Re: blog; Motor Controller Options for 2015
 

To be clear, last year there were three speed controllers made by two companies. This year there are two speed controllers made by the same TWO companies, not one. We decided to work together to better be able to meet the demand and have the absolute best products for the market.

This is a small market and the cost to enter this market is very high. Instead of competing all the time, we decided to work together for the greater good of the FRC community. IFI believes this is a win-win-win-win for all parties involved.


With that said, I don't understand why people think it is ok to have a single source for the robot controller and power distribution but is such an "unfortunate" thing to have two FRC suppliers work together on two completely different speed controllers.

Paul

Re: blog; Motor Controller Options for 2015
 

I don't think that's going to happen. The economics just don't work out.

Let's take a look at what would happen if I (yes, me personally) were to try to get into the speed controller market.

Let's say that I haven't found the "killer app" for speed controllers, I just want to make something that's cheap, reliable, fairly linear, small, etc. Basically, I want to clone the new Vic. Maybe I adapt the package to the requests people have made in this thread, but nothing too out of the box.

First of all, I'd have to invest in commercial CAD software. Eagle (ECAD) runs ~$200 per license, and SolidWorks is probably several thousand dollars, without a whole lot of bells and whistles. Now, I can actually start designing. I'd estimate that it would take me about two weeks working part time to get a solid first revision of a speed controller.

Once I've got a first rev ECAD and MechCAD, I send it around to a few friends to take a look at, make revisions, and start to order prototypes. I can get a run of probably 15-20 speed controller boards at something like $300 for the lot (you don't usually pay per board for prototype PCBs, it's essentially a fixed price for a board of a given size of parts). FETs are going to run about $10 per H-bridge, plus annother $15 for other components, assuming we don't put anything too too fancy on the board. Some basic logic chips, LEDs, caps, etc. Now, let's say that I beg some machine time from a local team I'm friends with rather than pay probably ~$1000 for a run of 15 prototype aluminuim enclosures. Still, I have to pay for material, and maybe some tooling, so let's say that that adds annother $150.

I just paid around $825 assuming I make 15 prototype boards and assemble them myself (and not including the price of the CAD software!). This is just a broad estimate, without doing the research, it's my best guess, but in reality the real number could be anywhere between $500 and $1200. That's around $55 per prototype board, assuming $825 total in prototyping costs. Maybe I give these controllers to some teams to run at offseasons, and make updates to the product as requested for the regular season.

Alright, we've just invested some money in prototype speed controllers, now let's take a look at what kinds of costs well incur from the real production run. In order to make any money on these things, I'm going to get them made offshore, and assembled and tested there as well. We can probably get the board costs down to ~$5 a board for a large run, and let's also say we can bring component costs down, to $15 per board. Assembly is probably going to add $2-3 per board, so with board MFG, components, assembly, and testing, let's say that were paying $25 per board. I'd expect VP to be paying something similar, maybe as low as $15 per board, and certainly no higher than $30 per board. Then I have to pay to get the enclosure made. The Vic casings are aluminuim molded, so that means high startup costs, and low unit costs, so I'm not really sure what that would end up costing per unit. I'll just estimate $10 per unit, including ano. None of this includes the probably ~$1500 trip I have to take to make sure all of this stuff is getting made to my specifications.

Let's ship all this stuff over to the US for, let's say, $5 per unit, assuming we don't break the bank for air shipping. We paid about $40 for each unit. Now, we have to pay some guy to assemble the PCB in the casing, and put it in a nice little cardboard box. Maybe that adds annother $5 per unit, maybe a little less. Were looking at a total cost of $45 per unit. Maybe my estimates too high, but I'm guessing even for VP, they must be paying over $30 per unit that they can actually sell.

So, how do we price them then? Well, even if we assume we have a slightly better product in terms of features than VP, we definitely can't price them any higher $60. No one would buy a new unproved product from a completely new company unless they got some really awesome features or a really awesome price. Let's say I price them at $49.99 per unit in the hopes that I can undercut VP.

Ok, I cross my fingers that the things work well (they do), people like the product (they do), that China doesn't screw up (they don't), and I actually get them in stock in time (I also do). These are really big gambles, but let's just roll with it. Let's also say that I am really confident these guys will sell and order a run of 500. That's probably 3-5% of the FRC market. At $45 in costs per, that's $22,500 invested just in inventory. Finally, imagine these guys do sell like the dickens (again, a big gamble), I've just made $2500. That might, just might, be enough to pay off my development costs.

So, after a season of hard work, big investments, big gambles, and a lot of luck, how much dough did I rake in? Well, if I'm lucky, I didn't lose anything. This is for a probably $20,000 investment from my personal savings (not that I have that much!), and hundreds of hours of work.

