Archive for the ‘Technical’ Category

Oh That Stuff: Operational Amplifiers

Tuesday, March 17th, 2009

Operational AmplifierThese past two weeks at work have been a re-introduction to ECE 320: Electronic Devices and Controls. I have to do something that is very simple conceptually, but continually reminds me why I am not a hardware guy with how complex it actually is.

The problem:
I’ve got a signal that runs from 0 to 18 Volts. I have an analog to digital converter board that can handle voltages from -10 to 10 volts. I need to essentially subtract 10 volts from my input signal. In C:

signal = signal – 10;

done!

In hardware this involves 4 resistors and operational amplifier and expensive 10 volt reference, two regulators and power supplies to handle the power requirements, and more capacitors than you can shake a stick at. Save me now.

Well, I got some help, cause we employ people who actually know how to do this stuff and he came up with some designs for me. But I had to check them out. He told me what they would do, but I had to confirm, build it, and test it. I was very impressed with myself in how much I remembered. I was able to, in the span of half an hour, figure out the whole thing, including all the intermediate steps. It was just like taking a test in college. I can now add another class to the list of classes where I learned something that I have applied directly in work. There are a lot of those by the way. Maybe OSU CSE isn’t as bad as I thought it was in school.

The circuit didn’t work when I tested it. Today we discovered I managed to screw up and pick the one, in stock, op amp available that is not at all what I need. It was a current feedback amplifier, whatever that is. Apparently there are more than one kind; but can’t blame the ECE department for not telling me that. I’m not supposed to have to worry about this stuff.

Paul Romer Agrees with Me

Wednesday, February 18th, 2009

I’ve been catching up on some back issues, well I suppose they are posts, on Greg Mankiw’s blog. I ran across a link to this, Let’s Start Brand New Banks. It sounds pretty familiar to me. Although, it works from evidence instead of first principles, but I still feel justified. If only they would follow our advice.

Economic Something or Other

Wednesday, November 26th, 2008

I read an article today that made me feel much better about what the Federal Reserve is currently doing to help the economy. In short they are doing two things:

  • Dropping loads of money into the economy to fight deflation.
  • Lending directly to non-bank entities

As some of you may recall this second point is something I suggested in my No Bailout post. The lending programs the article briefly mentions are exactly what the Fed should be doing right now. They are well within the Fed’s role as the lender of last resort. The Fed’s job, when facing deflation (a problem economists don’t really know how to solve directly) is to increase the money supply so drastically as to create inflation (a problem economists solve on a regular basis). Since the usual means of increasing the money supply, lending to banks, isn’t working because the banks aren’t lending, its perfectly reasonable to lend directly to others.

The other news this week is that Citibank accepted a bailout from the treasury. This is particularly unfortunate for me as I am both still against the bailout, as enacted, on principle and I am a Citibank customer at some level. I think the treasury’s approach to the bailout is all wrong. The only point of the bailout should be to get us, as an economy, through the shortage of credit until new credit suppliers can enter the market (which requires high interest rates). The goal should be to allow creative destruction of badly managed credit suppliers and only prop up the one or two best managed (as determined by least likely to go bankrupt), and we do that only because we need some credit in the mean time. It seems that the treasury’s idea is to prop up everyone. It may, hopefully as a customer, be the case that Citi is one of the two that should be propped up, but clearly AIG is not. AIG failed early (and often) clearly indicating it should be let die; even now, there still is no reason to throw good money in after bad.

The other bailout, the auto bailout, I’m totally against that as well. The situation with the auto companies is not at all like the banks. First and foremost there is a shortage in the credit market and a surplus in the auto market. The other big difference is that not all car companies are going to fail. The large foreign companies, Honda and Toyota show no signs of failure. This means that the public interest in having companies producing new cars will be served regardless of how the big three American companies do. I argue that there is not a public interest in American car companies existing compared to car companies in general, as that would imply a nationalist impulse that has no place in a worldwide capitalistic system. Since the public interest is served by spending zero public dollars, there is no reason to spend public dollars.

Many argue there is a public interest in protecting jobs, that is true, but misspoken. There is a public interest in having jobs. Protecting existing jobs is a bad idea; it decreases the rate of increase of the standard of living. If GM were to fail, all of a sudden there would be opportunity to start a car company in America and stand a chance a new entreat. Also, the other companies remaining in the market would be less likely to fail. If we were not in a credit shortage people would line up to replace GM with their own ideas. These new companies would still need parts suppliers, and dealerships and all the rest of the infrastructure. This will take time, years for sure, but its not like there won’t be others to produce cars and consume parts in the meantime.

