issue.


    2. tank you Sir... grated by information...

  5. What kind of mounting stud did you use? How are they secured?

    1. These studs are from Wurth Electronik Part 9774100360 M3 With internal thread.
      They are soldered onto PCB then secured with M3 screw.

  6. Hi,

    FB1 and FB2,
    Description FERRITE BEAD 600 OHM 2A

    Part Number ILHB0805ER300V

    ILHB0805ER 300V seems to be 30 OHM.

    Which is the correct answer, 600 OHM or 30 OHM ?