I could have made probably $1200 from my actual job, coaching swimming, instead of working on this project, and to earn that money, I wouldn't have had to put up such a crazy amount of money with such a crazy amount of risk. I could have probably make even more money if I got a paid technical internship (designing speed controllers for example).

Basically, my point is that the speed controller market requires a lot of capital, and if you don't have the experience and established reputation like VP and CTRE do, it just becomes really, really risky to get a product that works well at a price that could sell well. While I don't agree with all of VP or CTRE's decisions, I don't think they're putting out a bad product or gouging us. They're probably making a big of money off of speed controllers, but I don't blame them. They've got full time engineers to pay, and a business to run. They're selling good, reliable products, and as a consumer, I feel like I'm getting my money's worth.

But anyway, given the costs and risks involved, it doesn't make that much sense to get into the speed controller market at this point.

Edit: Paul beat me.

Re: blog; Motor Controller Options for 2015
 

Can you comment on what your projected stock levels will be by January, 2nd 2015?

Re: blog; Motor Controller Options for 2015
 

He doesn't have to. Like I said above such information is generally closely guarded.

It's in the best interest of vex to fulfill everyone's order.

Re: blog; Motor Controller Options for 2015
 

This is not something we typically do.

By the way, VEX has NEVER run out of Speed Controllers during the FRC season since I have been the President.

Instead of looking at the perceived negatives of CTRE and VEX working together, I urge the community to look at it from a different perspective.

Consider the things VEX has historically been good at and the things CTRE has been good at and combine them. That is what the teams will get out of this partnership.


**And before anyone goes there again, comparing 2 SKUs for a product that there are limited supplier choices to 300 new SKUs in an environment where there is an established company and a completely unknown use case is just silly.**

Paul

Re: blog; Motor Controller Options for 2015
 

I do believe that meeting supply is possible, and that if anyone can do it, it's VEX, but people have the right to be a little concerned.

Historically, these sorts of things have sometimes been poorly executed, like everything BaneBots (remember CIM-u-lators in 2012?), game piece availibility (poof balls, orbit balls), VEX gearboxes (2 years), VEX/AM hex bearings, FIRST Choice (twice, and they knew the exact number of teams both times...), and a few others.

IMO, the biggest possible issues wouldn't be stock, but possible issues with the products themselves. They've gone through beta testing and I'm sure they've been designed really well (a group project between Talon and Victor engineers), but issues that pop up during production can be hard to spot.

That said, I really don't think there will be an issue. All companies/people involved know what they're doing, and I have faith in them. Even if something does go wrong, complaining about it on the internet won't solve anything.

Re: blog; Motor Controller Options for 2015
 

I created a new thread here to share a recent experience with TIM (Thermal Interface Material).

Re: blog; Motor Controller Options for 2015
 

Thank you for the reply, Paul. The more transparent you are with this situation, the better your customers will feel about this. Please, excuse my bluntness, but I do I have a concern. Essentially, the two producers of FIRST approved motor controllers have joined forces to create an great new product. I am excited about the form factor. I am worried about supply. This is a rational concern considering that it seems every year something is in short supply in the FIRST community. Will there be a limit on the amount a team can order at one time. If folk order 40 at a time (yeah 40! read the thread) can you guarantee that all customers will have access to this new product?

I have a small supply of new motor controllers on hand from last year and a stock of used ones. If replacements are hard to come by, I am worried I will not have access to this new source of material when i need them for the build season.

Re: blog; Motor Controller Options for 2015
 

You have every right to be concerned.

Let me try to put you at ease. IFI has 14 years of FRC speed controller history. Unlike gearboxes, wheels, and motors; This history has proven to be very predictable. Our data, combined with CTRE's data from the last two years has kept that predictable trend going.

We are manufacturing enough speed controllers for even the most optimistic of forecasts. Now, getting the mix between the two may take a season to get right but we have accounted for that in our forecast model.

Again, unlike other products for this market, the historical data has proven very reliable. VEX has never run out of speed controllers during the FRC Season in the past and we aren't going to start now.

Paul

Re: blog; Motor Controller Options for 2015
 

That sounds like a challenge. ;-)

EDIT: Seriously Paul, thank you for commenting and talking to the community. I know I'm not alone in having an appreciation for manufacturers that take the time to reach out like you are.

Re: blog; Motor Controller Options for 2015
 

I feel like I have to ask, how much benefit are you going to see in a 3-4 month period by ordering 40 of these? Would it be any additional burden on your team to order maybe 15 or so for the season, and then order an additional 25 next summer?

