A PhD is a good way to enrich your understanding of a particular topic, but in terms of increasing employability, the knowledge you gain is less useful than the skills you develop. Skills such as critical/logical thinking, planning, evaluating ... etc. These are more useful because the knowledge you acquire will have a half life 2 years according to
http://www.newelectronics.co.uk/electronics-blogs/engineering-knowledge-half-life-when-is-it-okay-to-not-know-the-detail/20564/. Its obsolete very quickly.
Trying to leverage a PhD in industry implies that you want to be employed by a leader or innovator in the field of you PhD. Consider which employer/industries you would be able to target with your chosen topic of study.
How many are there? and what would you need to change in your life to work there? [relocate, ...]
Compare that with all the other EE employers/industries who need their engineers to be jack of all trades. 90% of a product will use hardware that is going to be considered of the shelf. A SMPS, a controller, a CPU like embedded architecture. If you can't find it off the shelf, a standard EE will just use app notes or crib from dev boards for their chosen device. Suppliers like intel offer free services to review and check your motherboard or CPU interfaces. ST Mirco has design centers that will do the same for anything related to what ST sells (I'm about to use them to review bluetooth antenna designs, inside enclosures with existing electronics).
In the 1% of cases where standard EEs get stuck, we just hire a contractor who specialises in SMPS, ADCs , CPUs or whatever field is required (for as little as a month).
All the contractors I've come across have decades of experience in that field. A newly PhD 'd EE is no comparison.
Hiring such a contractor for 1-3 months in 1% of cases is way cheaper than employing another standard EE full time, let alone paying the extra that's usually expected of someone with PhD.
As someone who is recruiting for someone to complement my skillset, I regularly review resumes/CVs. After a few failed interviews with PhD applicants, I find myself naturally favouring someone with proven experience over a new PhD with no experience.
The 2 years of experience you already have + 3 years (min) of PhD study will probably make you less employable than someone who has 5 years of experience. This is especially true for the majority of employers who want someone general.
As a final note; consider what happens when you get employed by the leaders or innovators in your chosen topic and 2-3 years later, the industry has moved on to the next thing. How quickly will you find yourself sidelined or redundant and how hard will it be to find a new job?