Hi bcrew13.
Maybe you said it yourself, "problems you know pretty well." I guess part of the question is "How well is pretty well?"
Here's one thing I have seen. A person does a bunch of problems on a timed basis. In reviewing them the person sees that while he or she made mistakes, he kinda knew how to do the problems. So he feels pretty good.
Well ... the thing is that knowing conceptually how to handle the problems is NOT the same as really knowing how to do them and get them right.
Some careless quant errors are truly basically silly mistakes. Often people make them when they are doing calculations and get most of the way to the answer. Then, excited that they have done the lion's share of the work, they blow the last few steps or choose an answer that does not match what is being asked.
Other supposedly careless errors are actually the result of not really being that good at the types of problems one is getting wrong.
Whatever is going on with you, you can increase your quant score significantly by slowing down when you do practice problems and really seeking to achieve a high hit rate, maybe at least 85%. If you don't get at least 85% right untimed, how are you going to score Q49 on a timed basis?
Also, keep this in mind. The GMAT quant section is not a test of whether you know certain math concepts. It's test of whether you can get to right answers consistently. Maybe that seems obvious, but I know from my personal experience that at first one can be in a mindset that says, "I need to know about how to do these problems," when actually one more needs to be in a mindset that says, "I need to be good at getting right answers."
So that's much of what you need to be working on, getting good at getting right answers.
One resource I have seen people use to increase their quant scores is the quant part of the question bank at
https://bellcurves.com/. You can access it by going to the GMAT section and setting up a practice account. Then you just work through questions in areas in which you are weaker, section by section. If you do that for a month, and really focus on getting right answers, you will likely hit that target, but you really need to build depth of understanding of and develop accuracy in each category. So take as long to do questions as you need to in order to get that high hit rate.
To figure out which categories to work on, go back over the CATs you have taken and see what you consistently get wrong and what takes you the longest. Then become an expert in those things. Once you have done that, I bet you can find more areas that you are pretty good at but not Q49 good at.
Many of the explanations in that question bank are not that good, but you can post questions here and also just look at the explanations already here to get great ideas for how to handle all types of quant questions.