Specific Process Knowledge/Etch/DRIE-Pegasus/picoscope: Difference between revisions
No edit summary |
|||
| Line 119: | Line 119: | ||
* The shorter cycle times (Process A has 11 seconds full cycle and Process D4 has 3.2 second full cycle) emphasize the need for the faster process parameter sampling rate. For instance, a dep cycle with one measurement point for the C<sub>4</sub>F<sub>8</sub> flow is no good. | * The shorter cycle times (Process A has 11 seconds full cycle and Process D4 has 3.2 second full cycle) emphasize the need for the faster process parameter sampling rate. For instance, a dep cycle with one measurement point for the C<sub>4</sub>F<sub>8</sub> flow is no good. | ||
* The oscillations in pressure now last the entire dep cycle - it never stabilizes. | * The oscillations in pressure now last the entire dep cycle - it never stabilizes. | ||
* Optimizing a recipe to arrive at a minimum reflected power, for both coil and platen, using the Pro software datalogging may be somewhat illusory as it only picks a fraction of the spikes occurring at every change of plasma conditions. | * Optimizing a recipe to arrive at a minimum reflected power, for both coil and platen, using the Pro software datalogging may be somewhat illusory as it only picks up a fraction of the spikes occurring at every change of plasma conditions. | ||
What to do about this? | What to do about this? | ||
| Line 128: | Line 128: | ||
Changing the pressure in the plasma has a great impact on the RF matching conditions as illustrated for [[Specific_Process_Knowledge/Etch/DRIE-Pegasus/System-description#Why_RF_matching_is_extremely_important_in_the_Bosch_process | an oxygen plasma]] elsewhere. The obvious place to start is therefore the pressure. Also, it barely makes sense to claim that the pressure is 20 mtorr in the dep phase if it oscillates as shown for the Process D4 above. | Changing the pressure in the plasma has a great impact on the RF matching conditions as illustrated for [[Specific_Process_Knowledge/Etch/DRIE-Pegasus/System-description#Why_RF_matching_is_extremely_important_in_the_Bosch_process | an oxygen plasma]] elsewhere. The obvious place to start is therefore the pressure. Also, it barely makes sense to claim that the pressure is 20 mtorr in the dep phase if it oscillates as shown for the Process D4 above. | ||
Compared to many other vacuum systems the pegasus tools require a massive turbo pump | |||
With a huge turbo pump | |||
The pressure in the process chamber is regulated by the throttle valve or APC (Automatic Pressure Valve) that sits in front of the turbo pump. With a large valve plate sitting across the opening between the process chamber and turbo pump, this valve is capable of changing the pumping speed of the system. | |||
By precisely changing this opening, it can | |||
It changes the opening | |||
capable of opening and closing very precisely and very quickly. If it | |||
If a process is running at some | |||
The oscillations are caused by the fact that we use the 'fixed pressure' mode. | The oscillations are caused by the fact that we use the 'fixed pressure' mode. | ||