I think something that is able to produce a lot of heat needs a good thermostat to keep it under control. A 10% swing in a 60 watt bulb is not going to be seen as dramatic as that of 120 watt one, and 10% is a good number for electronic components. I would think about a good thermostat and have another go with the heating rope. I am going to try a 130-watt heater with a fan built in, used in compact electric fire places. The heating element will be on a dimmer and an electronic thermostat to control the heat produced if the swings get to wild.
However I'm on day 3 now with my old incubator and I have yet to see more than .3 degree drift. I have 2 40-watt bulbs, but there the long slim ones (dollar store), if one blows then the other will adequately heat while I change the blown one out, I had 2 25-watt ones but one alone did not work so I swapped for the 40's, so far so good. I would also check your vents, I found a better vented box held temperatures better when I used the dimmer switch and the fan, without the fan I had more stability with less vents.
My personally theory why you had trouble is because the dimmer switch is not very accurate as no one cares if you have the light producing 30% or 40% of it's lighting ability, it's just dim, but when you measure the 'dimness' then the differential is quite apparent.
However I'm on day 3 now with my old incubator and I have yet to see more than .3 degree drift. I have 2 40-watt bulbs, but there the long slim ones (dollar store), if one blows then the other will adequately heat while I change the blown one out, I had 2 25-watt ones but one alone did not work so I swapped for the 40's, so far so good. I would also check your vents, I found a better vented box held temperatures better when I used the dimmer switch and the fan, without the fan I had more stability with less vents.
My personally theory why you had trouble is because the dimmer switch is not very accurate as no one cares if you have the light producing 30% or 40% of it's lighting ability, it's just dim, but when you measure the 'dimness' then the differential is quite apparent.