While I think we are probably not going to see dire shortages, I don't really get the rationale of purchasing so many at a time. If this is something that a lot of financially able teams do, it may actually have a small effect on supply.

Re: blog; Motor Controller Options for 2015
 

I'm not Adam, but I think I know where he's going based on past posting history. I would guess that only about 30 will see use during build, the rest are spares. Those 30 are for characteristics testing, a practice robot or two or three, and the competition robot, maybe a couple to loan to other teams. Due to the speed controllers being an unknown quantity* as far as shaving resistance, heat tolerance, I would guess extra spares as well. I'd also hazard a guess that many of those will end up on next year's robot as well, meaning fewer purchased next year.

For perspective, let's assume that a "typical" robot has about 10 motors of various types--minimum 4, maximum 22 not that anybody ever uses that many. 10 seems to be fairly consistent for a medium-complexity robot. All 10 motors need speed controllers--also typical, though occasionally someone'll use a Spike on one or two. Of course, you want some spares--raiding last year's 10-controller robot isn't always a practical option--so let's say 15 for competition. (See, they're already at your proposed quantity, and I haven't even finished.) Now, let's factor in the practice robot that many teams build--that's another 10 controllers; share the spare pool. 25 controllers and counting. And then you throw in that they like to build a full offseason robot after the season, just kind of testing new tricks, and still show up with primary and practice to offseasons. 10 more controllers, so 35. I'm sure I can come up with a good reason for another 5. And I'm not on the team.

See, that wasn't so hard--40 speed controllers, including spares, covers one year for 'em. More, if they reuse the ones off the practice and offseason robots. And it's a lot easier to just cut one check for a bunch than to cut two for the same item several months apart, particularly if it's a "non-consumable" type of item.


*As in, so far so good, no mass complaining yet--but the suppliers' reputation is very good. OTOH, robotics teams are very creative in finding ways to fry speed controllers for some reason.

Re: blog; Motor Controller Options for 2015
 

If I didn't care how much I was spending but I wasn't completely oblivious to cost (only mostly oblivious), I'd probably order something like 48, because then I'd have 12 of each controller for a regular bot and practice robot, and I wouldn't have to decide which controller to use right now. I can understand the number 40 from a team that is well established and has the means to raise a reasonable amount of funding.

At this point I'm thinking 8-12 Talons for our team. Ideally I'd want 20-24 talons, but I think we will end up mixing and matching and building up our stock of these controllers over a period of about 3 years.

Even if we ended up strapped for cash and unable to buy a single new controller, I wouldn't feel too bad about life. We have enough to get us through next year, because prices on controllers have been good in the last two years, plus we have been able to get some donated controllers through the IFI voucher, AndyMark product donation voucher, and FIRST Choice. And best of all, Talons and Victors are reliable. The motor controller situation has been better than ever in 2013-2014; 2015 looks even better.

Note to IFI / CTRE: if you make a DC motor controller for FTC to compete with the overpriced / unreliable HiTechnic controllers, I promise to order a bunch of them. Seriously.

Re: blog; Motor Controller Options for 2015
 

Sonic, the drive in chain, advertises itself as the Ultimate Drink Stop (or at least they did), similar to VEXpro, the ultimate FRC gearbox stop. Sonic also is in the business of selling tots. People LOVE tater tots even though they're allegedly terrible for your long term health, but that's irrelevant. VEX has been in the business of selling motor controllers for what feels like as long as this year's group of incoming high school freshmen have been alive.

Sonic, while specializing in hundreds of thousands of possible combinations for drinks, may find it difficult to predict sales of some of the niche mixes. However, the tater tots are obvious, consistent, and easy to predict sellers because of a lack of outside competition (no other chain restaurant sells tots) and little to no change in the base product.

VEXpro, while making dozens and dozens of different ways to get FRC motors to work the way teams want them to through some sweet, high-performing gearboxes, may find it hard to predict sales for some of the niche mixes. However, the motor controllers are obvious, consistent, and easy to predict sellers because of a lack of outside competition and little to no change in the base product.

What I'm trying to say is that I'm in a Sonic drive thru right now before a robotics meeting.

Re: blog; Motor Controller Options for 2015
 

I can certainly see why a team would use 40 controllers or more in a year. I still don't see the need to purchase all 40 at one time, when demand is at it's peak and supply is still an unknown.

To large teams' credit, I'm many of them would re-sell or donate spare controllers to in-need teams if there was a shortage, but I don't see the need to contribute to a potential shortage.

This is probably all useless conjecture, though. Like I said, I doubt there will be a shortage like there was with the Talons from CTRE. I'm just keeping the possibility in mind.