Questions about making a propeller based product

jeffjohnvol

My dad had an idea for a add-on product to replace a manual set of controls on a popular boating item.· The logic that it would do would be pretty simple, and could almost be done with a basic stamp which I have used before, but one thing it would need is a timer, and my programming skills are better than my electronics, and would rather have something on-chip like the prop does.

Also, I would like to someday do this kind of development work as a side business when I retire.· Lord knows I wouldn't be able to quit my job now and do this and make a living, but it would be fun to do for some side income (if any!), and its better to use a more versatile chip like the propeller, even if I have to learn a new language.

I also want a chip fast enough to read quadrature encoders.· I had a stamp and it wasn't fast enough to capture the transitions.

Anyway, here's my question.· I don't mind spending money up front for the development boards etc, but when a product is made, are there decent alternatives to get an integrated board that has the prop with the necessary hardware mass produced (say 100 at a time or more) that are more cost effective than are sold here?·· The one-off's I see here are good for a single project, but for the prop chip plus a board it would be in the $50-60 range which would be a large chunk of the margin.

I've seem some interesting products that Bean has made which are intriguing, especially the USB device.

If my quesiton requires more specifics, I apologize in advance, but would be happy to try to give more if needed.

Thanks.
Jeff
·

Mike Green

A lot depends on what else you need in your specific system. If you use a PropPlug for programming and testing, you just need a 4 pin header on your board. The processor itself needs a crystal in your case (to do timing accurately) and an EEPROM. That's it (except for a 3.3V power source). You'll need a pullup resistor for the EEPROM and some 0.1uF filter capacitors. Use the Protoboard for a model.

Designing your own PCB is really the only practical way to mass produce moderate quantities (100's) of a system. There are companies that will assemble the boards for you. Again, the Propeller specific parts really only include the processor, crystal, EEPROM, a couple of passive components, and a header for programming.

Leon

It sounds like the Propeller is overkill for your application. If all you need is a few I/Os and an accurate timer any of the much cheaper MCUs like a small PIC or AVR should suffice.

Leon

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Amateur radio callsign: G1HSM
Suzuki SV1000S motorcycle

jeffjohnvol

Leon said...
It sounds like the Propeller is overkill for your application. If all you need is a few I/Os and an accurate timer any of the much cheaper MCUs like a small PIC or AVR should suffice.

Leon

Leon, that is probaby very true.· I was just contemplating that the prop would be better to use on more complex projects later on and would be fun to learn.· I didn't want to invest too much time/money and then find out I need to use a prop later on for a project that needs more realtime processing.

On a cost savings basis, I would probably be better off seeing if someone on the SX··board (or other pic sites) wanted to design/build this thing for me (for a fee of course).· The logic of this thing would be soooooo simple.· Almost laughable to use an MCU.

Thanks.
Jeff

Leon

I could probably do it for you, but you'd be better off finding someone near you. I'm in the UK.

Leon

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Amateur radio callsign: G1HSM
Suzuki SV1000S motorcycle

Post Edited (Leon) : 6/22/2008 12:46:34 PM GMT

jeffjohnvol

Mike Green said...
A lot depends on what else you need in your specific system. If you use a PropPlug for programming and testing, you just need a 4 pin header on your board. The processor itself needs a crystal in your case (to do timing accurately) and an EEPROM. That's it (except for a 3.3V power source). You'll need a pullup resistor for the EEPROM and some 0.1uF filter capacitors. Use the Protoboard for a model.

Designing your own PCB is really the only practical way to mass produce moderate quantities (100's) of a system. There are companies that will assemble the boards for you. Again, the Propeller specific parts really only include the processor, crystal, EEPROM, a couple of passive components, and a header for programming.

True about what is needed.· I guess I was hoping there may be some $12 generic board that has the stuff you mentioned that people use in the field to run·a programmed prop with a header for the dig inputs, and maybe a built in ADC [noparse]:)[/noparse] .

Newbie question here:· Even though I have programmed a stamp before, I've never done the final step of loading the code into the stamp and have it run on its own by just being powered up.· I assume its a simple process with the prop/stamp?

Thanks.
Jeff

jeffjohnvol

Leon said...
I could probably do it for you, but you'd be better off finding someone near you. I'm in the UK.

Leon

[noparse]:)[/noparse]

Leon

I don't know about the Stamp, but all you have to do with the Propeller is to use F11 to compile the program and load it into into the EEPROM. It is then loaded into RAM and executed when the Propeller is reset.

Leon

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Amateur radio callsign: G1HSM
Suzuki SV1000S motorcycle

scotta

Go with the prop. It's cheep and has plenty of room for expansion.

My question is: What on a boat needs a quadrature encoder. Aren't
they enough trouble as it is [noparse]:)[/noparse]

"Sailing is the fine art of becoming ill while going slowly nowhere at great expense"

grasshopper

jeffjohnvol:

Making your own circuit board is quite simple and cost less that paying for some one elses most of the time. I designed a board and had 4 made for less that $100.00. Sure that is a little costly but for 4 new prototypes its not that bad.

here is a picture

Mike Green

jeffjohnvol,
The Protoboard is the closest to what you're asking. The only thing missing is the headers which you can easily solder in place. You can add the components needed for as many ADC inputs as you need. They require two resistors and two capacitors mounted as close as possible to the processor chip. Realistically, you could probably get 4 to 8 mounted on the Protoboard.

