[labnetwork] SPTS (STS) Advanced Silicon Etcher

Tue Aug 18 04:13:03 EDT 2015

Hi Richard,

The jumpers as far as I remember effect the voltage to the motor and thus the speed (there are also some POTS that do the same). Normally they would have been setup in the factory by STS and wouldn't normally need to be changed (the AMC modules were used on many different matching systems and were setup to match the system). You could however fit the links if they are missing and see the effect just make sure you do it to the right AMC.

It sounds like you are trying to run a switched process from your description. If you have a problem with the coil matching then I would just try and optimize that part first so make a recipe with just the coil RF on. This is how I would tackle this...

*         Start by connecting a 50 ohm load to the RF generator. Take the RF cable out of the matching unit and plug it direct on the load. You should see the forward power you ask for and 0 Watts reflected. This proves the RF power supply is fine.

*         Without any power running open the coil matching unit. Remove the thick HV connections to the capacitors and connect a capacitance meter across each capacitor.

o   Drive the Coil and Load capacitors to 0. Note the capacitance value

o   Drive them to 50 and note the value

o   Drive them to 100 and note the value

o   Check the two Tune caps are in sync. IE you are getting roughly the same values. Make sure both the Load and Tune caps are spanning near enough the full range of capacitance

o   Any problems above then let me know and I can help you adjust this

*         Assuming the capacitors are OK. Make the HV connections again and re-assemble everything. Set the load and tune start positions in the recipe to 50% and run the process. If it doesn't strike or you can see the capacitors are shooting right to the extreme then press hold before it times out. Some tools have a software bug and when the RF doesn't strike and an error occurs the capacitors don't return to the start position. This means if you hit resume you are starting not from the start values you wrote in the recipe but from where you ended when the error appeared. You can get around this by trying to strike but pressing hold before the error appears. Try changing the start values (you can do this with the process on hold), 60 / 40 or 40 / 60. If nothing works then continue...

*         Try to find the capacitor match position by manual turning. Change the recipe so the capacitors are matched manually. Now when the process starts you will get a high reflected power but you can manually begin to tune the capacitors. Beware this is tricky and requires a lot of practice and you should be fast, a high reflected power for a sustained period can damage the RF power supply. I would start by moving the load cap to 40, watch the ref power carefully if it drops then small adjustments until it begins to rise again, keep it at the lowest ref power value. Do the same with the tune adjust it either higher or lower and watch the ref power if it goes up turn it the other way... it is not easy but you optimize the plasma this way. You may need to go back and forth between the load and tune making small adjustments (a bit like tuning a radio) When you know the values for 0 watts ref then in the recipe enter a 5% lower value for the load and 5% higher for the tune (remember to set the matching to automatic and restart the process). This way when the plasma turns on the capacitors move quickly to the optimum position and don't overshoot. (If the 5% values don't work then try the actual values)

I think I will stop there as there is a lot of info, I have also included a RF faults doc (from STS) but I'm not sure how helpful that is and also a slave motor PCB schematic but I'm not sure if that is the one in your system.

One final thing, I don't know your process but it can also be difficult to strike the plasma at a really low pressure. Try an initial strike pressure of around 10-15mTorr.

Best regards,
Roy

From: Richard Battaglia [mailto:rlbemc at rit.edu]
Sent: 17. august 2015 17:09
To: Roy Cork
Subject: RE: [labnetwork] SPTS (STS) Advanced Silicon Etcher

Roy,

What do the jumpers on LK1&2 do?  What the unit does is seek to find a good tune point on the coil, which it does get to (about 12 ref).  Then when the platen comes on the coil loses its tune and searches and searches never actually finding it.  I can watch the unit while it is running and the coil jumps from 12 ref to 120-335 ref.  It then lets the platen run, which only tunes from 1-12 watts reflected on 16 watts forward.  The platen does finally reach tune but after approximately 10 seconds and then that part of the process is done and it shuts off.  After that time the coil is still on, and if it has a descent tune, will continue to run to the next step.  If the coil has not tuned well while the platen was running it will sit there and run for a bit with approximately 483watts reflected with 520 watts forward.  After about 7sec it will time out and give an error.

This was the original issue I had when I started all this that let me to the Tune Slave Cap and the fact that it did run smoothly but instead "steps" in one direction and runs smooth in the opposite.

I hope this explains it a little better as to where I stand right now.

Thanks
Richard Battaglia
From: labnetwork-bounces at mtl.mit.edu<mailto:labnetwork-bounces at mtl.mit.edu> [mailto:labnetwork-bounces at mtl.mit.edu] On Behalf Of Roy Cork
Sent: Wednesday, August 12, 2015 9:25 AM
To: 'labnetwork at mtl.mit.edu' <labnetwork at mtl.mit.edu<mailto:labnetwork at mtl.mit.edu>>
Subject: Re: [labnetwork] SPTS (STS) Advanced Silicon Etcher

Hi Richard,

I'm not sure if this is the right way to reply to your message in the forum but I'll try (for some reason your message isn't appearing in my thread).

I have attached the drawings for the AMC modules (sorry they are not totally clear but they are photocopies of photocopies). There is a block overview along with the capacitor adjust PCB (located on the side of the coil MU) and 4 drawings of the module, the important one is Page 3 module_main. In my experience that is where things go wrong.

You wrote "I watched the voltage going to the motor and it is in fact getting a pulse as opposed to a constant voltage.  That being said, Tune Cap 2 looks like it is stepping more than just adjusting slowly.  I watched the resistive pot and it "steps" past where it needs to be and then tries to adjust back to the resistive value it was looking for.  The problem here is that it always overshoots the needed resistive value so it never actually adjust correctly. "

It could be the gain for the motor control is set too high. That is it accelerates too fast and overshoots the setpoint, turns back goes past again and so on, this kind of looks like its pulsing. Check on the AMC analogue board that jumpers LK1 and 2 are set correctly (also try the opposite to what they are and see what effect it has)

Setting LINK's
                                    LINK1 TUNE CAPACITOR       IN for VACUUM TYPE
                                    LINK2 LOAD CAPACITOR       OUT for AIR SPACED or SF6 Types

Also take a look at the small POT PCB on the side of the coil MU (drawing attached), this can also be limiting the range of the movement (try comparing both load and tune POT values).

Finally if the motor is always slowly creeping I have also seen a problem with a leaking OP amp. Try changing the components (IC's) in the load circuit for ones in the tune circuit and see if the problem swaps over.

Let me know how you get on. I have some more docs I can dig out if you need more help.

Roy
DTU Danchip
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://mtl.mit.edu/pipermail/labnetwork/attachments/20150818/d4bc2086/attachment.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: RF Faults and rectification.doc
Type: application/msword
Size: 49664 bytes
Desc: RF Faults and rectification.doc
URL: <https://mtl.mit.edu/pipermail/labnetwork/attachments/20150818/d4bc2086/attachment.doc>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: CD310170.03 Slave Motor Drive PCB.pdf
Type: application/pdf
Size: 342467 bytes
Desc: CD310170.03 Slave Motor Drive PCB.pdf
URL: <https://mtl.mit.edu/pipermail/labnetwork/attachments/20150818/d4bc2086/attachment.pdf>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: amc notes.ppt
Type: application/vnd.ms-powerpoint
Size: 900096 bytes
Desc: amc notes.ppt
URL: <https://mtl.mit.edu/pipermail/labnetwork/attachments/20150818/d4bc2086/attachment.ppt>