[time-nuts] Surface Mount OCXO Questions
John C. Westmoreland, P.E.
john at westmorelandengineering.com
Thu Oct 31 02:54:05 UTC 2013
Graham,
Good points - yes, I have this part currently in the design: TPS75833KTTT
(LDO from TI) - putting another one down
(just) for the OCXO isn't a problem.
And a nice 12V rail isn't a problem either since this is for a radio with a
nice 12V source. Could I boost the 3.3V rail to
12V or maybe 5V to 12V - sure - but your point about the switcher is well
taken and I agree. Having a nice, fat, analog
ground plane isn't a problem either.
And, this is just a 'dev' board so we can do what we need to make the OCXO
work as good as possible.
>From their spec sheet: '... the supply voltage sensitivity and load
sensitivity is 5E-11 for a 5 % change in voltage or
load impedance.'
Thanks,
John
On Wed, Oct 30, 2013 at 6:47 PM, Graham / KE9H <timenut at austin.rr.com>wrote:
> John:
>
> Look at the ppm (or however they express it) as to the sensitivity of the
> frequency stability of the OCXO relative to Voltage input.
>
> Say the oven power drops from 3 watts to 1 Watt as the oven comes up
> to temperature. At 3 Volts, relative to 12 Volts, for a given resistance,
> it is four time the Voltage change due to the higher currents, and an
> additional
> four times the percentage of the operating Voltage as a ratio. So
> additional
> design consideration for Voltage control/stabilization is needed.
>
> If you have a solid (wide, thick, multi-layer) ground, then that can
> work. To reduce the voltage drop feeding the OCXO, you might consider
> putting a dedicated LDO regulator, right at the OCXO, that shares the
> ground
> reference with the OCXO, so any voltage drop in the feed side is removed,
> as well as any Voltage variability with current in the ground system.
>
> As to why they are selling the 3.3V part, they probably started selling it
> before they had some customers get into performance issues per the above.
> But once offered, they have to continue to support their customers.
>
> I think they are just telling you that it is somewhere between 4 and 16
> times
> easier to get the full performance out of the part with a 12 Volt power
> feed
> than a 3 Volt power feed, not that you can't get full performance with a
> 3.3V feed.
>
> I am sure their parts meet specs, you just need to understand them.
>
> P.S. - I would stick with linear regulators feeding the OCXO, not a
> switcher.
>
> --- Graham
>
> ==
>
>
> On 10/30/2013 7:37 PM, John C. Westmoreland, P.E. wrote:
>
>> Graham and Time Nuts,
>>
>> (thanks for the answers.)
>>
>> I have another question - I am looking at a part from MTI. I wanted to
>> use
>> one of their 3.3V parts. They are telling me to use the 12V part because
>> the 3.3V part can have an issue with ground loops due to the higher
>> current
>> requirements at that voltage for the oven.
>>
>> Have any of you experienced this? Makes me wonder a little why they offer
>> the 3.3V part. It would seem good layout can control any possibility of
>> ground loops becoming a problem.
>>
>> Thanks and Regards,
>> John W./AJ6BC
>>
>>
>>
>>
>> On Wed, Oct 30, 2013 at 6:54 AM, Graham / KE9H <timenut at austin.rr.com
>> >wrote:
>>
>> John:
>>>
>>> All SMT OCXO's will either have a recommended PCB footprint in the spec
>>> sheet
>>> or will refer you to a recommended footprint in another document.
>>>
>>> Some don't care about a ground plane under the part, some require it
>>> with no crossing signals, some require an open thermal hole underneath
>>> the oven. I have seen all three cases. As usual, it is suggested that
>>> you
>>> read the [] manual.
>>>
>>> Best regards,
>>> --- Graham / KE9H
>>>
>>> ==
>>>
>>>
>>>
>>> On 10/29/2013 9:18 PM, John C. Westmoreland, P.E. wrote:
>>>
>>> Hello,
>>>>
>>>> I was wondering if I could get some recommendations on surface mount
>>>> OCXO's
>>>> vs. the traditional through hole.
>>>>
>>>> I was also wondering on the board layout - if you found it necessary to
>>>> leave a thermal moat so to speak - and what worked best. Maybe the OCXO
>>>> has an internal air barrier that maybe would make this unnecessary - not
>>>> sure.
>>>>
>>>> Your input and experience appreciated.
>>>>
>>>> Thanks,
>>>> John Westmoreland
>>>>
>>>
> ______________________________**_________________
> time-nuts mailing list -- time-nuts at febo.com
> To unsubscribe, go to https://www.febo.com/cgi-bin/**
> mailman/listinfo/time-nuts<https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts>
> and follow the instructions there.
>
More information about the Time-nuts_lists.febo.com
mailing list