There is a variety of boards available for the Propeller, usually developed by someone for their own projects. Unless your needs are much the same, the particular design isn't quite what you need and you have to make modifications or build your own. ADC inputs particularly are a problem for prototype boards because the additional parts have to be close to the processor chip, so the design has to commit I/O pins to that function or at least plan for it.

Leon

I still can't see why one would use a chip costing $12 chip in a product when a $2 chip would work just as well. It could make as much as a $50 difference to the selling price!

Leon

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Amateur radio callsign: G1HSM
Suzuki SV1000S motorcycle

grasshopper

Leon it is so that he can learn how it works. I am fairly sure that any 2 dollar IC will not out perform that of the propeller. Plus look at all the great resources he has available. I have personally, never recieved as much support from Microchip, Ti or others for that matter. Parallax is becoming one of the best and I am moving my company towards the propeller even if I have to pay a few more dollars. Well we employ the RCM2130 Rabbit module and hands down cost 10 times as much and does 10 times less.

Leon

An ordinary MCU will perform just as well as a Propeller in his application. It will be able to use a smaller PCB and will consume much much less power. There are lots of situations where a Propeller is an excellent solution, but it's not suitable for every application.

Leon

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Amateur radio callsign: G1HSM
Suzuki SV1000S motorcycle

Mike Green

Leon,
Have you looked at the datasheet for the Propeller? If you're not using all the cogs or if you're using them intermittently, the power consumption drops a bunch. For many applications, you don't need to run the processor clock at 80MHz and the power consumption drops rapidly as the clock slows down. With an 80MHz clock, you're talking about an instruction per cog per 50ns. If most of your I/O is serial communications and encoders, you can easily slow the clock down significantly and markedly reduce the power consumption. Jeff Martin's encoder driver has formulas in the comments for clock cycles vs # encoders and max. RPM. At 32MHz, you can have 16 encoders at 1048 RPM. The current draw per cog is about 1/2 at 32MHz vs. 80MHz. With a 32MHz clock, the FullDuplexSerial driver still should be able to handle 115KBaud.

Post Edited (Mike Green) : 6/22/2008 4:28:49 PM GMT

Leon

It's still overkill for the OP's project, if it's something he wants to put into production. I can't see the Prop using less power than a nanoWatt PIC in his application even if it only uses one cog, and it's going to cost far more.

Leon

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Amateur radio callsign: G1HSM
Suzuki SV1000S motorcycle

Graham Stabler

If he could program other controllers is might be a no brainer but otherwise perhaps the propeller is not such a horrendous choice, the next product might be something more complex. It really depends on the application and the market, if this thing is going in a fancy brass box to fit on a yacht then he can probably go with whatever he wants. "Production" does not have to mean made in china retail prices or even large volume.

Power consumption is unlikely to be a problem I'm guessing but again we don't really know the application.

Graham

Leon

The OP actually agreed with me that there wasn't really any point using the Propeller, and was going to find someone to help him get develop it with a PIC.

Leon

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Amateur radio callsign: G1HSM
Suzuki SV1000S motorcycle

jeffjohnvol

Mike Green said...
jeffjohnvol,
The Protoboard is the closest to what you're asking. The only thing missing is the headers which you can easily solder in place. You can add the components needed for as many ADC inputs as you need. They require two resistors and two capacitors mounted as close as possible to the processor chip. Realistically, you could probably get 4 to 8 mounted on the Protoboard.

There is a variety of boards available for the Propeller, usually developed by someone for their own projects. Unless your needs are much the same, the particular design isn't quite what you need and you have to make modifications or build your own. ADC inputs particularly are a problem for prototype boards because the additional parts have to be close to the processor chip, so the design has to commit I/O pins to that function or at least plan for it.

Mike, you have a valid point on the ADC's.· I guess I just have nightmarish flashbacks working on my EE degree and we had to build our own ADCs.· IC's make it much easier. [noparse]:)[/noparse].· Thanks for the tips on the caps.· I wouldn't have thought of that.· Sadly, my EE degree turned into a database programming carreer.· Anyone need SQL help?· LOL.

jeffjohnvol

scotta said...
Go with the prop. It's cheep and has plenty of room for expansion.

My question is: What on a boat needs a quadrature encoder. Aren't
they enough trouble as it is [noparse]:)[/noparse]

"Sailing is the fine art of becoming ill while going slowly nowhere at great expense"

Sorry, the encoder was an experience with an earlier project.· Just thinking about getting experience on a chip that "could" handle one.· If you look at my earlier posts (a year ago?) I talk about a tram lift project.· I ended up using an absolute encoder, that varies from 0 to 5 volts and rotates 360 degrees continuously.

Thanks.

jeffjohnvol

Leon said...
It's still overkill for the OP's project, if it's something he wants to put into production. I can't see the Prop using less power than a nanoWatt PIC in his application even if it only uses one cog, and it's going to cost far more.

Leon

Leon, Actually, you do have a point.· I probably need to find a more complex project to use the prop for, and jumped at this one.· Truth is, I could probably do it with some nand gates and a 555 ckt.

If this idea of his has any legs, it would be more profitable to use the sx chips.·

Thanks.
·

jeffjohnvol

grasshopper said...
jeffjohnvol:

Making your own circuit board is quite simple and cost less that paying for some one elses most of the time. I designed a board and had 4 made for less that $100.00. Sure that is a little costly but for 4 new prototypes its not that bad.

here is a picture

nice.· [noparse]:)[/noparse]