Is it possible and was it a practice to use SON-I lamps on dimming circuits? For example a SON-I 70W in combination with a two tab 50/70W ballast and a switching unit like the Philips Chronosense or Tridonic ZRM U6M or ZRM U6L? I wonder if somebody have heard of this because circuits like this but with superimposed ignitors and SON /E lamps were quite common here and perhaps in Germany but I haven’t seen it on SON-I lamps.

When switching over from high to low current, there is a significant risk of the arc getting extinguished.
First the switching transient may generate irregularity/gap in the supply current, which may lead to the lamp failing to reignite after that.
Second the reduction of the current tends to lead to cathode temperature reduction, which tends to increase the cathode fall. At the same time the arc tube temperature, so the gas pressure, does not decrease that fast, so there happens to be a moment with higher arc voltage drop, too high to successful reignition, so arc extinction.
The superimposed ignitor then uses to intervene, generate few HV pulses and helps to reignite the arc just when it fails to reignite by itself, so prevents the arc from extinguishing.
With SON-I you do not have that support anymore.
Plus SON-I (and equivalent) do have bimetal controlled ignition aids (antenna wire or a conductive strip on the outside of the arctube,...), which when connected would lead to fast arctube degradation, therefore the bimetal switch disconnecting them once the lamp ignites and so heats up. By operating that lamp on a dimmer circuit, the arctube operating temperature may drop so much the bimetal connects the starting aid back and allows the excessive degradation to happen. Most lamps will have that bimetal threshold temperature low enough, but the point is, it has quite a large production variation, so still a significant amount of lamps may have that threshold too high for the dimmed operation.

Thank you Medved!

I was thinking the same that for a long term it would not work properly with this lamps. The relays are approved for SON/ E lamps and HPMV but there is no reference to the SON/I lamps in the literature.

Thank you Medved!

I was thinking the same that for a long term it would not work properly with this lamps. The relays are approved for SON/ E lamps and HPMV but there is no reference to the SON/I lamps in the literature.

If I rememver well, Osram explicitely listed the selfstarting HPS along lamp types explicitely not compatible with any dimming method at all...
The pulse start HPS were allowed to be dimmed down to 70% by voltage reduction with max 2% per step in min 5 minute interval (written for European series choke ballast market) or down to 40% by impedance and/or electronic power regulation ballasts.
MH were specified down to 60% power by electronic ballast only (I did single step down to 40% and the lamp tends to extinguish), MV down to 40% by impedance or a phase cut.
In any case all lamps must be run at 100% for at least half an hour before any power reduction could be applied.
Hope I remember these numbers well, but they definitely were in such relation...

So far I know is that HPS can be dimmed down to 50% on appropriate magnetic gear and down to 20% on electronic gear (Philips DynaVision)