[time-nuts] Re: FTS 1050A problems when powered by DC?

[zfe at gmx.net]

I meanwhile can explain some of the causes of my „problems“.

The LEDs:
The anodes of the of the green and yellow LEDs as well as of the Limit
DAC and Limit DET are  connected to the unregulated 24V Bus (i.e tab 4,5
of the transformer which is about 24V DC with respect to the  star
ground). If the 1050A is powered by the DC input or battery there is no
tension at this bus due to a diode - thus these LEDs do not work then.
Except for the LEDs, the unregulated 24V Bus seems only be used for the
DC-OUT Pin B.

The frequency jump:
When AC powered the „V_osc“ regulator gets an input voltage of about
48V. When DC powered, in the recommended range, the input voltage is
only about 26V. This causes a jump of about 20mV at the output of the
regulator so roughly a 0.1% jump in V_osc. According the specs of the
1000B OCXO this is, by an order of magnitude, not enough to explain a
frequency jump of about 5*10^{-11}.
Switching from DC to AC causes also a jump of about 1mV in the control
voltage of the OCXO (of about 5V in my case) which, according the
manual, would lead to a 4*10^{-11} frequency change. This fits quite
well. Moreover the log of the control voltage has the same ramping as
the frequency jump (see first attachment).
The jump of the control voltage can not be explained by the jump of the
output of the 10V reference, which is only 0.7mV and shows no ramping at
all. The cause should be thus somewhere on the DAC board - more
investigations needed.

The frequency oscillation when DC powered:
I have no indication where that comes from up to now. I found no
oscillation of matching period (of about 4 minutes) or strength in any
of the voltages.
The only discontinuity in current consumption, I have found, is a 2mA
jump to the DAC board (from +15V of the DC-brick). The jump is quite
frequent when the time constant is set to 1s (although there is no
external oscillator attached) and has a 40s period when the time
constant switch is set to „manual“ (see the second attachment - first
part switch to 1s, last part switch to manual).
The oscillation strength and clarity depends on the used DC supply. But
I have not yet found which quality of the supply causes a difference. It
does not depend on the voltage, I went op to 48V, so the same as with
AC. I added some capacitors in parallel which also shows no effect.
And even with AC-power there is a slight bump in the Allan plot at the
appropriate location. So AC-power is also weakly a victim of the effect.
As other idea of the cause of the oscillation I suspected a short glitch
in power that triggers the oscillation. E.g. from the tickle charger.
There are very short exceptional glitches visible on the scope, but not
regular and somehow fitting the required period. Moreover I am not sure
the glitches are from the voltage rail or only some artifacts.
The charger is not on a dedicated board but only a part of the A3
control board. I have to examine that board further to see how to
disable the charger.


Am 18.06.25 um 22:09 schrieb zfe--- via time-nuts:
> Although it makes no sense at the first glance my 1050A (opt 015 and 002
> but no batteries installed) have some issues when powered from the
> DC-input, while they perform unsuspicious when powered by AC.
> 
> The most visible issue is that the „oven ready“ and „osc int“ lights are
> off when powered with DC. This happens with all three of my 1050A. The
> oven, however, performs as expected according the control instrument.
> 
> But the more severe issue is that with DC power the stability is
> degraded. There is a larger drift with some regular oscillating
> modulation. The first picture is magnification of the first part of the
> second. It shows a measurement with DC power. The 1050A was powered by
> that DC source for about 3 weeks before the measurement.
> 
> The second picture is the complete measurement. At about 6500s I plug in
> AC. There is a jump in frequency and after stabilization the stability
> looks much better than with DC.
> My first idea was that the SMPS DC supply  could be  the culprit. So I
> attached a linear HP lab-supply to the DC input and disconnected the AC
> power at about 20000s. The frequency dropped and the oscillating drift
> is back -Â  even more pronounced!
> 
> 
> I powered with 29V DC and the supplies had more than enough power. The
> DC performance is still within the specs  but AC is considerabely
> better. But this is only with the 1050A that performs best with DC, the
> other two have more severe DC stability issues.
> 
> Did anybody noticed a similar problem and has a hint where the problem
> could be located? … or is this a feature ;)
> _______________________________________________
> time-nuts mailing list -- time-nuts at lists.febo.com
> To unsubscribe send an email to time-nuts-leave at lists.febo.com
> 
-------------- next part --------------
A non-text attachment was scrubbed...
Name: 1050A_3_ContV.png
Type: image/png
Size: 39483 bytes
Desc: not available
URL: <http://febo.com/pipermail/time-nuts_lists.febo.com/attachments/20250715/8cbc1166/attachment.png>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: 1050A_3_current.png
Type: image/png
Size: 148173 bytes
Desc: not available
URL: <http://febo.com/pipermail/time-nuts_lists.febo.com/attachments/20250715/8cbc1166/attachment-0001.png>