observation · computed
RETRACTION: the 640 Vpp / 150 kHz / 60 W / 2 A figures are NOT SOS Productions' bench — do not attribute
SOS Productions stated explicitly in #neuralstan20-testing on 2026-09-17: "FTLOG thats not my bench. Delete that data in reference to me."
RETRACTED, do not reason from these against SOS Productions ever again:
- 640 Vpp
- ~60 W, ~2 A
- 150 kHz
- "linear gas-per-watt ratio"
- 32 V LC supply
- flat-core 1,262 mH / 2 uF
- EC ferrite vs air-core choke comparison
These appeared in a profile summary, not from anything SOS Productions reported in conversation. Origin unknown; possibly cross-contaminated from another builder's numbers or from a design-lane summary. They must not be quoted back to him, used as a starting point, or carried into any calculation attributed to him.
RULE REINFORCED: profile/summary figures are second-hand until the user confirms them in the live conversation. Ask before reasoning from them; when a user disowns a figure, it is dead permanently, not just for that turn.
No replacement bench data for SOS Productions is currently held. Anything needed must be asked for fresh.
Basis
- Recorded
- Published
- 23 Sep 2026
- Stanbot
- v3
- Source Ref
- SOS Productions, Discord #neuralstan20-testing, 2026-09-17
- Notebook Id
- 3409
retraction attribution sos-productions bench-data