Advice for hiring EE to take through-hole prototype to next stage

[JohnnyMalaria]

[I can't find a more suitable place for this]

This is a question about how to find and engage a professional EE to take a through-the-hole prototype to the next stage.

For the past year and a bit I have developed a lab instrument that uses a combination of off-the-shelf modular components (e.g., 4 channel DDS for RF generation, 2 channel LF function generator, 8-channel ADC etc which are all interfaced to a Windows laptop by USB) and my own through-the-hole LF circuits to perform analog amplification, filtering and IQ demodulation of a photodetector signal.

It's at a point that I consider to have demonstrated proof-of-concept and so I have a good idea of what the specifications would be for a professionally-designed next version. It currently is housed in a 3U case and I want to reduce the size significantly. I have already provided private demonstrations to interested companies at trade shows and wish to do so next year which is why I desire something more professional.

I'm pretty sure that much of the functionality could be done with an FPGA. I tried to dabble in this but the learning curve was horrific and I haven't the time to educate myself sufficiently to come up with my own robust solution. Likewise, I don't have the time to learn surface mount techniques and PCB layout especially because there is some RF circuitry involved. I'd require control via USB using Python code that I've written. So, there's quite a varied functionality required.

Where on earth do I begin to find someone? I know that's a pretty broad question but I'd value any pointers you can provide.

Thanks.

(I'm in North Carolina, US)

[tautech]

Tim's the bloke for this.
https://www.eevblog.com/forum/profile/?u=92330

[CopperCone]

prepare to pay out the ass if you want someone to do a FPGA. I would recommend ad AD9106 if you can use it lol

FPGA is like the thing you learn along with DSP if you want to have one of those jobs where no one knows what the fuck your doing and your paid alot

And for specifications I assume you have things like noise, linearity, phase related measurements, power consumption, emi requirements (does it need to fall into a class A, B), etc figured out? If you don't have a concrete spec written with those kind of mathematical numbers then you can get fucked out of your money hard. Otherwise if your spec is too tight the engineer should write down what is possible with the budget/technology you have so you know what you wanna pay for, otherwise you might get a slacker, you probobly wanna get a evaluation that you pay for to make sure your within theoretical and budget limits then pay for specific design labor. Good to pay separate so you don't feel bad asking someone else if the design specifications they think they can do are wroth the money.

[T3sl4co1l]

How much bandwidth, size, budget, etc. are we talking here?

Tim

[JohnnyMalaria]

How much bandwidth, size, budget, etc. are we talking here?

Tim

I always struggle with the budget question - it comes down to what exactly can be done like when you go to buy a car with a budget in mind and the see something nearly twice the price that you didn't know existed and is the thing you absolutely must have

Anyway, breaking the system down into its different sections:

RF

This is what I asked about a few weeks ago and you kindly provided an almost perfect fit.

2 x RF power amps delivering +32dBm into a 50ohm load. Frequency = 40+/-1MHz. The inputs come from a DDS generator. The one I use at the moment generates about 1Vpk sine.

Photodetector signal

The output is a broadband signal due to lots of noise in the MHz region. The signal of interest is typical 1-50kHz. The AC component ranges from approx. 1 - 10mV at has a large DC component (>1V). Currently, I pass the signal through an 8-pole Sallen-Key bandpass filter (2kHz/30kHz) with +40dB gain.


Quadrature detection

The filtered and amplified detector signal is demodulated by two quadrature multiplier ICs and low-pass filtered before going to an ADC sampling typically between 8kHz and 32kHz.

DUT impedance analysis

There's a low frequency square or sine voltage signal from a FeelTech DDS that ranges DC-1kHz and up to 20Vpp connected across the DUT whose impedance ranges from ~500kohm down to ~10ohm. Current is measured with a uCurrent device. Right now, there are six signals fed to the ADC which has to be simultaneous sample-and-hold. The digitized data are transferred via USB in the background to permit real-time analysis.

My primary goal at the moment is miniaturization.

[T3sl4co1l]

Do you have a complete schematic for that, and want it on one PCB?  Would you want any optimization of the schematic too?

And yeah, a lot of that could be integrated in an FPGA or DSP if you want it really small, but that takes a lot more coding effort, which will drive up the cost.

Budget would be a sliding scale, anywhere from a few thousand dollars for a rough PCB layout, to maybe a hundred thousand for a very in-depth en-shrinking-ification including firmware.  All depends on how much you need.

Tim

[Ice-Tea]

I'll happily second Tim as a recommendation.

And I'd also toss in you need to have an idea about how many you plan to sell and what a fully populated board can cost to still make economic sense in certain batch sizes.

[nctnico]

[I can't find a more suitable place for this]

This is a question about how to find and engage a professional EE to take a through-the-hole prototype to the next stage.

Where on earth do I begin to find someone? I know that's a pretty broad question but I'd value any pointers you can provide.

Posting an advertisement in the 'jobs' section would be a good start. You will need to have quantifiable goals though like size and power consumption. The best thing to do is to specify functionality and not the components someone should use. Perhaps there is a better way you didn't think of.

[JohnnyMalaria]

Thank you all. Now I have a better idea of how to proceed and I'll start thinking about next steps.