The Supertester

[tsaksa]

Does your team have difficulty trouble-shooting their robot's electronics?

Does your software team need help simulating input from encoders or other sensors?

Does your team need a small and low cost multi-purpose test device?

Something specifically customized for FIRST electronics?

And something you can customize or modify for your own use?

Team 997 is proud to introduce what our team calls the Supertester. It is being prototyped right now and we are trying to judge demand so we can plan how we will build production models. We do not know yet if we will be selling finished tester or kits, or perhaps even both, so please provide us input.

All summer long our team has been developing a product that we think is very useful when working with FIRST robots. It is microcontroller based, has a 2X16 character backlit LCD display, and comes with several built in features. Here are a few things we have programmed into the Supertester right now.

1. PWM output that can drive a motor controller or servo.

2. Ability to read and display information about a PWM signal coming from another device.

3. Simulates the output from a quadrature encoder.

4. Reads input from a two channel encoder and displays speed and direction

5. Outputs a single channel square wave up to 5 kHz.

6. Inputs and outputs SPIKE control signals

7. Tests limit switches

8. Reads and displays analog signals from potentiometer, gyro, or other 5v analog sensor.

9. Scans an I2C bus and tells you the hex address of any devices on the bus

10. Has a built in option to read data from a ADXL345 accelerometer to test and experiment with it

11. And it is fully hackable. It uses the Atmega328P which can be modified from an Arduino Uno.

That is it for now. Remember this is not yet available in volume, but look for more information soon on our website,
http://chsrobotics.org/
We are also setting up a special email address for getting more information at
supertester997@gmail.com

And a video
http://www.youtube.com/watch?v=vgV74...ature=youtu.be

Here is a low quality photo. We should have something better soon.

[safiq10]

I think you guys might have just created the a part for a future KOP. I'm very intrested in seeing where you guys will go with this! If you ever need testers team 2950 is up for the challenge! Great job keep it up!

[Aaron.Graeve]

This looks both interesting and promising. I could definitely have used this when helping to debug other peoples robots (primarily to check pwm signals to motor controllers). Depending on the finished product, I may get one. Additionally, the capabilities of this may be usefull for CSA's at events.
Could we see more of the user interface, or is that stil in heavy development?
You all look to be doing a very good job so far; keep up the good work and this may become a KOP item, or a staple for teams at the very least.

[tsaksa]

The basic user interface is simple. There is a next key and a select key and two function keys. you navigate the menus by hitting the next key and reading the display. PWM tests, Spike tests, Encoders, etc. When you get to the one you want you press select. If you select an option like encoders where there are a lot of different function a sub menu comes up and you select again. The two extra function keys are not used until you are running a test. There is also a knob that controls features that require analog control.

The students should have a first version of the users manual ready soon and i will ask them to post it. That should answer a lot of questions. But new functions are still getting added. I think you may be surprised how much it can do.

By the way, we are not sure about quantities and price yet. We think we may need to charge about $150 each for these. It would be interesting to know if we are likely to sell many at that price.

Also, at some point we also plan to distribute a copy of the code, schematics, etc. that should allow teams to write their own functions.

Almost forgot to mention. We will be showing this at the Gladstone Robotics First Fare. Chiefdelphi.com/forums/newthread.php?do=newthread&f=53

[otherguy]

We think we may need to charge about $150 each for these. It would be interesting to know if we are likely to sell many at that price.

I understand custom electronics aren't cheap to build in small quantities, especially a polished product like you have here. It looks very clean.
A >$100 price point puts this out of the nice to have around the shop category, at least for my team.

I certainly see the value of having this around, and am not questioning your pricing, but I wonder if you would consider selling this product in different configurations/options?

I'm thinking an unassembled kit (PCB/components/case) or PCB/components w/o case could help bring the price down a little to allow more teams access to this great tool. Would this be something you would consider doing?

[tsaksa]

Originally Posted by otherguy

I understand custom electronics aren't cheap to build in small quantities, especially a polished product like you have here. It looks very clean.
A >$100 price point puts this out of the nice to have around the shop category, at least for my team.
Clip

If we could cut some things out to sell it at a lower price, what would you be willing to do yourself? Obviously installing and soldering the components yourself would save a lot of work for us. It also adds some documentation and support issues, but hopefully we can get around those. What about machining the case? Would you do that to save some cost? I will assume you would want us to supply a pre-programmed Atmega chip. Perhaps we can cut other things out as well. For example, the tester is quite usable without a knob on the potentiometer, even though it looks somewhat less finished. Also, do you need a battery door on the case, or even provide your own case? We are currently using a really nice case with a slide open battery door. It makes the tester look very professional, but adds over $12 to the price and more if it gets any custom machining.

