Dell Crouch - Personal Website

The Personal Website of Dell Crouch Jr. 38 years of Advanced Engineering and Development in the Lead Acid Battery Industry.

Wednesday, October 26, 2005

On Data Collection

It used to be so tedious. There were no computerized databases. If you wanted to find something, you had to hope that someone had filed it correctly. The searches and queries that we use today didn't exist. As I look through copies of census records from 1930 back, I am reminded that data is data, whether it is a listing of family members or a a simple measurement of temperature over time.

In my career, I've seen the transition from analog data collection and measurement to digital. The result has been a flood of numbers that now causes scientists and engineeers to decide how much data they want. It is now possible to gather too much data! This was unheard of 38 years ago when I got out of college and started working. Here is a simple example. On the electric vehicle program, I was the guy who tested the battery pack for performance and durability. There were 26 batteries in a single series string. The objective was to measure the voltage of all 26 batteries at the same time and record the data. First comes the measurement problem - the load on the batteies changed once every second to sumulate driving in traffic, up and down hills, etc. So, the 26 batteries had to be scanned in less than on second to allow time for data collection and recording. There were also 8 temperature sensors in the battery pack as well as the current and the elapsed time. All of this has to be scanned and saved. So, we now have a spreadsheet with 26 + 8 + 1 + 1 = 36 columns by one row for every second of the test. If the test lasted for 2 hours, that was a 36 column by 7200 row spreadsheet! Early versions of Excel couldn't even handle that many rows (This was in about 1994, before Windows 95!). In addition, hard drive space was used up very quickly because this testing was automated - the batteries ran 24/7, creating a new data file for every discharge and for every charge!

The solution to this problem was to decide how much data we needed. We called this a "smart save rate". We decided that if the data in second number 400 was very similar to the data in second number 401, we would only save one set. This rule applied throughout the testing.

When I started my career, there was no such thing, unless you worked for NASA or the Department of Defense and used a "mini computer" for data collection. (I should note that this "mini computer", it's storage and input/output only filled one room instead of two.) The VAX from Digital Equipment Corporation was an example.) Everyone else used less expensive and MUCH SLOWER equipment. At ESB in Yardley, PA, we had multichannel strip chart recorders from Esterline Angus like this one:

These machines recorded one of 20-40 channels every second. So, if you wanted to monitor channel 3, you only saw it every 20-40 seconds! This was useless for any phenomenonn that did not take hours to occur. To find out where a test actually ended, we would roll out the chart on a desk or table. We knew that the chartspeed was 1 inch per hour. So, we would start at the beginning and measure the length to the last data point. Then, we would use a drafting template called a french curve to draw a curved line through the channel 3 points and find out when the test actually ended at 1 volt. For the elctric vehicle program where the load was changing every second, this instrument would have been useless. Bear in mind that, in 1973, this instrument cost about $4000, a lot of money in those days.

Of course we could have used three-channel recorders with continuous recording of all three channels. BUT, for 26 batteries, 8 temperatures, and current, we would have needed 12 of these recorders running 24/7 - not practical or cost effective. Never mnd the fact that I ofter had 3-5 packs on test at any given time, each of which would have required 12 recorders!

When we had to test every battery pack that we made before shipment, we automated the process even further. I wrote a macro to take the data files and make, save and print graphs. No more set the border (click), enter the titles (click), set the scales (click) , etc. Since all of the tests were the same and all of the files were in the same format, this was the only way to go.

So, the next time you feel like cursing your computer for its very exixtence, consider the alternatives.

Tuesday, September 20, 2005

Thank You!!

First, thank you everyone for a great birthday! I'm now ready to go to the next step in my career.

I'll probably use this site as a publishing platform. By nature, I'm more of an essayist than a diarist (yes kids, there is such a word - I know they don't bother to teach vocabulary in the classical sense anymore).

As you might have noticed, I do have opinions on a wide variety of topics.

More to come

Dell/Dad

Monday, September 12, 2005

Happy Birthday Dad!

This page is your Blog. I like to think of a blog as a journal that I share with an audience. What audience is that? Well for yours, it would start as a place that visitors that you sent your URL (dellcrouch.com) to find out more about you. As those visitors find out about you and enjoy your content, they'll come back to see what you add. And that's what great about blogs, there is an expectation of constantly added content from someone they want to read from. As the word gets out that Dell Crouch knows a thing or two about Chemistry (and maybe genealogy, Politics, and Automotive topics depending on where you take this), people will start to circulate your link. And as your viewership increases, so does the popularity of your blog.

This is your website though, and you may just be interested in posting some articles and other work related documents. But I figured being someone who's not interested in retiring, who is still actively involved in the lead acid community, might enjoy having a place to continue to work, explore, research and discover on the web.

Cheers to the next exciting step in your career, and Happy Birthday Dad!

Dan & Shana, Sara & Steve, Matt, and David.