iWoz (35 page)

Maybe two weeks after that, we got an office in this older part of Los Gatos, the town where I lived. It was right up against the Santa Cruz Mountains on the last little corner with a few shopping places. The space was small—maybe around 900 square feet—and it was right above a Swensen's ice cream place. And that's where Joe, Laura, and I moved ?n.

It was great. It felt just like the early Apple days, so exciting. We were building something no one else had thought of yet. Who had ever thought of playing your remote control into a device that would learn its code? That had never happened. I mean, today it's more obvious because we have universal remotes and things, but not back then.

The first thing we did was to start meeting with representatives of components that might apply: infrared sensors; infrared transmitters; microprocessors. We started looking over the data sheets and the books and figuring out what microprocessor we'd use. We started making some choices and got together an idea that was more in our heads than on paper, but it wasn't like a fin

ished design you could actually breadboard and wire together and build something from. This process was exactly like what I did with the Apple II.

There were a couple of areas that were tougher. One problem we had to tackle was how do you receive infrared signals in the remote? We didn't have that much expertise in this area. I didn't, and Joe wasn't so sound on how you build a sensor for infrared. So we actually hired a consulting firm in Sunnyvale to help us basically read an infrared signal. If you placed your own remote control right next to our receiver, the signal coming out of yours was a very strong signal.

In the same way that the closer you get to a lightbulb the brighter it is, so it is with a remote control. The consultants designed an intricate circuit with an awful lot of parts and filters. And I said, "If you're close and it's powerful, then why can't you just detect it with a lot simpler circuitry?" Just go straight to a photo transistor. You know me. I like to do things with the simplest possible circuitry. You don't need all these special amplifiers that need power all the time. Just go straight to the photo transistor, which is a transistor that senses light instead of electronic signals.

And that idea actually worked out.

They had to put in a couple of little parts and capacitors to filter it just right to avoid the signal bouncing around in a weird way. And they came up with a very good circuit that worked reliably. You could play your remote control into our little receiver device and it would capture the signal very accurately. It could determine how many microseconds the infrared signal was on and how many it was off.

Then it could trace the signal and make a time recording of it, the signal from your remote control.

• o •

The time also came for us to come up with a plastics design for the remote. Very early, just after we moved into the second build

ing on Alberto Way, we started going around to a few design companies to see if they would show us some samples, or some ideas we could look at.

One of those companies was Frog Design, which had done the Macintosh. We called them up and they said, "Sure, we do third- party development for people other than Apple."

They said they'd take anyone who walked in the door and talk to them, design for them. We told them what we wanted, and they did it up in a few models. A couple of them were a little too fancy-looking for my taste. I wanted the most normal-looking design, just totally straightforward, where every button is sort of square, a real symmetrical design.

I wanted it to look like an every person product, not something that looks like it's from outer space, you know what I mean. And we liked some of the other products they came up with to fit that description.

But in the end, they dropped us.

It turned out Steve Jobs was over at Frog for some reason and saw a CL 9 prototype. From what I heard, he threw it against a wall and put it in a box and said, "Send it to him." As if Apple owned it. The Frog guy told me that Steve told him they couldn't do any work for us because Apple "owned" Frog. Not true, and everyone knew it. But Frog told us they felt uncomfortable doing it without Apple's permission—Apple was a big customer—so they weren't going to do it.

Well, I wasn't going to argue. I don't truly know what the real story was, but I thought, Good, fine. We'll go somewhere else. And we did go somewhere else.

• o •

Of course, I had to choose a microprocessor for the device. I ended up choosing two processors. So the remote later went down in history as the first remote control with a dual processor!

Anyway, thinking about the two microprocessors and working

with Joe, I decided it might be nice to have one microprocossoi for small tasks like reading the keyboard and keeping time, and another to do the heavy-duty work. The larger processor I used was an updated version of the old MOS 6502 I'd used for the Apple I. The other was a smaller, cheaper processor. I think we paid 50 cents apiece for it in quantity. It was a 4-bit processor meaning, remember, it could process only 4 bits of data at a time. That was all we needed for these smaller tasks.

However, a little processor like that is hard to write a program for. Man, was it hard to control. It was almost as hard as program ming the state machine in the floppy disk. Nothing was built into the hardware, and when you don't have the hardware resources, you have to take advantage of what you do have inside the chip. And you wind up with weird instructions that do things in weird ways. That's because the chip didn't already have built into it well-thought-out instructions a human could easily understand and use. That was to keep the cost to a minimum.

But the program on the 4-bit microprocessor wound up doing two basic things: keeping the time of day, monitoring the keyboard, running the LCD display, and enabling power to some of the rest of the circuit, and it also communicated with the bigger, 8-bit microprocessor, telling it what keys had been pressed and receiving data to display on the screen.

We sat down and sketched out on paper where we wanted the lines of letters, numbers, and a few special words to pop up on our LCD. And we found a company that would make us an LCD. We gave them our layouts, and they eventually brought us back LCDs with a bunch of connection pins. And that LCD would actually connect to the same 4-bit microprocessor chip that was reading the keyboard.

Now, the real guts of our product—memorizing all these different infrared codes and repeating them when you pressed buttons—was going to be done by the second, more powerful

microprocessor. Because of the updated version of the 6502, I thought, Great. I am so familiar with this! It had a very beautiful architecture inside. The way it was structured to be, with very few transistors inside doing a lot of work, it was just so good and it did the right job.