Given the credit shortage it is hard to start a new business right now. That means markets can not properly respond to the long term pressures that they face. That is the problem we must solve, and are working to solve. We need to solve the credit shortage in order to allow new players to enter markets, then the market will resolve any problems in the auto or other industries.

Update:
This Newsweek story provides examples of past American industries that have undergone creative destruction. Despite coming at the topic from a more pragmatic, less theory and market driven it echos my main idea regarding the automakers.

No Bailout

Wednesday, September 24th, 2008

I am not in support of a bailout of any kind. I also live in an ivory tower when it comes to economics, and a short, unexperienced or well taught one at that. I am confused somewhat by the push for a bailout. Why am I confused? Well, on the one hand, I can see where the people who are going to get free money want it, and I can even see how the republicans, who generally are against free money for people, are ready and willing to give free money to banks. I understand that Bush is once again saying, “my way or the highway.” Why wouldn’t he? it has worked every other time he’s done that. That stuff is not confusing.

What is confusing is the economic reasoning for a bailout. The entire point of capitalism is that it is the best known method for increasing the standard of living of a society. It does this by selecting ideas that are winners and ideas that are losers by rewarding people who have good ideas, and making people who try to long on bad ideas go broke. Of course, one would hope people with bad ideas get the picture before they go broke, but sometimes they don’t. This idea is called Creative Destruction. I just recently finished Alan Greenspan‘s book, The Age of Turbulence, which stated the importance of this in no uncertain terms. It also stated that as you increase socialism and other safety nets to insulate people from risk, creative destruction occurs less often and the ability of the capitalist system to increase the standard of living decreases.

All that is happening here is creative destruction. Some firms, are going to go out of business. They are going to go out of business because they had some bad ideas and they ran to far with them. There is nothing wrong with these firms going under. They actually increase the standard of living in the long run by going away. Of course, not all firms will fail completely, some will merge and what not, and be able to survive. This is because they did not run as far with or have as bad of an idea as the ones who did fail. They will, however, suffer pain, as a result of the bad risks they took. Everyone who took bad risks will suffer pain, and as a result people will not take these same bad risks again. That’s how its all supposed to work. That’s clearly an ivory tower way of thinking.

But what about the credit market? When firms go out of business it is usually because there is a glut of supply, too many firms are in the market, producing too many goods. This was in fact the problem a year or two ago (or more). There were too many people willing to lend money and not enough people to which to lend money. To solve this they expanded the pool of people they to whom they were willing to lend. Clearly that was a bad idea, and the market is telling everyone who did this that it was a really bad idea, and it is telling them this in a big way. It seems obvious now that it was a bad idea, but we only know that because of the market.

The problem now is that there isn’t enough credit. People can’t get loans. There is a lack of supply, because in effect all the firms who took bad risks can’t loan money, so they are no longer in the market. The call for a bailout is grounded in the idea that we need these firms to re-enter the market so that supply will increase, and people will be able to get loans. I agree that we need more firms in the credit market, or at least that we need to increase the supply of credit. But why do we need these firms? Firms are in the end just people. These people have proven themselves unfit to be in a credit market; that is they have bad ideas about how to make money in the credit market. We should not help them to re-enter a market that they so clearly have failed at.

But how do we increase the supply of credit without them? There is no reason why new firms won’t stream in to the credit market now that there is a shortage. These new firms will need promises to their investors that they won’t repeat the mistakes that were made. They will need to not take as many risks. Investors, who are reeling from taking too much risk will flock to these new firms. This will provide them with the capitol they need to provide the loans, for which people are apparently clamoring. These new firms will flourish with their new ideas, and the old firms will die; creative destruction. As I said before, die is a simplification, some old firms may survive by doing the same thing the new firms are doing. However, if they can’t survive that way, we should not help them, just let a new firm take their place. This whole thing will take time, I don’t know much, but I bet that we can have new firms in the credit market in 2-3 quarters.

What do we do for the next 2-3 quarters? Well, it won’t be easy. There is a shortage of credit, and so people who need credit will have to stand in long lines, and accept higher interest rates (pay more for it).

Who will they borrow money from? Anyone who will lend money of course, be it small banks who see a profitable new opportunity to expand their business, or the vestiges of the old firms still limping along. There is also the federal reserve of course, they are the lender of last resort. They always have money to lend (that’s the point). In the past, large banks have been their customers, but there is no reason why they can’t make loans, at appropriate rates, to smaller customers, or anyone really. Lastly, don’t forget overseas investors.

