[time-nuts] MTI 260-0624-D OCXO

Fri Feb 19 14:19:26 UTC 2016

Hi Mike,
Would you be so kindly and post some pictures of the opening process of 
that OCSXO ?
73
KJ6UHN
Alex
On 2/18/2016 7:54 PM, timenut at metachaos.net wrote:
> Bob,
>
> The vendor has said that they did not want the unit back. So...
>
> I opened it up. Crudely, I admit. I learned a few things. I was concerned
> about the outer case heating up too much when I was trying to remove the
> solder. Turns out that the outer case doesn't touch anything except the pins,
> so it can get pretty hot without any damage. Unfortunately, I used a bit of
> brute force to remove the casing after scraping away as much of the solder as
> I could and after cracking the solder seal with a hammer and screw driver.
> Even using more heat, I'm not sure of the best way to remove the case. Neither
> solder wick nor a vacuum desoldering tool is likely to remove all of the
> solder between the top case and the bottom.
>
> Even so, I got the case off. Somewhat bent, even bent the bottom circuit board
> a bit. I discovered that there is a 2mm hole in the top that allows a variable
> resistor to be adjusted. You need a pretty long, small screwdriver / adjusting
> tool to do that, but that is probably for setting the center point. I would
> try to remove the solder rather than drilling, or drill upside down to prevent
> solder flakes from falling inside.
>
> Also, I found that the -D on my part number appears to correspond to the board
> revision, which is marked "Rev D". So the -C and -D parts probably have the
> same specifications. And, on the inside there is a marking "92.0" which I
> believe would be the set point for this specific crystal. So if I took the
> crystal out, I would know where to design the set point for a custom unit
> (currently beyond my skills, but who knows...).
>
>  From there, I removed the bottom casing. That caused additional damage, some
> lifted traces and even one very small part (tiny, SMD, who knows?).
>
> But, I then soldered wires directly to the board, making patches for the
> lifted traces.
>
> I plugged it in.
>
> I turned it on.
>
> Success!
>
> I gave it 12v which should supply 2A, but it dropped the voltage down to just
> over 9V. Even so, I got a nice sine wave out at around 4.999790Mhz according
> to my (uncalibrated) scope and around 800mV (into 50 ohm, DC). That was with
> nothing attached to the adjustment pin. After some time, the frequency
> stabilized at 5.000014Mhz and the voltage came back up to 10v. When I attached
> the adjustment pin to ground, there was no change. When I attached the
> adjustment pin to Vref (which is at 6.15v), the frequency dropped to
> 5.000010Mhz. So apparently, these units adjust negatively. But also have a
> very wide adjustment range of 4Hz (8e-7). I am assuming that the adjustment
> range is 0..6v. I attached the adjustment pin to +10v and the frequency
> dropped another 3Hz for a 1.4e-6 adjustment range.
>
> I don't expect that this specific unit will be very useful given the damage
> that I caused opening it up. But, probably there were just bad connections
> internally, so if I were much more careful in the disassembly, I could
> probably have fixed it. Something to keep in mind for the next unit.
>
> I may be wrong for what the variable resistor does. Turning it made no
> difference in the frequency.
>
> I also suspect, but don't know that the difference from 5Mhz is due to the
> frequency counter being uncalibrated. It could also because I damaged the unit
> or just because it is very far off from where it should be.
>
> Still, I learned a lot, and well worth the time spent.
>
>
> Mike
>
>
>> Hi
>> The one advantage you have in testing a used OCXO is you have a pretty good idea of how
>> you are going to use it. If phase noise does not matter to you … no need to test. I’d always check
>> that it tunes on freq with reasonable EFC range left over. I’d also make sure that it warms up
>> properly (oven works) and that it has a reasonable output. What goes on the list past that …
>> it depends on what you need.
>> The gear you have will check aging and get it set on frequency fine. It will check it for “wander”
>> as your lab heats up and cools down. With a GPSDO and a simple phase lock, a DVM may be
>> all you really need to do most of that. You will not have a proper ADEV, but you will know it works
>> pretty well (or not ..).
>> A lightbulb oven / bench / fridge /freezer proces can give you a wide range TC if you need it.
>> If you do get into phase noise, a sound card system will get you going. For ADEV, the 10811’s
>> tune far enough that a single mixer system with your 53131 will give you good data. Both
>> of those will involve some building, but not a lot of money.
>> There is one thing about the 10811’s: They are not sealed units. They tend to soak up humidity
>> when stored in most parts of the country. You may want to run them for a month or three
>> before doing any fancy testing.
>> Bob
>>> On Feb 18, 2016, at 7:40 AM, Adrian Godwin <artgodwin at gmail.com> wrote:
>>>
>>> I have a small collection of 10811 and similar oscillators here, collected
>>> from hamfests rather than china (and mostly pre-ebay).
>>>
>>> What sort of testing regime would you put them through ? I don't have
>>> anything as exotic as a timepod but I do have an HP53131A, the Tait Rb
>>> source, and a KS-24361 set up. And always keen to get hints on the
>>> acquisition of new toys !
>>>
>>>
>>>
>>> On Thu, Feb 18, 2016 at 12:09 PM, Bob Camp <kb8tq at n1k.org> wrote:
>>>
>>>> Hi
>>>>
>>>> It is *much* better to have an OCXO that you do not need to test to death.
