# Can P gain too high cause roll of death?

I've been sperimenting with PIDs on my quadcopter and I've noticed that pitch P = 65 causes double front and back flips which looks like a roll of death (or in this case a flip of death) and lowering pitch P to 60 solves the issue. Why is that? Can I mitigate this with filters or D gain?

I've also noticed that certain values of P and D mix well and make the quad smooth but if I bump up the I gain to twice as high as my P gain then my quadcopter bounces back. Is this reasonable or there's a problem somewhere else?

• I didn't say that P is pitch, I've said that pitch P is... Read carefully. I know what P in PID stands for. Useless comment by the way – Fredo Corleone Apr 8 '19 at 10:26
• You used P in two different ways in the same paragraph, it's not surprising that it's easy to misinterpret what you are saying. It might be clearer if you used $\theta$ for pitch rather than P. On Robotics we are fortunate enough to have MathJax support enabled, so it's really easy to just add $\theta$ instead of P. For a quick tutorial, take a look at How can I format mathematical expressions here, using MathJax? – Mark Booth Apr 8 '19 at 13:16

Yes the proportional gain can cause overreaction. As it is only concerned with instant correction a larger gain value of P can make it over compensate. Basically I like to think of it as, in any instant point the grater the difference between the actual position or signal the more you have to accelerate to get there as quick as you can.

This can be corrected by using a derivative gain. The D gain is only concerned with compensating for the rate of change. This make it good to reduce the overshoot of your system (preventing death rolls).

What you are noticing in the last part of your question is a natural ratio of the gains. A good method to find those well performing ratios is to use the Ziegler–Nichols tuning method.

Also, if you are interested in control systems I would strongly suggest looking at Brian Douglas' YouTube series on control systems.

Actually i'm not fan with the PID algorithm in this field especially drones and rc plane. Think about fuzzy logics.

https://searchenterpriseai.techtarget.com/definition/fuzzy-logic

• I'm not sure this actually answers the question asked by the OP. – sempaiscuba Mar 30 '19 at 14:13
• It doesn't answer my question but I'd like to know what you mean. :) Why you don't like PIDs? – Fredo Corleone Mar 31 '19 at 22:15
• Welcome to Robotics user11249082. Thanks for your answer but we prefer answers to be self contained where possible. We are also looking for comprehensive answers that provide some explanation and context. Very short answers cannot do this, so please edit your answer to explain why it is right as answers which don't include explanations may be removed. – Mark Booth Apr 8 '19 at 13:22