Working Principle Of Neon Transformer Circuit

[iraquois]

Hi. I am new to this forum and I couldn't find where to post this message. Sorry if it is in the wrong place. My intention is to make and understand the working of a 4kV 10mA 40W neon transformer circuit. I know it can be done with a flyback transformer and some pwm modules like tl494 sg3525 uc3845 NE555 etc... But Does anyone know what is inside a modern neon transformer. Actually my buildings are always produce 4kV from a power supply which has 12-24V and 2-3 amps. But as you know neon sign transformers are powered from directly 220V AC mains. Are they buck 220V to lower levels and then boost again or are they directly boost it to 4-10kv. I will be appreciated for any idea to show me the way I should start with. Thank you for reading.

[Mads Barnkob]

Hi and welcome to HVF

A quick google search for "neon electronic transformer" tells me that there properly is two types. The dirt cheap and the actually safe properly engineered.

There seems to be a lot of copies of this design floating around, which means that some companies have likely copied that into a product.


Here is one that actually seems to be a http://www.evertron.net/ manufactured unit where the schematic is available at repairfaq

[davekni]

Mads,

That upper simple circuit is almost exactly what I've built in low-voltage versions (12-48V) for misc. projects such as a tiny eddy-current levitation demonstration and small induction heater.  I don't have the base resistors (which must be milliohms), and usually add reverse CE diodes.  It's a nice simple ZCS oscillator for series-resonant loads.

[ritaismyconscience]

I think the 1st circuit is used in most fluorescent lamps. Is there anything wrong with it (other than no fuse)?

[iraquois]

Thanks a lot for your answers. Second circuit is a little bit odd to me. Will it boost 120 VAC (170VDC) to 4KV with a turn rate of 56/75 ?

[davekni]

Most of the complexity of the second circuit is ground-fault protection for the HV output.  Looks like it shuts down if either end of the output arcs to ground (ie. through a person) or to the transformer case.

The 56-64 turns comment appears to be a range of values for the primary winding, perhaps for different models.  The turns ratio secondary/primary will be much higher.

[iraquois]

Most of the complexity of the second circuit is ground-fault protection for the HV output.  Looks like it shuts down if either end of the output arcs to ground (ie. through a person) or to the transformer case.

The 56-64 turns comment appears to be a range of values for the primary winding, perhaps for different models.  The turns ratio secondary/primary will be much higher.

I understand. Thank you.

So last one question. I see in a lot of schematics that there is a expression like in the picture which is marked with red circle. It seems it is for feedback but I don't know where exactly it is connected to. Does anyone have any idea?
https://imgur.com/CkNZIRq

[davekni]

For the example in your link, I expect it's exactly as the text says, a piece of copper foil glued to the case.  I suspect the case is painted or anodized.  The copper foil makes a capacitor to the case.  If one of the HV outputs arcs to the case, capacitance to the foil transfers that signal to the electronics, triggering fault shutdown.

Many SSTC designs use a similar "antenna".  For SSTCs, the antenna isn't for fault detection, but for oscillation feedback.  The antenna is a tiny capacitor to the SSTC output.  (The frequency and distance for SSTC antennas make them behave as tiny capacitors.  The distance is much shorter than a quarter-wavelength, so electromagnetic radiation is insignificant.  That's why I put "antenna" in quotes.)

[klugesmith]

In old fashioned mains-frequency NST's, the capacitive detector is for a feature called Secondary Ground Fault Protection.
Tesla coilers like to whine about them.
They sense when the complementary HV outputs are out of balance.

They became mandatory in the 1980's, I think, to reduce the risk of starting fires in buildings.
For example, if HV insulation failed and led to a persistent arc, to something like a nail.

Later the rule was extended to cover high frequency power supplies.
I guess that's the detail we are talking about here.  The schematic does say "plastic case".

https://iaeimagazine.org/magazine/2000/2000january/secondary-ground-fault-protection-in-neon-signs-and-outline-lighting/
"It is important to note that SGFP addresses a risk of fire and not a risk of electric shock. Injurious electric shock from neon circuits has not been reported as a field problem. The SGFP requirements are different when compared with ground-fault circuit interrupter (GFCI) requirements for line voltage circuits that are intended to address a risk of electric shock.

"The development of UL 2161 has been a complex technical issue that included combining the requirements for transformers and power supplies into a single standard, significant upgrades in both the transformer and power supply requirements and the development of requirements for SGFP. Many of the revisions to the Standard did not involve SGFP but were related to including requirements to address new technology and adding new tests, test conditions and construction requirements to address the field reports."

The B6 picture is my own, from a public place.  Anybody recognize it?

[iraquois]

For the example in your link, I expect it's exactly as the text says, a piece of copper foil glued to the case.  I suspect the case is painted or anodized.  The copper foil makes a capacitor to the case.  If one of the HV outputs arcs to the case, capacitance to the foil transfers that signal to the electronics, triggering fault shutdown.

Many SSTC designs use a similar "antenna".  For SSTCs, the antenna isn't for fault detection, but for oscillation feedback.  The antenna is a tiny capacitor to the SSTC output.  (The frequency and distance for SSTC antennas make them behave as tiny capacitors.  The distance is much shorter than a quarter-wavelength, so electromagnetic radiation is insignificant.  That's why I put "antenna" in quotes.)

Thank you a lot. Do you think this circuit work without the feedback loop. I will use a transformer winded around a old TV flyback core. So I don't have that much space. It seems to me they use this for SSTC. Am I right?

[davekni]

You need the oscillation feedback (C1+R3 and C2+R4).  You do not need foil tape nor the circuitry from C5 through R5.

I haven't figured out exactly why opto-couplers U1 and U2 are used, along with their associated circuitry.  Appears to change duty cycle of oscillation between positive and negative line voltage half-cycles.  Perhaps there's some advantage to skewed (not 50%) duty cycle, and the line cycle alternation is to even out power dissipation between Q1 and Q2 (the power FETs).  Anyone else know this circuit's operational theory?

[iraquois]

Thank you I see .

Anyone else know this circuit's operational theory?

Yes, it would be nice if someone explain it