Designing for the Toy Industry: “It sure would be great to have three LEDs in this thing. But they’re 2¢ each, so we have to take one out.”

Will Sakran’s bench

Will Sakran is an electrical engineer in Brooklyn, NY with a small design shop where they do schematic capture, electronics design, and circuit board layout.

I had a chance to catch up with him at a meetup in New York City last week where he demoed his sound-circuit product Foxonix to an enthusiastic crowd. At a dusty table between a pile of PCBs and an army of soldering irons, we got into it. Here’s how it went:

I’m always curious about how people without day jobs make a living so I ask the nosy question: how do you really make ends meet?

Will tells me they’re in three different business areas. Mostly making money doing contract work for toy companies. They also make and sell some of their own products like Foxonix. The third area is pitching toy ideas.

Sophi: Fun! What’s a typical project you’ll work on in the toy industry?

Will: Yeah, so a toy company might come to me with a play pattern description, “We’re making this play set, and we want to build some electronics into this play set so that when you open the door you get this sound effect, or when you press this button it plays this piece of music.”

We write a program to control all that content, to control the lights, maybe a motor. So for those kinds of jobs, it’s usually one product, one electronics design, and one program.

Ever wonder who invents toys with annoying sounds? You’re welcome.

Sophi: When the toy company gives you that description, are you given a quantity spec, like “We would like to sell, say, a million of these things”?

Will: Not usually. If it’s a work-for-hire, then I’m not always in the nitty-gritty of how many they plan to make. But I do assume it’s going to be a lot.

Sophi: I have to know, what’s “a lot”?

Will: The minimum chip order for a lot of these kinds of projects is around 20,000 chips. And depending on what the product is, it could be making 200,000 of something. In the toy business, the margins are crazy tight. When you get your cost breakdown, there’s always a round or two of cost reductions after the whole thing gets priced out. Sometimes you’re sort of haggling over a penny or a penny-and-a-half.

“It sure would be great to have three LEDs in this thing. But they’re two cents each, so we have to take one out.”

When I start a design, I usually start with the play pattern that the toy company gives me. “There’s this thing we want to make. It’s going to have this much audio in it. Two songs, three LEDs, these activation points.” Based on what they would like their toy to do, I figure out what we need to make that happen.

So I’ll draw up a schematic with a speech chip, showing the buttons and a switch, and three batteries with the LED connection. From that schematic the Toy Company will take it and turn it over to their team in (usually) Asia or whatever vendor they’re working with, and they’ll price it out. But it falls on the engineer to absolutely know what is reasonable cost-wise as far as a toy design goes.

Sophi Kravitz: Interesting, so it’s the purchasing team that you’re sending your schematic to?

Will: Yeah. On the front end when we’re defining the design, that’s where I take the play pattern, sort of the rough, high-level spec, and I turn that into a schematic that describes electronically what that design would be. If I build a prototype it’s from the parts I have on hand or I order parts from Mouser or Digi-Key. Sometimes I go out to royalty-free sites and buy some sound effects and some music.

I’ll put together prototypes for a demo so that their marketing and sales teams have something to show at Toy Fair or whatever trade show it is. I do see the electronics and the programming through to the end.

Sophi: You mentioned the other night that you use Altium to generate your BOM to send to the purchasing team, wherever they might be. Some components will have the same functionality but might require less passives or fit better on the board. Who makes the decision, ultimately, about which component to use?

Will: Ultimately it falls on the vendor, which is the factory that ends up making and assembling the product. What they’re going to do is get circuit boards made, and somebody there will buy all the parts and put this thing together. So while I might specify a 0.1 microfarad SMD capacitor from Mouser with an actual part number, what they probably will do is source an equivalent one that they can get for 0.5 cents. And that’s totally okay as long it’s a functionally equivalent part and it’s less expensive. For some parts though there is no substitute, if I’m using an Atmel micro controller and the program’s going to be written for that micro controller, then you can’t just put a TI part in there instead. In my BOM’s comment section I’ll say, “It’s okay to use an equivalent part,” or “Do not use an equivalent part,” or “This resistor has to be half watt.” Or if there’s a specific thing that can’t be changed, I always think it’s a good idea to call that out. But for things like resistors and capacitors and inductors and these super-inexpensive, really common parts, substitutes are usually okay.