The only real problem is that there needs to be capitol, money to lend. There is an easy way to generate capitol; raise interest rates until people who do have money are willing to lend it. There should be no free money from the government. There is no reason to think that the only people who have money are the government. There is plenty of money in the country, and even more in the world. If we want people to loan it to us so we can buy houses, we need to provide them with an attractive interest rate. When we do that we’ll get the money, and we can loan it out, to anyone who wants the money badly enough to pay for it. As time wears on the risk premium that is currently going to be required will diminish and things will return to normal. Change leads to normalcy. We need to change the big names in banking, if we just sit tight, this will happen for us. If we meddle, we’ll loose in the long run.

What might really happen? From what I’ve read, which is only a little, they are considering a type of reverse auction where the banks will all place bids. What they are actually bidding in is too complex for me right now, but the way it works out is that the banks who are better off will be able to off the government a lower bid. The lowest bid wins, and every bidder may make trades to the government at that rate. Anyone who can afford that rate, who can avoid bankruptcy by trading some amount at that rate will survive, and those who can’t still go out of business. As you can tell, Bush did not come up with this plan, which is being called his. This plan is reasonable, in that those who took the worst risks still get creatively destroyed. It lesses the short term pain, and the cost of some, intangible amount of long term prosperity. The risk to reward trade off is probably there in this plan, especially since who knows how much prosperity, how far in the future we’re talking about.

So am I for the plan? Well, I still think we should do nothing, but what they plan to do is not brain dead. What the democrats are pushing for, more regulation would be bad. It all sounds good, but markets don’t like regulation. I don’t much care how much CEO’s get paid. As the paper pointed out to day, they are losing a lot of money right now, if you count stock. If you put a cap on how much you can pay them you could end up with a shortage of qualified CEOs. One may argue we already have that, in which case we clearly need to pay more to get the actual good ones, not less. What we really need is a populace who is educated enough to participate in the credit market on both the lender and borrower sides. Until we have that, the only sensible regulation, increased transparency, is worthless. This is roughly the same remorseless conclusion that I came to in my last econ rambling. What I want to see then, are some educational riders on this bill. Assuming the risks the taxpayers are taking on comes up good and we win big. Lets require all that money we made go to education.

Maybe someday soon I’ll have time to see what Greg Mankiew thinks of all this, but I’ve been too busy at work to keep abreast of his position.

The Kitchen

Tuesday, September 23rd, 2008

Last week two of my three room mates returned. With their help we’ve managed to unpack more than 95% of the house. It really feels much more like a home now. As I have time to make them I’ll post views of the other rooms in the house. The kitchen was one of the first rooms to get fully unpacked, here it is.

The program I’m using to create these views is Hugin. It is a gui front end for Panorama Tools, and it makes it so much easier to stitch together photos and it would otherwise be. Although, it does still take some time to setup, and it takes hours of processing to generate the full resolution source images, which then take time even to shrink down.

The setup is really identifying points in the images that are the same spot in real space. It has an excellent tool to do this, but realistically you need 5-15 points for every combination of two images that overlap. There are 17 images in the above composite. They don’t all overlap, but its still quite a few points to do manually. For simple constructions it will do this automatically fairly well, but on complex setups I find I get much better results if I do it all manually. One wrong point can cause an unsightly distortion. For example, on the one for my room, I started from the base points and added my own to shore things up a bit. There are distortions in both, that I could remove with more time, but this one does look better. I’m somewhat disappointed that I don’t have any displays that are even remotely close to being able to show the whole image at more than a fraction of the real resolution.

Low Probility Events

Tuesday, June 24th, 2008

Last Sunday while playing a game of Settlers of Catan with Knights and Cities my friend Post rolled a corner. I went on to win the game of settlers, but we gave Post an honorary win for the dice roll. Which really, will be more memorable than the game. Of course, if it ever happens again I’m going to have to buy some new dice.

I mean, statistically speaking, it should be the most exciting thing that happened all week. This was of course a week in which the Boston Celtics won their 17th NBA Championship, Tiger Woods won the US Open in some sort of playoff sudden death thing, and I went to the beach after having dim sum for the first time. All somewhat unlikely things, but without crunching the numbers I’m going to go with the dice roll as the rarest event of the four.

It’s a good thing all those sporting events happened last week or that list would have been really short. Things with me have been pretty normal, following the familiar patterns of which I have grown tired. My new mantra, of sorts, is that I am moving in September.

I have, however, failed to make friends with enough people to avoid moving in with some more craig’s list randoms. Last week I started looking and I found a seemingly low probability find (aka it sounded awesome). They responded quickly, to say that they were busy and would let me know when they’re free. I’m thinking that was a line since I haven’t heard anything. But I’m not jaded on the housing search yet, hopefully I find a place before that happens.

