So you need *either* a high accuracy 90 deg phase shift *or* a high accuracy quadrature sinewave oscillator, as just part of the overall design.
Unfortunately, both are difficult to do in the analog domain with even 1% accuracy over a single decade frequency range and you want a 4 1/2 decade frequency range!
It is of course possible to generate the quadrature sinewaves digitally, and the max. frequency is low enough that it should be possible on a fast MCU running software DDS driving a couple of DACs, but how much filtering will then be required to meet your 'no noise' 'requirement'?
"... but for you guys its easy"
![ThumbsDown :--](https://www.eevblog.com/forum/Smileys/default/icon_thumbsdown.gif)
Then there's the issues involved with speccing the power amplifiers. 12V to 24V supply and a '200W' inductive load. No impedance data provided.
At a first guess, firming up your requirements to get something an EE engineer can quote for would be several hours work, mostly face-to-face clarifying your requirements, helping you to distinguish between ones that are essential, and others that are simply nice to have, and putting numbers on the specification limits. Several hours work for an experienced consulting EEE in a 1st world country is likely to be in the ballpark of $500 before the actual design process even starts .....
Once you've got proper engineering specifications, you'll still have to commission the design. Some desperate 3rd world undergraduate might knock out something that half-works for a similar fee, or they might simply vanish with your down-payment 'for parts and materials'.
If you want a qualified engineer as project lead, and reasonable assurances the project will produce working deliverables, you are looking at an order of magnitude or more higher costs.
I'm therefore not surprised Stray Electron has been pushing you hard in the direction of quite reasonably priced off-the-shelf modular test equipment.
However, frequently people not familiar with a particular field of engineering incur excessive costs because they are narrowly focused on implementing 'one true solution' they have dreamed up when anyone with even technician level experience in that field would make very different choices. I'd give 50% odds you have an an
XY problem, and giving a 'big picture' description of the whole system, what it needs to be able to do and how this quadrature power driver circuit fits into the 'big picture' would let us offer cheaper/better suggestions to get similar results.
The only reasons *NOT* to tell us more is if you intend to patent the system, or if its commercially sensitive and you cant get a release from your legal dept. to discuss it publicly, or if its subject to any government security classification.
N.B. I don't 'have a horse in this race' as I'm not in any way interested in taking on consultancy work in this field.