The Apple II had my little development system that I wrote myself, and I could type instructions in quickly and test them out. What if I could have that for this microprocessor? So we actually designed our board in such a way that you would be able to hook up through a serial port so that we could connect a terminal or computer directly to it. That would enable you to type and see data on the screen, although the remote control was really the computer. (It was like a little cousin of the Apple II.)

What terminal? Well, I decided the Apple II C would be a great terminal. There were programs that could make it behave like a terminal that talked to other computers.

Remember how I told you that in the Apple III had added this mini-assembler that let me type in things like LDA for loading the A register, or #35, which meant 00110101, the binary language of Is and 0s that computers can understand? That program and many other development tools were built into the Apple II, but would really be useful in the remote control as well.

I had a friend I'd worked with at Apple, John Arkley. He was a consultant, and he offered to convert the debugging and other code I'd written for this new 6502 microprocessor. We paid him and he did it.

And it was great. I would hook a little Apple II C into our breadboard—our wired prototypes—and type away and do the debugging. It's like I had a new little Apple II inside of the remote control. It had all the fun of an Apple II.

And when we finished designing it, this product was great. We were all just blown away by how great these tools turned out.

So then we had a manufacturing issue. Who was going to man

ufacture the device? And suddenly an old friend of mine showec I up from my childhood days with the Electronics Kids. Remember my neighbor Bill Werner? He was the one who did all the toilet- papering of the houses with me and got all that phone cord to create the house-to-house intercom system in our neighborhood.

By high school, though, Bill had kind of gone in a bad direction. Not like me. He got bad grades, got a motorcycle, got in trouble for burglarizing an electronics store, got into some bad stuff. But now he'd turned his life around, and we ended up hiring him—he had worked at the Silicon Valley manufacturing firm Selectron. And we hired his wife, Penny, too, to do some secretarial work. So our team was building.

Selectron was the kind of company we needed. It did manufacturing, like I said, and that was the one thing we had left—figuring out how to build this device in mass quantities.

• o •

Meanwhile I got a call one day from a venture capitalist in England. You see, years before, in the early days of Apple before we went public, he'd called me up and offered to buy some of my stock at a low price, and I'd said yes. But he hadn't bought it.

Well, he called at a slightly later date and asked again if I would sell him the Apple stock at that price. I'm not sure what the offer was, but it was low. By this time Apple stock was easily worth ten times whatever he was offering, even though it wasn't yet public. He said, "You promised to sell me some at this price. Will you?"

So I kept my word. His venture capital company made a ton of money on the London market, a ton of money.

Now, at CL 9,1 told him all about this new company I was starting, and he said, "Can I visit you?" I told him sure. He showed up. I remember thinking, Man, this guy is really staid. Just very formal—so reserved in his language and manner. He was English, okay. I guess he was stuffy compared to us, and you can imagine how loose we were.

Anyway, I described to him what we were doing, and he immediately said he wanted to invest. I told him I wasn't taking any money, that I was financing it all. But he actually begged me.

Well, when people beg me and say they want to be a part of something, I always give in.

After his investment came in, I suddenly had another big investment from the big Silicon Valley venture firm New Enterprise Associates (NEA), which had also done 3Com, Adaptec, and Silicon Graphics. This guy from England had brought his friends in, you see. So all of a sudden we had two or three million dollars.

So we pulled that off in a few months, and we began to realize we were going to need a bigger place to set up shop. I called an old friend of mine from Commodore, Sam Bernstein, a guy who'd written articles for newspapers. He was sort of a reporter. And I always liked the way he thought and the way he organized his thinking. So I asked him to come on board—this was early on— as president. We got along splendidly.

• o •

We ended up keeping CL 9 in business for about three years, maybe a little more. There are still people out there who talk about how amazing our product was. I don't regret doing it for a second. I ended up selling the company to someone, but they couldn't raise money and closed it down.

But at the time, I had other challenges to think about. I had two small children at home (Jesse and Sara). So it was hard making sure I had enough time to devote to them.

I mean, after the 4-bit microprocessor project was done, it was time to do the 8-bit. And I set out to do it and was just having a lot of difficulty getting started on that job. I had my kids I was giving a lot of attention to. And my relationship with Candi was starting to get rocky. We were fighting. We weren't getting along at all. We had fights about how to raise the children, especially. And we were talking separation.

Well, I had an idea to just take off and hang out in a hotel room somewhere beautiful for a week. I planned to just disappear from the world and go to Hawaii and write the code.

So I went to Hawaii, the Hyatt on Kaanapali Beach, and I sel up my little Apple II C so I could start typing the new program in. (Someone was watching the kids.) I thought solitude would help me finish the project. At least I hoped so.

But what happened was, I didn't do a single thing that entire week. I literally sat there looking out my window and watching whales every day; I got used to the hotel schedule. I swear, about ten times a day somebody would come into the room to restock the minibar, change the sheets and towels, check this, check that. All day there were these major interruptions. I hated that.

So after that week of doing nothing, I thought I should stay another week. I found out I could keep that same room that I loved for another week.

Well, guess what? I wound up staying there for four weeks and not doing one single bit of code. I did nothing there, absolutely nothing. I just enjoyed being there. While I was there the
Challenger
space shuttle disaster happened—it was January 28, 1986—and that was really extremely upsetting to me. But whatever the reason, I did nothing.