I will pass you request on to our team, but be aware that our coach right now only wants one price and configuration. We have had a working prototype for some time and are really struggling with the logistics of manufacturing and distribution.

[Jim Wilks]

I think $150 for a fully assembled unit is still reasonable.

Only a small percentage of teams get deep enough into controls to be able to use something like this intelligently, so your potential "market" may be small.

Please do continue with your efforts!

[cfair]

At $150, we'd probably buy one. I would love to have a standalone device to test jaguars with the CANbus, similar to what it sounds like it already does with a PWM signal. But modifying it to provide CANbus functionality might be a good off-season project for students. Whether we in fact put up the money depends on what we have left at the end of a season. I think you'd find a lot of folks who have a "use it or lose it" situation at the end of the season, and would welcome the chance to bring in a hackable device like this.

[otherguy]

Originally Posted by tsaksa

If we could cut some things out to sell it at a lower price, what would you be willing to do yourself? Obviously installing and soldering the components yourself would save a lot of work for us.
...
What about machining the case?

Soldering components wouldn't be a problem here, but if the board is all surface mount components, maybe that doesn't make sense to offer as an option, as I would suspect most teams/people don't want to be bothered with hand soldering or setting up to reflow a single board. If you're using mostly through hole components, then offering a board w/ bags of parts makes sense to me.
Offering it without a case or battery pack wouldn't be a problem either. I don't know much about the interior of your box, but I could see mounting everything in a project box from radioshack, or just on a board, with a piece of lexan on standoffs above it. Battery power (if pack isn't provided) can be wired to a battery holder from RS. If the buttons are mounted directly to the board normally, they could be mounted to the lexan sheet, and wired up w/ jumpers. Alternatively access holes could be drilled in the lexan to allow access to a board mounted component.

Long story short, I think a "some-assembly-required" option would be a workable option, even if it meant some re-design work for the purchaser.

I will pass you request on to our team, but be aware that our coach right now only wants one price and configuration. We have had a working prototype for some time and are really struggling with the logistics of manufacturing and distribution.

Understood, carrying and supporting multiple configurations is certainly more work on your end. Just throwing my $0.02 into the mix since you asked. Whatever would bring the cost down would make it more attractive to us. I can't really speak for other teams, but we have the tools and knowledge in-house to assemble something like this from piece parts. We don't need a polished product, as long as it's still functional and operates reliably.

[tsaksa]

Originally Posted by cfair

At $150, we'd probably buy one. I would love to have a standalone device to test jaguars with the CANbus, similar to what it sounds like it already does with a PWM signal. But modifying it to provide CANbus functionality might be a good off-season project for students. Whether we in fact put up the money depends on what we have left at the end of a season. I think you'd find a lot of folks who have a "use it or lose it" situation at the end of the season, and would welcome the chance to bring in a hackable device like this.

This tester does support some I2C, but not anything on CAN Bus. I am not sure if the Atmega can be made to directly interface to CAN without some extra circuitry. That might be a nice thing to consider for a future version if this one did well, but we did not want to take on too much in the design.

[yash101]

That's neat! Quite a bit of practical use for something like that. That makes me think of something that might also be practical:

Write a Java interpretter for the Parallax Propeller and make that load the code in chunks from a flash chip. That would increase the capabilities of that chip, especially after the Java code is converted to bytecode.

[tsaksa]

At this point we are not really looking at any other processor options. We chose the Atmega chip because so many hobbyists are familiar with the Arduino which uses the same chip. Many students would have easy access to the tools needed to modify or extend the capabilities of this tester themselves. There is a lot of Arduino expertise out there that we leveraged to build this tester, and other teams could leverage to modify it.

If this version is successful we might do another product in the future that has more capabilities. If so we can consider other target processors at that time. But for now, what you see is what you get.

[yash101]

Well, I was just making that up as a personal "joke" project so that I can call C by the name "Java," the language required by our team!