Sophi: Does the purchasing team on this kind of quantity utilize design engineers as part of their team?

Will: I think they do have engineering teams. Or they have some people on staff who are at least really savvy when it comes to building these kinds of products. They can probably look at the schematic and know by looking at it what kind of part needs to be used. An example would be if they look at the schematic and they see a resistor in series with an LED and it’s a current-limiting resistor, they can look at that and say, “I can use this kind of resistor” without having to go back to the engineer that developed the schematic.

A lot of these factories are like all-in-one. They’re huge. They make circuit boards. They do pick-and-place. They bond chips directly to the board. That’s all done in one place, and then it goes over to the next part of the factory where it gets assembled into the housing. And then there’s people screwing everything together. There are a lot of one-stop shops in the toy industry.

Sophi: Who puts the mechanical items like USB plugs or nuts and bolts onto the bill of materials?

Will: There’s a whole package of information that gets sent over to either a team that works in Asia specifically, either in Hong Kong or in China. And that’s usually the way it is with big toy companies like Mattel or Hasbro. A lot of the design and development is done here in the States, but then this whole packet of documentation gets sent over to another office, which is in Hong Kong. That office is the go-between between the U.S. designers and the Chinese manufacturers.

Part of this package is renderings, what the item’s going to look like. It includes control drawings, which have all the mechanical dimensions for all the parts. It can include 3D SolidWorks files if the U.S. designers have taken the design that far. It probably includes some marketing materials, including quotas and how many they’re intending to be made in a production run. It includes the schematic and the play pattern and the bill of materials. So all these things get packaged up and sent over to the vendor, to the factory. Then they go through it with a fine-tooth comb, and they price out everything. So when you get a cost breakdown back from the factory, it does include the cost of the screws that hold the thing together.

So any other mechanical parts like button covers that have to be molded or ears that have to be molded or purchased, linkages, things like that, that stuff gets determined by the factory.

It’s not unusual for them to help carry the design through to the end. You might just specify an action, like a motor turns on, and the eyes move up and down. They’ve probably solved that problem 50 times. So rather than the toy company defining what every little piece of linkage looks like to make that happen, we just leave it to the factory to do it.

Sophi: So the factory might have modular things like blinking eyes or linkages, slow linkages, fast linkages, medium linkages…

Will: I wouldn’t be surprised if they have a whole toolbox of mechanisms that they can draw from. But tooling is always part of any development. Unless you’re like physically making the same thing out of plastic, you always have to make new tools. So it’s conceivable that you would throw a new set of gears or linkages into a tool that has to be made for that product anyway.

Sophi: You mean you might just make it bigger or smaller in the CAD file, but make a new tool.

Will: Right, right. Even for a simple product you probably have two or three sets of tools that you need to make to mold plastic parts. So that’s a good opportunity to put in something extra that you might need for internal parts and kind of build that into the tools. That way you do have some freedom in making new parts if you need a new gear or new linkages or you need to make them longer or shorter or bigger or smaller. You could squeeze them into an already planned for set of tools.

Sophi: What’s an embarrassing design mistake you’ve made?

Will: The very first board design I ever did (back in the earlier mid-’90s), you still pulled a book off the shelf, “Oh, here’s the TI logic data book.” You flipped through it, like, “Oh, yeah, I’ll use this part.” So I designed up this part into the very first circuit board I ever designed. And the part already had gone obsolete. By the time we got the boards made and delivered, that part was nowhere to be found. So I had this board that had an empty spot on it that was never gonna be filled.

Sophi: How many did you order?

Will: It wasn’t too many. The reason I tell this story is because now whenever I plan to use a part, I always make sure it’s not too terribly old and not going obsolete before designing it in.