Selecting your Electronics Design House

Electronics design is relatively one of the less expensive parts of custom design & engineering work. For instance, the cost of design and tooling for a good plastic injection moulded case with keyboard may exceed the design cost of the electronics inside. Having said this, all the good electronics design staff cost real money - the client is paying for the expertise and experience that the designer has to offer.

It is quite possible to contract and employ an Electronic Designer without being especially knowledgeable in electronics. Insist on understanding the principles involved, even if not the precise details of operation. Experience demonstrates that there are two breeds of design engineers - the desirable type are fountains of information, the undesirable types hide information to make themselves indispensable. If you employ a "Closed-Book" engineer, you will find it difficult to ever un-employ him. While there are occasionally valid reasons, practices which highlight an undesirable engineer or design house are:

Security through Obscurity Not telling people about something doesn't necessarily make it secure. For instance, perhaps a manufacturer might make a device with a "secret" communications protocol, and think that by not making it public no other company will be able to retrofit equipment to their proprietary system. (In the first place, they would be alienating customers and potential third party support in the industry, but that is not the point) But if there was no actual encryption or other security, all it takes is someone with the right knowledge, and perhaps a little luck, to come along and find the back door, and the security is lost. Security through Obscurity is very little security at all. It is the sort of security a dilbert-type manager dreams up. In the field of Cryptology, "Security through Obscurity" is widely questioned as a principle. Most such systems have proven to be poorly designed. Perhaps the best model is the 20 Year old Data Encryption Standard (DES). Complete details of DES operation have been available since its release - for 20 years it was algorithmically secure. As of 17 July 1998 brute force attacks have been developed that can crack a ciphertext-plaintext pair in about 56 hours - but that is not actually a failure of the algorithm, DES has stood the test of time brilliantly. For security choose algorithm or patent - not obscurity.

• Circuit diagrams unavailable
• Source code unavailable or poorly commented
• Numbers rubbed off of integrated circuits
• Parts lists unavailable, or without parts costs
• Factory test and calibration not described (if complicated)
• Secret principles - "Security through Obscurity"

Electronic Design houses can be best found using google or other search engine [You've found one], through some specialist magazines, and by asking around for recommendations. There used to be "Electronic Engineers," in the Yellow Pages, but that category is now deleted in some editions due to the small number of entries.

In Selecting your design house, use these criteria:

1. ABILITY - Check the designers ability to produce and deliver by examples of his work. Check the examples are his own work. Don't be fooled by flash tools (they are like fast cars), buzzwords, or bravado - what counts is what gets delivered.

2. TIMESCALE - Turnaround time for recent projects. Simple microprocessor circuitry should take 8-12 weeks from spec to prototype - longer if a significant amount of firmware is required. To get faster delivery usually results in exponentially rising costs, - or cut corners.

3. BUDGET - Your design house may be reluctant to estimate costs on your project before having all the details, but they will often be more forthcoming with the amount charged for another, similar sized job completed recently. For most work we quote, and considering the history that engineering work has of running over time and budget, that provides you, the client, with a lot of security.

4. COMMUNICATION - Effective communication is absolutely essential to your project being successful. Find a design house that speaks your language. Spend the time to work over your specification with them. If you have a clear spec and like what you are reading, then why not use an Australian Electronics Designer? We will make the effort to keep in touch by phone, we can ship prototypes anywhere in the world in 48 hours, and we spell out everything you could possibly want to know about us in black and white light blue, right here on this website.

5. DEPENDABILITY - Will the company or individual be around to provide you with engineering support in 5 years time? It costs quite a bit to get new engineers up to speed on a previous project - it is worth investing your work with a reliable vendor. (Actually, its quite funny - on several projects we have outlasted all the client in-house engineers - they have all moved on to other careers, and the sub-contractors are the only members of the original design team still available to the management!)

6. SPECIALTY - Industry Specialization (e.g. go to an eft-pos machine manufacturer to get an eft-pos PCB designed) is useful, but can pose problems - will there be a conflict of interest? A good design house works across a large range of industries and projects.
   Of much greater importance is whether the design house is capable of producing the project technology you are after. Look for design houses with experience producing products with a similar level of technology - e.g. SMT, PC based, realtime acquisition, battery powered, or whatever else may be appropriate for your task.

