Damon's Laboratory Notes

My first poll

Tue, 6 Jul 2010 03:34:00 GMT

What type of content would you like to read on damon4.com?Market Research

Choosing the best pcb size

Fri, 4 Jun 2010 14:26:00 GMT

A while back I got a consulting job through my employer, strictly creating the artwork for a pcb. The client was in the tele-communication field. The circuit was used in the party line telephone systems they use in rural areas. It consisted of two 14 pin chips and a handful of discrete's. I think it counted rings and rang the phone to distinguish a particular phone on a party line. Anyway, the client wanted me to take his circuit and create a pcb. It had to fit in a pcb about 1"x1.5". If I could not fit it in that size he gave me two other sizes that were a little larger, then a little larger still. The client had a pre-existing little housing for the pcb from a previous product and did not want to make a new mold for a new case.

Now here is my dilemma. Which size do I start out with? Placing some outlines of the components on a 1" x 1.5" rectangle proved that by laying the components so they touch each other they would not fit in the space. I talked to some other engineers who had more experience routing pcb's and they gave me this advice: assume success and start with the ideal size, particularly if the client is willing to pay for your time, and this one was.

Sure enough, after the second day I got a bright idea on how to make the components fit on the pcb. You see one of the components I had to place on the board was a large MOV and I had to lay it down because of a height consideration with a lead bend. I wound up laying the MOV flat on top of a chip and another component and that freed up enough space, with some other clever placement and routing, to complete the pcb in the 1" x 1.5" area. The other thing that helped, and I could do this ONLY because it was a small job, was to use Autocad. With Autocad I was able to be more creative with the artwork then a dedicated pcb software would allow.

The outcome of this story? I thought the client would be ecstatic and give me accolades, but all I got was my paycheck. I have since learnt that a paycheck IS something to estatic about and to always try to give the client his ideal product first.

MIL STD 704

Wed, 2 Jun 2010 01:28:00 GMT

As I read Military specs I am reminded of an Alan Greenspan quote "I guess I should warn you, if I turn out to be particularly clear, you've probably misunderstood what I've said”. Most specs seem to remind me of that. But not 704.

704 details the power system specification of an aircraft and actually seems to be written for an engineer to read. It covers the AC and DC power busses on the aircraft and their allowable limits for both normal and abnormal operation.

Bruce Lee on Water

Sun, 2 May 2010 23:06:00 GMT

One of my favorite quotes is from Bruce Lee.

“Empty your mind, be formless, shapeless - like water. Now you put water into a cup, it becomes the cup, you put water into a bottle, it becomes the bottle, you put it in a teapot, it becomes the teapot. Now water can flow or it can crash. Be water, my friend.”

Forces that shape a Design

Fri, 30 Apr 2010 16:32:00 GMT

Sometimes the forces that shape are designs are odd. An interesting example comes to mind. One of the products I once designed and manufactured was a digital motion control trainer for colleges and voc tech. The trainer had experiments in motion programming and closed loop control, etc. What we had decided to do was to take the all digital motion controller we developed for industrial applications and embed it in other products/markets. So we designed a prototype, made an outline of some lab experiments, wrote the first three labs and showed them to some schools. We got an order for 20 of them right off the bat!

I think what sold our product was the look and feel of the product. We were proud of our industrial motion controller and wanted that to carry over to the educational product. We wanted it not only to be the best inside, but to look good and to feel good. We paid attention to all the details, right down to the switches and potentiometers. We delegated someone to get samples of pots and switches from various manufacturers and filter them to the best 4. We then voted on the best feeling pot. The pot we settled on was a custom item from Bourne. They had a 'silky' feel and we loved them.

Now here comes the problem. We chose a semi custom pot. You see our product needed 6 of each of these pots and each pot had a different value. We only needed 20 of each value. There was a $1400 NRE from Bourne since each configuration of this pot was custom and a minimum quantity of 100 pieces. Thus we were looking at easily a $9,000 cost just for the pots on our box. The solution we came up with, in retrospect, sounds easier than it was to actually implement at the time. We went to great lengths to redesign the circuitry so we could use a single value pot and save $7,500. On some of the pots we had to design additional op-amp circuits just to make the values match. It was worth it though considering it was $7,500 we pocketed!

The story of 9 finger Bernhard

Fri, 20 Nov 2009 00:36:00 GMT

I have not posted in a while, I have been busy after my work in China with a ton of new work here, but I do have a story to tell. A while back my business partner, Ed Kafrissen, and myself got a contract to deliver a product to Telerobotics Corp. The president, Carl Flatau, contracted us to deliver the electronics for a force sensing control system for his robotic gripper product. This device we gave him was pretty slick, his product was pretty slick, but that is a different story.

Carl was a stickler for quality. All the devices we delivered to him were 100% tested before we delivered them to him of course, but he wanted to know details about our manufacturing. So Ed suggested I visit the sub-contractor we hired to stuff and wave solder our pcb's since at least a few pcb's came to us with some sort of problem from the assembly house.

