I've been dealing with a difficult problem for weeks at work. I've been thinking about how to implement a sensing circuit (sorry, I can't be too specific). I had developed a solution a couple months ago, but it wasn't very robust. A good solution to the problem just wasn't coming to me. After sitting down a little bit today and looking at the problem - the solution came to me - in 5 minutes. It is simple, removed 2 of the primary variables that made the problem difficult, and focuses on the one variable that the sensor needs to focus on. It's a cheap solution, and works on a single voltage rail. Finally after a little detailed analysis, I showed that it works better than I originally intended - a nice surprise.

But the solution is still bitter sweet. The bitter: They still want to have a software override for the sensor. Indicating to me that they don't trust the circuit or my ability to design (at least at this point in the program). The sweet: They will literally spend millions on the software override at least until they get comfortable with the circuit. In the end - they will be forced to use the circuit, and it will quite literally save the program.

I guess the reason I am so excited about this design is that it was the first time in my professional career that I solved a difficult electronics problem with a simple circuit. I've spent many many hours studying Jim Williams' work and the work of the early early Tektronix and HP oscilloscope engineers' and I've marvel at it's complex simplicity. This is the first time I've really felt like I'm following in their footsteps, and I hope someone looks at my work someday and thinks the same.

Sorry for this little brag thread, but this is a great turning point in my career. And not too many people can fully appreciate how I feel, but some of the people here will. I really do take great satisfaction in making simple solutions using analogue technology that is all so often thrown into the digital or software realm of control these days. I've always considered myself an analogue engineer - but today I feel like I AM an analogue engineer.

 

Congratulations!!

 

It's interesting that the assumption is made that simple solutions just can't possibly work as well as the one which cost's millions of dollars and countless man-hours.

What comes to mind is the 1st Mars rover program where they used airbags, which protected the rovers and allowed a safe landing no matter what the terrain., with no landing control system required, and they used proven off the shelf. items instead of "re-inventing the wheel" for every system.

 

Don'tcha love it when a plan comes together..

 

It's interesting that the assumption is made that simple solutions just can't possibly work as well as the one which cost's millions of dollars and countless man-hours.

Isn't it? I marvel at this every time I encounter it. It's like people make an evaluation of how important a thing is, and then it automatically follows that the thing aught to have a level of cost or effort required that is directly proportional to the level of importance. You try to tell them that the cost or effort required to obtain it is actually much less than they were expecting, and they can't be convinced. They feel as if it is a violation of the laws of nature, and/or that some mystical force will even the score at some point in the future if great effort and cost is not expended in the present.

 

You guys hit the nail on the head.

I explained that this circuit would do the job during the meeting yesterday and they need not worry about all of the software overrides, and software freaked. Saying things like - 'software always get priority' and 'we need to be able to override'. I knew instantly that the topic was off-limits for future discussion.

It will be easier and quicker for me to put in the overrides to the circuit than it will be to argue with them. I think I will send a well thought-out email next week, but I'm sure it will fall on deaf ears.

It is quite a compliment to have a program spend millions of dollars on something that needs no more than $50k in design and test, and when the day comes that the circuit needs to work - it will - and I'll get my 'I told you so'.

 

Congrats. Finding simple solutions to tough problems makes is all wothwhile. Hopefully, someday you can share your insight with the rest of us.

 

A 25 cent problem doesn't neen a 10 dollar solution. And sometimes everyone needs that reminder ... Those problems cross the spectrum of humanitiy's endeavors.

 

Congratulations on your design. I hope it serves you well.

I do think you've missed a real problem: you now need to complete the social engineering aspects to get your idea implemented. You have to get someone up the food chain with the proper pay grade to approve your work. From what you've said that is a cultural problem, not engineering.

I would handle it by having a talk with the highest person who will give you the time. Your goal is to get them to "explain" to you the 'software always get priority' and 'we need to be able to override' issues.

You need to be polite and respectful of everything you say, buy question everything in a tone "but why can't we..." Be passive and let the ideas come from the other person.

Keep in mind that in the end they may well be correct, and this central controller needs to be in control at all times.

 

Congratulations on being one of a dying breed, the analog guy.

Perfect example: http://forum.allaboutcircuits.com/blog.php?b=532

My wife said her employer needed a thermostat for stress testing their finished products.
Because I was a State Certified heating and cooling designer, I just happened to have the LM723 thermostat built and working. I offered it to whatever jerk was sent to talk to me.
He asked about long term repeatability and I quoted the spec sheet for the 723 chip, .1% per thousand hours.
He said he couldn't possibly pay for something that accurate.
I had not yet told him I was willing to give him the prototype, the documentation, and the rights to copy it for $50. (1985 dollars)
The next January, the company calendar arrived with a photo of $3,000,000 worth of melted equipment as the cover photo. That company went bankrupt 3 years later.

The LM723 thermostat is presently controlling the space heater in my bedroom. It has been working properly for 35 years without a repair. Most amazing: it hasn't even fouled the aluminum electrolytic capacitor on the power supply in 35 years!

The moral of the story? You can't sell a $50 solution for a three million dollar problem.

Edit: I didn't expect any, "Thanks" because I've posted this before, but that was a few years ago. There must be several new helpers on this site since then. And a most humble, "You're welcome". (If you can do the solution in analog, it often kicks the butt of the digital solution in simplicity, price, and reliability.)

ps, this edit caused the "thanks" to disappear (tindel and Metalmann)