Some decades ago, I developed a high frequency stabilised power supply for a range of Osram spectral lamps. These are arc lamps that run from a fluorescent lamp type choke and starter at 50Hz. The power supply ran at 20KHz but used LC resonance at about 70KHz when starting.
My thought was to make the power supply work wider voltage 85 to 250Vac by running at about resonance, it would also start more quickly. And because the lamps are rated 10W to 60W not require PFC controller.
My concern though is that because the LLC is optimised for zero current switching, the switching time would have jitter? This jitter would affect the quality of the light output perhaps, or it may be trivial compared to the little flicker that the arc lamps tend to have anyway.
I have uploaded the output circuit and the controller's low voltage power supply derived from the lamp's sensing current transformer but not the controller, which is in any case a conventional current mode controller with no slope compensation of the 1980s and 1990s.
If I am on the wrong track, please advise an alternative controller?
Solved! Go to Solution.
Thank you, Jozef,
You have given me some interesting links. It was not eye sensitivity to flicker so much as electronic device sensitivity to flicker was my concern. The power supply that I made previously after many integrations is robust and resolved the issues very well. Its efficiency is poorer than it could be, though.
You mention a Mercury type power supply, is that;
https://www.nxp.com/docs/en/brochure/75017226.pdf
I can see that to start the lamp, the controller must work at LC resonance. When running, the frequency can change (be a little inductive, the datasheets explains). I can not see if they will do that?
Anyway, I shall consider your recommendations and may pursue other avenue?
Dear Jozef,
I have uploaded the Osram datasheet from their website. There is also a YouTube video of one of these lamps starting and running on my blog see the link below.
https://blog.andrew-lohmann.me.uk/2018/07/electronics-high-frequency-arc-lamp.html
Thankyou
Andrew
Dear Andrew,
thank you for the datasheet and additional information. I see the spectral lamp type is similar to the fluorescent lamp type.
From the datasheet you shared, you need to supply the Sodium spectral lamp with Alternating Current, so you need an AC/AC controller. Unfortunately we do not have any such type. We have plenty of AC/DC controllers, so we can offer you one for the Mercury type, which is the only one, which can work with both AC and DC current.
Please check this link for our AC/DC controllers portfolio. For the powers from 10W to 60W a flyback type would be more suitable, than the LLC type. For example TEA1755T is suitable for power supplies from 0W to 250W. E.g. the TEA2017 LLC and PFC combo is suitable for powers from 90W to 1000W. For lower powers the LLC controllers have low efficiency.
Regarding the jitter, yes every power supply produces some jitter, but human eye will not be disturbed by that. Current converters works in kHz, latest in MHz. The human eye will not perceive the jitter.
"The human eye can see electromagnetic radiation roughly in the frequency range of 400 to 800 teraherz.
The eye has trouble decerning modulation of the radiation roughly above 60 to 70 Hz, but it depends precisely where in your field of view the modulation occurs.
Light modulated above the cut off frequency is not invisible. It is perceived as steady light.
That is the very simplified answer. There is a lot more to it, and no you can't see 120 Hz, that is a strobe effect and probably a conspiracy to sell high speed monitors to people who think they know what's going on but don't. It is the result of high speed shutters and poor CGI techniques, but let them buy and enjoy, their graphics cards."
With Best Regards,
Jozef
Thank you, Jozef,
You have given me some interesting links. It was not eye sensitivity to flicker so much as electronic device sensitivity to flicker was my concern. The power supply that I made previously after many integrations is robust and resolved the issues very well. Its efficiency is poorer than it could be, though.
You mention a Mercury type power supply, is that;
https://www.nxp.com/docs/en/brochure/75017226.pdf
I can see that to start the lamp, the controller must work at LC resonance. When running, the frequency can change (be a little inductive, the datasheets explains). I can not see if they will do that?
Anyway, I shall consider your recommendations and may pursue other avenue?
Dear Andrew,
the UBA2016/15/15A would be suitable to supply the one spectral lamp, which can be supplied with DC current. These have the DC output. Unfortunately they are no longer manufactured and we do not have any replacement.
With Best Regards,
Jozef
Dear Andrew,
I do not have this information. I apologize for inconvenience.
With Best Regards,
Jozef
Dear Andrew,
could you please share a datasheet of the Lamp you want to create a power supply for? I am not sure if I understand the meaning of the Arc Lamp you are referring to. Do you mean the Fluorescent Lamp?
With Best Regards,
Jozef