EEVblog Electronics Community Forum
Electronics => Beginners => Topic started by: fmzambon on September 06, 2020, 10:01:26 am
-
Hello,
I'm designing a Nixie clock from the ground up using 6 tiny (see picture) Nixies ;D
I know that many nixie clock designs use an external wall adapter that produces a low voltage DC supply and then use a step up converter to generate the 170+ V needed by the nixies. However this prevents one from using the mains as a timing reference, which should have a very good long term accuracy. So I'd like to integrate the whole power supply in the clock to have access to the mains frequency.
Since this is the first time that I deal with the mains I thought I'd ask for a review of my power supply circuit, lest I do something stupid.
Connector X1 to the left is the mains input. The center conductor is earth while the two outer ones are 230VAC. Since our mains plugs here in Italy are completely symmetrical, there is no guarantee whatsoever as to which one will be live and which one will be neutral.
I'm not sure about the primary side input protection. I added a 50mA fuse (I didn't want to use a glass fuse across the mains, and 50mA is the smallest size I could find with a ceramic case) and a 275V MOV. Would that be a sensible level of protection? :-//
Safety capacitor CY1 connects mains earth with the logic ground. Again, is this a sensible thing to do? Is the capacitor value reasonable?
The transformer ( https://www.hammfg.com/files/parts/pdf/229A120.pdf (https://www.hammfg.com/files/parts/pdf/229A120.pdf) ) is wired in a convoluted way due to layout reasons, but I believe that I have the polarity correct. It is a 6VA transformer wired as a 230V to 120V step down. It has a 30% regulation which, to my understanding, means that the unloaded output voltage can be up to 30% higher than the nominal value.
On the secondary side there are two supplies: CX1, C1, C2 and CX2 form a capacitive divider that feeds a low voltage AC into the Schottky bridge rectifier formed by D1, D2, D3 and D4. Then there's some capacitance and the low voltage goes off unregulated to power the CD4000 digital portion. There's no regulation here because the 4000 logic has a huge supply range and because the current consumption is tiny. I'd be wasting most of the power just to meet the minimum load requirement of a 7805. As such the "+5V" supply line can be anywhere from 5V to 10V, depending on the mains voltage.
G1 and D5 are a battery backup to keep the logic part powered in case of mains failure. Should that happen, a backup 32768Hz crystal oscillator timing source takes over until the mains return.
The bottom part of the circuit is the HV supply. The 120VAC that comes out of the transformer secondary is rectified and filtered by C5. C6 may or may not be necessary, depending on where I can place C5 in relation to the regulator. C5 is a 450V 33uF electrolytic ( https://www.mouser.it/datasheet/2/420/United-Chemi-Con-1511290.pdf (https://www.mouser.it/datasheet/2/420/United-Chemi-Con-1511290.pdf) ), while C6 and C7 are 450V 1uF metallized film polypropylene ( https://www.mouser.it/datasheet/2/315/ABD0000C255-1131244.pdf (https://www.mouser.it/datasheet/2/315/ABD0000C255-1131244.pdf) ).
The LM317HV is NOT there to filter out the ripple. It is there to cater for fluctuations in the mains voltage. The nixies that I plan to use have a fairly high sustaining voltage (around 154V), so the anode resistors are calculated to produce the correct current with a voltage drop of 180V - 154V = 26V. Without the regulator, even small fluctuations in the mains voltage could lead to important variations in the anode current, which could easily overdrive the little nixies. The 317 is there to swallow up any voltage fluctuation and produce an output voltage that is never more than 180V. It can be less than that if the mains voltage is low, though.
Finally I use zeners for the 317 instead of the classic two resistors approach because without the zeners I would need to make R2 a very high value. Which is fine in itself, except when one considers that the 317's adjust pin current has to go though R2. Since that current can be all over the place, the output voltage could be off by several volts.
Thoughts? Comments? Suggestions?
Andrea
-
I know that many nixie clock designs use an external wall adapter that produces a low voltage DC supply and then use a step up converter to generate the 170+ V needed by the nixies. However this prevents one from using the mains as a timing reference,
I'm not sure I understand the reasoning here. I'm sure you could extract the 50/60 Hz line signal as well as boost a low voltage up to the voltage required by the nixies.
The nixies don't need a lot of current. With this approach you are putting a lot of power in your nixie PSU. If something goes wrong a lot of current could flow.
-
Why not use an AC adapter instead?
-
I'm not sure I understand the reasoning here. I'm sure you could extract the 50/60 Hz line signal as well as boost a low voltage up to the voltage required by the nixies.
The nixies don't need a lot of current. With this approach you are putting a lot of power in your nixie PSU. If something goes wrong a lot of current could flow.
I mean that several nixie PSU designs that I've seen use an external AC/DC adapter, so they only receive DC. Then they step up the DC to whatever voltage is needed by the nixies.
But if the only input that I receive from the power connector is DC, how can I get the 50Hz timing reference?
Unless I'm missing something, I need an AC input in order to see the line frequency. Which brings me to...
Why not use an AC adapter instead?
The difficult thing finding a suitable one. The 120VAC output voltage is convenient because, when rectified, it's almost the perfect voltage to drive the nixies.
But most 230VAC to 120VAC adapters that I've seen are intended to power appliances that take several tens of watts or more, so they are fairly bulky and overpowered for the application. I only need a couple of watts tops,
I could use an AC adapter with a lower output voltage, but then I'd need to step that up, which is perfectly doable. But if I'm already using a transformer somewhere in the power supply, why not have it output the voltage I need?
That's why I went with the idea of integrating the mains supply using a small 6VA transformer. Correct output voltage and not oversized for the application.
On top of that, I have a few ideas for other projects that could use mains power. So I'd like to take the opportunity to gain some experience with a low power mains power supply before moving on to something more ambitious.
-
Why not use an AC adapter instead?
The difficult thing finding a suitable one. The 120VAC output voltage is convenient because, when rectified, it's almost the perfect voltage to drive the nixies.
But most 230VAC to 120VAC adapters that I've seen are intended to power appliances that take several tens of watts or more, so they are fairly bulky and overpowered for the application. I only need a couple of watts tops,
I could use an AC adapter with a lower output voltage, but then I'd need to step that up, which is perfectly doable. But if I'm already using a transformer somewhere in the power supply, why not have it output the voltage I need?
That's why I went with the idea of integrating the mains supply using a small 6VA transformer. Correct output voltage and not oversized for the application.
On top of that, I have a few ideas for other projects that could use mains power. So I'd like to take the opportunity to gain some experience with a low power mains power supply before moving on to something more ambitious.
The idea would be to derive all of your supply voltages from a low voltage AC source just like a DC source, but which also provides the AC signal for timing.