Charging for the First Checked Bag

Wednesday, May 21st, 2008

American Airlines will start charging $15 for the first checked bag…

Come on! That is a lot ridiculous. You might as well just increase fares by $15 across the board. The only real difference is the marketing angle. People are used to fare increases, they don’t like them, but they expect them. Increases in extraneous fees that you must pay just make people feel like you are ripping them off. The marketing can’t possibly favor the fee, now that it is no longer, hidden having been reported in the press.

I, personally, like to travel light and make a quick getaway from the airport. Until the asinine no liquids TSA rule was made I never checked a bag (when possible). The current, exceedingly annoying liquids rule results in the same behavior. The alternative, of course, is to purchase shampoo and such at your destination. This takes time, but probably only costs on the order of $5. By charging for a checked first bag American makes this option more desirable, probably more desirable than checking the bag. Although, it may be easier to for business travelers to get reimbursed for the fee than for the cheaper incidentals.

Of course, encouraging people not to check a bag does nothing to decrease the weight of the plane (saving fuel), it just makes them carry it on. With the planes packed as full as they are these days, there is already not enough space for everyone to carry on their two bags. This will only make things worse in that area, and further increase the importance of getting seated early in the boarding order. When the space on the plane is exhausted there will be an interesting choice — do you charge for baggage that must be checked at the gate due to over crowding? That is a damned if you do, loophole if you don’t situation.

The only saving grace of all this is that, for now,

Rival Delta has no current plans to match American’s fee for the first checked bag…

But we all know how this oligopoly works, everyone follows suit rather quickly. Of course, given the bone headedness of this idea compared to the easier to implement fare hike, maybe they wont all be lemmings. Come on, the cost to implement charing for a first checked bag has got to cost more than $15 per bag.

Updates that Suck (and Don’t)

Friday, May 2nd, 2008

I recently bugged Chris about updating this wordpress install, which has been complaining about being out of date for months now. He kindly updated it for me today. I have to say that it has some nifty new features, and is a lot more dynamic working (at least when you’re in the admin area). However, it has a problem. They changed a lot about how the page I spend most of my time on (the write a post page works). They made it suck for instance to preview the post while editing it. They also made it suck to include images.

In news of updates that don’t suck, Mario Kart Wii Rocks! I’m off to play it now.

New Hard Drives

Friday, May 2nd, 2008

Last month Best Buy Online ran a great deal on some 1Terabyte Seagate hard drives. I have been looking for new drives ever since I hit the 90% full level on my current 1Terabyte mirrored setup. 90% full is the magic number where various unreliable internet sources say that file fragmentation can become a performance issue. Not that I need speed out of my huge media storage, but its nice to have. The main problem with 90% full is that you run out of space of course, so I pounced on this deal, but not before they were sold out. So I back-ordered the drives.

The website said to expect 2 weeks, but we all know that’s a bullshit number. I sent them off a nice email at 2 weeks to see what was up, and the response made things seem grim. Like they would much prefer me to cancel my order than them send me the drives. But low and behold yesterday they arrived! It was certainly unexpected. Here’s what I got.

I could just add them to my current setup to get a whopping 2Terbyte partition, but I don’t actually need that much more space. The thought of all those drives (which I have room and connections for) heating up the inside of the case and wasting a bunch of power makes me uneasy. Instead of that, my current setup uses the two new 1Tb drives in place of two old 500Gigabyte drives. This maintains the heat and power usage of the previous setup and still increases my space by 50%. I of course can always re-add those 500Gb drives later if I need to, and I fix up my power and cooling situation.

1.4Terabytes! 60% Free

Quantum Physics is Useful?

Tuesday, April 22nd, 2008

RLE Logo

Today at work we had an amazing talk I almost think I kind of understood. It was by an MIT Professor by the name of Jeffrey Shapiro on research being done by the Research Laboratory of Electronics @ MIT. It offered a ~5 side explanation of quantum mechanics. The rest focused on varying degrees of practical applications. Most of this was on quantum cryptography and how you actually build a world wide, internet compatible quantum encryption key exchange system. I have been wondering for a while how you make that work in reality, and now I have some idea. Crazy!

The final application was for an idealized quantum laser radar. Instead of a idealized traditional laser radar where you send out a light beam and read a return signal indicating a target or not, you instead entangle the transmit beam with a beam that you keep. According to his math, cause hell if I really know whats going on, you get additional detection capabilities when you do this — by measuring the return signal with the kept entangled transmit beam. These benefits come despite that the return signal you receive is so drowned out in noise and losses that it is no longer entangled [1]. At which point the theory of why you should get improved results breaks down. He has no idea why the math still shows that you get better results. Sounds like pretty exciting stuff if you ask me.

1 He said so. 2008. Jeffrey Shapiro.