Hardware design - 1997

Normally you would start with a list of subcontractors, rank them according to your criteria, and then contact each about your project. This may involve email, but would usually involve a phone call as well. Keep the information provided brief, but not deceptive. Avoid sending a subcontractor a 100 page document and asking for his feedback - you may be astounded by the silence! From the responses, and based on your ranking, you should pick a short list.

When you phone, make sure you are speaking to the guy who will be doing most of the work. Unfortunately, if you are speaking to a Sales person, they are likely to be quite good at telling you what you want to hear, and you won't really be getting the feedback you need. It is quite important to engage the engineer, and prove your commitment to the project also. During that inital phone call, unconsciously, the engineer is interviewing you as much as you are interviewing him.

Engaging an Engineer If your work involves much more than just copying a circuit diagram and laying out a quick PCB, it would likely benefit from a bit of creativity and flair. There is passion involved - the sort of guys who do this work do it because they enjoy it - they like taking the concept that you have, bashing the unworkable bits out and then turning it into something new, clean, elegant and complete - then switching it on, and watching it work. Their success is your success - they have made the design operate, you can get out there and sell (or use) the thing. You need to help this along. Engaging the engineer in your project will do it - so sell your requirements to him - ("if we had xyz feature we could sell in this new market") and be open and honest about the details of your project. It is really very unlikely that a subcontract engineer is going to take your ideas and try and market them himself. It would kill his reputation. But, you need trust on both sides; if you don't feel comfortable with the person you are talking to, just politely move on. It is also about timing. Don't be too worried by an engineer who says "I will have to get back to you about that" - some of the biggest mistakes I have made have been in answering a question too quickly. But otherwise, keeping the momentum in a project can be very good. You can help this by ensuring the engineer has one point of contact in your company - a project manager - usually yourself. If the engineer asks a question or offers alternatives, try and address it in a tight timeframe, and don't forget about it - keep a list. Samples and interfacing equipment are usually critical - get those to your engineer as soon as possible. If an engineer is making the effort to stay in touch, giving him a few minutes on the phone, or a few lines by return email may be all the feedback needed to help him turn a "pass" level project into a "distinction"

A large amount of effort is required in the initial stages of a design project, both for client and design house, in examining the spec, ratifying the clauses and arriving at estimated production costs.

If the job is large, it is pretty normal to consider two or three subcontractors for the same job to the estimate stage, but it is unusual to have two or more design houses produce competing quotes for the same job without paying them a "feasibility study" fee. (Ballpark estimates are a different matter) Normally the selection process has narrowed the field to one favourite even before the specification is worked through in great detail.

When choosing a design subcontractor, be wary, also, of "trying a small job first". Every project should be considered on its own merits. If the client has difficulty establishing that the design house can reliably complete the larger project, completing a smaller task first does not really alter the situation, - credibility should be established by references and examples of past work. In "trying a small job first" the client might even find the nature (or size) of the project causes the design house to underperform by lowering the projects priority, - when a job is important to a client, it rubs off on the sub-contractors.

If you haven't got a good spec to start with, your choices are pretty limited - you are better off with an employee engineer if you expect to "design it as we go" - a real subcontractor knows that he is unlikely to get paid for all his time if there is not a clear objective at the start. One work around is for the client to engage the subcontractor on a per hour or consultancy basis. Or, better still - we have a page on Spec'ing your design work

At the risk of sounding a little self-serving, the one thing a client can do that helps any small business is pay them quickly - larger businesses probably don't notice, small businesses do.

The "outsource" or "inhouse" argument for engineering is as old as the design process itself. I think my most useful contribution would be to say that the decision really should be based on availability of inhouse staff. If you have the inhouse staff, uncommitted, and if they can can complete the project efficiently then don't outsource the work.

One caveat: Consider the big picture, - when you outsource you quite often get greater control over costs and risk, because they get quantified up front. When you operate inhouse some portions of the project cost will be invisible - if your project goes over budget, it is unlikely you will remember that 'Fred' didn't answer customer calls because he was working hard on finishing the new circuit board.

These comments are provided in good faith, and represent the opinions of Steven Murray, manager, AirBorn Electronics. The comments highlight a useful method for selecting an electronics design house, but others will have their own ways. We hope you decide to include AirBorn Electronics in your selection process.

Use of the third person masculine pronoun "he" in this text will not, I hope, stop anyone employing female engineers. It was just how the page was written. Please read all references to "he" as the more appropriate "she/he"