Pcb acrylic application
suleymancskn , 04-30-2021, 02:28 AM
Hello Fedevel family.
I am curious about your opinion on a topic.
I have a design with DDR3.
After production, we apply acrylic to the integrated pins.
I am worried that this application will cause problems where I set impedance.
In my tests I saw that 5um top solder mask changed the impedance by 2%.
Therefore, I wonder if I should apply acrylic in places where I do impedance matching? Does anyone have experience in this matter?
Thank you
I am curious about your opinion on a topic.
I have a design with DDR3.
After production, we apply acrylic to the integrated pins.
I am worried that this application will cause problems where I set impedance.
In my tests I saw that 5um top solder mask changed the impedance by 2%.
Therefore, I wonder if I should apply acrylic in places where I do impedance matching? Does anyone have experience in this matter?
Thank you
robertferanec , 04-30-2021, 08:04 AM
I do not have any experience with this.
qdrives , 05-03-2021, 08:02 AM
Experience: no. Rick Hartley mentioned it in a training (at least the solder mask).
suleymancskn , 05-03-2021, 10:23 PM
Hİ @qdrives ,
Can you give a little more detail?
In my design, there is a solder mask on the ddr3 lines.
What I am asking is how the acrylic application on the solder mask will affect the impedance.
Or is there a different method against oxidation of DDR3 pins routed from a QFP sheath processor?
Thank you.
Can you give a little more detail?
In my design, there is a solder mask on the ddr3 lines.
What I am asking is how the acrylic application on the solder mask will affect the impedance.
Or is there a different method against oxidation of DDR3 pins routed from a QFP sheath processor?
Thank you.
qdrives , 05-05-2021, 05:05 PM
Well:
1) I do not do high speed design so I have no use for the exact details. I do remember a bit for the training.
2) He was compared to ENIG, as ENIG has nikkel and that is ferro magnetic. The nikkel has more (bad) influence on the signal than the solder mask does.
3) If you can estimate the thickness of coating and know the properties, then you could calculate the effect.
4) You could do some signal integrity simulations with a mismatch in the impedance to check if it will be problematic.
5) If you have both coated and uncoated then you would have to go in between the two.
1) I do not do high speed design so I have no use for the exact details. I do remember a bit for the training.
2) He was compared to ENIG, as ENIG has nikkel and that is ferro magnetic. The nikkel has more (bad) influence on the signal than the solder mask does.
3) If you can estimate the thickness of coating and know the properties, then you could calculate the effect.
4) You could do some signal integrity simulations with a mismatch in the impedance to check if it will be problematic.
5) If you have both coated and uncoated then you would have to go in between the two.
Use our interactive Discord forum to reply or ask new questions.