That is where I met '9 finger Bernie'. When I shook his hand it was immediately apparent, and let me tell you an awkward moment. Bernhard (or Bernie) had many years of experience and a heavy German accent. I asked him why we were getting some defective boards. A few boards had chips in back-wards, or the wrong resistor value stuffed, ... just a hodge podge of errors. Our pcb's were mixed technology, SMT and thru-hole. Bernie grabbed one of my boards and used the stub of his missing finger as a pedestal to hold the board as he talked to me. Bernie then gave me a bit of advice that has taken me some time to appreciate.

He told me the pcb's had too many different components. He said that for his staff to give me the fewest defects, then my pcb would need to be simpler. He told me the ideal pcb had all the same value resistor. He chose resistors rather than diodes or some other component because the orientation of resistors does not matter. Second he said that he would prefer all the resistor's to be lined up next to each other. I did not laugh at Bernie (I also helped him set up a better quality control on his line for our product, but I pondered on his advice just the same.)

I wish I could have gave him his ideal pcb. Unfortunately I do not have the knowledge to design a board that performs all the functions necessary for a product with a single value resistor. As one step in the design process I do try to simplify my pcb's though now because he was right. A simpler pcb is cheaper to produce and will be more reliable.

It takes good engineering to simplify a board because it's a more difficult challange.

Talk to Damon

Fri, 7 Aug 2009 01:06:00 GMT

Questions about my web site? Give me a free call from any US location! Click below and enter YOUR number. Your phone will ring to mine.

Computer Architecture class and Mano Assembler

Thu, 2 Jul 2009 22:52:00 GMT

This spring I taught a computer architecture class to the students at the Nanjing University of Posts and Telecommunications. This was a great bunch of very smart students. One of the topics in the course was the Mano Machine (we were using the Mano book!). This is a fictitious computer that has a simple architecture. The architecture is simple enough that the students can grasp a complete chip architecture in a few weeks time. There is quite a bit on the internet about this including an assembler and simulator someone wrote. I teach them how to design/write a two pass assembler. Guess what? One of my students actually wrote the code to implement a two pass assembler for the Mano machine. Shao wrote it in Java. He created a Google codeplex project here if you want to download and run or build upon it. Shao Chong got an A in my class. Not only because of this extra credit, but he really understood the material. Good work Shao.

Vivian Ang's concert in Shanghai at Fudan University

Sun, 24 May 2009 10:20:00 GMT

My cousin Vivian Ang is a Concert Pianist. She is on tour in Asia now. I have the good fortune to be in China when she gave her concert at Fudan University in Shanghai. Shanghai is only a few hours from Nanjing University of Post and Tele communications. She is wonderful and Fudan is one of the most prestigious universities in China. Click on the picture to be taken to my photo album of my trip to Shanghai with pictures of cousin Vivian (in case you can't guess she is my cousin from my wife's side).

Persuading People

Sun, 3 May 2009 01:54:00 GMT

As an engineer you will commonly work with clients, managers, vendors and co-workers. It is your responsibility to provide your client or employer with the best advice you can on the engineering topics you are entrusted with. I have found that at times I have felt that I had a better direction in a design path, or even company procedures than the way things were going. Since you, as an engineer, are getting paid to provide your best advice, you need to bring this to the attention of those around you.

What are the best ways to achieve this? My advice is to approach this from an eduction point of view. I have always felt that if others had the same information I did, and they were reasonable people (which they invariably are), they will reach the same conclusion(s) as I did. Therefore I will not need to 'persuade' them. This takes more time then trying to force something down someones throat, but the payoff is immense. You will be viewed as someone to work with. Sometimes all you need to do is to provide a bulleted list of pro's and con's, or a decision tree, or a strategic graphic (a picture is worth a thousand words).

What if, even with your best effort at education, things still do not go 'your way'. Assess the importance of this decision or advice. Is it really a make or break direction for the company or you? In most cases I would think the answer is probably not. If not, take a back seat. Let someone else drive. You gave your best advice and best effort. Also, you may not know all the information yourself. Those you're trying to persuade may not want to put the effort in to educate YOU. Perhaps your direction is not the best. This may be particularly true if you are the new kid on the block. Things are the way they are for a reason. More than likely there is a whole history to the way people do business around you that you may not be aware of and will take time for you to learn.

Perhaps there are alternatives to what you're proposing that are almost as good. About the only additional thing I might try, if I felt THAT strongly about this, is another short, very focused, attempt at education. Then drop it. The worst thing is to be viewed as argumentative, then all that credibility that has built up is washed away.

There is another question you need to ask yourself. Is your goal to be viewed as right or is it to effect a new and better direction for the design or your company? Sometimes its best to let others take credit for your ideas and/or do things more quietly. On the other hand, perhaps it is an appropriate time to be recognized as being right.

I took a course once that dealt with some of these issues. The topic? "How to manage difficult people." Don't be one of them. Be effective.