>>>> There is simply to much testing to do. A modern factory does not test the
>>>> quality into the product (of any sort) they design in and build it in. The
>>>> same
>>>> is true of the normal customer for virtually any component. They make sure
>>>> the parts come from somebody they can trust and save a lot of testing time.
>>>>
>>>> That said, yes, I have a pile of scrap parts sitting in front of me as I
>>>> type this.
>>>> It’s a hobby. Time does not count. It lets me play with all sorts of toys
>>>> doing
>>>> the testing. I occasionally learn things in the process. Mostly I learn
>>>> that
>>>> my hope of a 100% perfect batch is still a dream ….
>>>>
>>>> Bob
>>>>
>>>>
>>>>
>>>>> On Feb 17, 2016, at 9:56 PM, timenut at metachaos.net wrote:
>>>>>
>>>>> Bob,
>>>>>
>>>>> In this case, I know how it was taken off the board - it wasn't. They
>>>> just cut
>>>>> the board around it. I had to remove it myself. Nice thing about that,
>>>> is that
>>>>> I got a nice plastic, pre-formed part that fits between the board and the
>>>>> OCXO, probably as an insulator.
>>>>>
>>>>> You are also right about the number of things that you need to test to
>>>> make
>>>>> sure that one of these is fully functional. However, this is a "starter"
>>>> OCXO
>>>>> for me and I don't yet have the equipment to perform the tests. The best
>>>> that
>>>>> I can do is to make sure that, when powered up
>>>>>
>>>>>   1. I see something that sort of looks like a sine wave at a reasonable
>>>>>      magnitude.
>>>>>
>>>>>   2. My 2465B CT frequency counter thinks that it is somewhere near 5Mhz.
>>>>>
>>>>>   3. If I apply gnd, or VRef to the adjustment pin I see the frequency
>>>> change
>>>>>      at least a small amount.
>>>>>
>>>>> Without more equipment, there isn't much more that I can do to test.
>>>> Unless
>>>>> you have some suggestions?
>>>>>
>>>>>
>>>>> Mike
>>>>>
>>>>>
>>>>>> Hi
>>>>>> Best guess is these things get taken off the board with either a big
>>>> torch or a charcoal fire.
>>>>>> You can ask Mr Google to dig up pictures of the process. Depending on
>>>> just how quick
>>>>>> they are, the insides of the OCXO can easily be reflowed. The
>>>> likelihood of it reflowing and
>>>>>> cooling back to a reliable joint … not real good.
>>>>>> Bob
>>>>>>> On Feb 17, 2016, at 5:16 AM, Andrea Baldoni <erm1eaae7 at ermione.com>
>>>> wrote:
>>>>>>> On Tue, Feb 16, 2016 at 07:58:21PM -0500, timenut at metachaos.net wrote:
>>>>>>>
>>>>>>>> I just received a 5Mhz OCXO from eBay (MTI 260-0624-D OCXO). After
>>>> testing it,
>>>>>>>> it is clear that it is defective.
>>>>>>>>
>>>>>>>> 1. It never heats up.
>>>>>>>> 2. The reference voltage is zero.
>>>>>>>> 3. Only noise is seen on the output pin.
>>>>>>> I had the same issue with some of them. It's very likely that the
>>>> internal
>>>>>>> solder connections from pins to PCB are broken, at least, the one for
>>>> power.
>>>>>>> It happens because the inner oven is heavy and there isn't any thermal
>>>>>>> insulator (besides air) to keep it from moving.
>>>>>>>
>>>>>>> Wheter the crystal has been damaged or not, it's unknown. I had one
>>>> where the
>>>>>>> crystal actually fell off from its supports too.
>>>>>>>
>>>>>>> I posted a link to photos of the internals, time ago. The link is dead
>>>> now
>>>>>>> and I don't think to have the photos anymore but perhaps someone
>>>> downloaded
>>>>>>> them.
>>>>>>>
>>>>>>> Best regards,
>>>>>>> Andrea Baldoni
>>>>>>> _______________________________________________
>>>>>>> time-nuts mailing list -- time-nuts at febo.com
>>>>>>> To unsubscribe, go to
>>>> https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
>>>>>>> and follow the instructions there.
>>>>>> _______________________________________________
>>>>>> time-nuts mailing list -- time-nuts at febo.com
>>>>>> To unsubscribe, go to
>>>> https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
>>>>>> and follow the instructions there.
>>>>>
>>>>>
>>>>> --
>>>>> Best regards,
>>>>> Timenut                            mailto:timenut at metachaos.net
>>>>>
>>>>> _______________________________________________
>>>>> time-nuts mailing list -- time-nuts at febo.com
>>>>> To unsubscribe, go to
>>>> https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
>>>>> and follow the instructions there.
>>>> _______________________________________________
>>>> time-nuts mailing list -- time-nuts at febo.com
>>>> To unsubscribe, go to
>>>> https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
>>>> and follow the instructions there.
>>>>
>>> _______________________________________________
>>> time-nuts mailing list -- time-nuts at febo.com
>>> To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
>>> and follow the instructions there.
>> _______________________________________________
>> time-nuts mailing list -- time-nuts at febo.com
>> To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
>> and follow the instructions there.
>
>