[time-nuts] Using a common power supply among few time standards

[Will Kimber]

Hi,

Usual (best?) is supply slight over voltage and use LDO/RC filtering in
each unit as Bob suggested. But that would require either supplying
24v/12v/5v/-12v (with some margin 10%?) or converting some single supply
of 24v or 48v  in each unit to required voltages at each piece of
equipment.  The later requires switching supplies to get all voltages.


Anybody fancy 400Hz or even higher frequency 48v AC supply?

Will

On 7/02/20 11:52 am, Bob kb8tq wrote:
> Hi
>
> Well, I’m sitting here plotting the stability of some eBay “frequency standards”. The one thing I *can*
> measure is just when the power cords get bumped.
>
> As long as you can hold sub mili-volt sorts of stability on the supplies “as delivered” it pretty much does not
> matter how you get there. If you get up into the 10’s of mili-volts then it can indeed matter on some devices.
> Some devices have less sensitivity than others. By the time you get to 100’s of mili-volts, you likley are impacting
> most devices in a manner you can easily see in the output.
>
> Yes, there are a lot of assumptions in all that. The first is that it is a 12V supply and not something odd like 48V.
> Scale the numbers if you are using an odd voltage. The next Is that the disruption is not simply the lead drop
> from the oven current. That *will* have an impact, but it’s going to get mixed up with temperature stability. Step /
> pop / jump sort of changes will come through to the output fairly quickly. Third is that the supply is actually going
> into something that matters (like an OCXO heater) and not just the supply lead for an opamp. There are a lot
> more, but they get a bit fiddly ….
>
> One alternative is to generate something like 13V and run a good LDO at the “point of use”. Depending on what
> you find for LDO’s you may be able to get quite close to 12V. The power “wasted” becomes very small in that case.
>
> Bob
>
>> On Feb 6, 2020, at 5:28 PM, Taka Kamiya via time-nuts <time-nuts at lists.febo.com> wrote:
>>
>> I have been addicted to home brewing GPSDO, Rubidium, OCXO, and etc, etc, etc.  I don't mean making of these from scratch but taking a surplus unit and encase them in ready-to-use form.  Most of them are in U2 rack case.
>>
>> One part of it that I really do not like is constructing a power supply for each and every unit I make.  Good power supplies are essential to clean and stable timebase, I know.  It is also heat generating as most of what I make use linear regulators.  But it's tedious and mandane to do it every single time!  So, I wanted to make a large enough power supply to provide +24V, +/- 12V, and 5V.  My plan was to have independent feed line for each timebase unit.
>> Is this even a good idea?  I have no experience in this to make a reasonable judgement.  This plan would mean these power voltages will be fed to each components without any local regulation.  While I do plan to put ferrite chokes and bypass capacitors at input, but since the whole point is to eliminate local power supply components, there will be nothing else in terms of isolation.
>> Similarly, I was also thinking of having a large battery bank for 24 volt supply.  What type of battery is irrelevant and so as charging method for this question.  I have heard of this being done.  But for units that require 24 volts, that means there is no room for any type of regulation.  Most 24V batteries output varying voltage depending on state of charging.  I know 12V battery can produce as high as 13.x volt.  Is anyone doing this type of thing on this list?  If so, would you be willing to share details?
>>
>> Appreciate any input.
>>
>> ---------------------------------------
>> (Mr.) Taka Kamiya
>> KB4EMF / ex JF2DKG
>> _______________________________________________
>> time-nuts mailing list -- time-nuts at lists.febo.com
>> To unsubscribe, go to http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com
>> and follow the instructions there.
>
> _______________________________________________
> time-nuts mailing list -- time-nuts at lists.febo.com
> To unsubscribe, go to http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com
> and follow the instructions there.