Options for debugging PSoCs

Ouch :-( ...Jim Thompson

--
| James E.Thompson, CTO                            |    mens     |
| Analog Innovations, Inc.                         |     et      |
| Analog/Mixed-Signal ASIC's and Discrete Systems  |    manus    |
| Phoenix, Arizona  85048    Skype: Contacts Only  |             |
| Voice:(480)460-2350  Fax: Available upon request |  Brass Rat  |
| E-mail Icon at http://www.analog-innovations.com |    1962     |
             
                     In wine there is wisdom,
                    In beer there is freedom,
                   In water there is bacteria
                                               - Benjamin Franklin

Yeah, After thinking about it last night, I am just about ready to go back to the PIC24. My problem was that I have an RGB LED that I am measuring the reflection with a phototransistor to do color classification. Unfortunately, the variation from one LED and photo transistor to another means that I couldn't just build one standard amp to work with them. I needed something I could calibrate to each unit. The PSoCs programmability seemed ideal, and let me incorporate the filters internally, saving parts count. Unfortunately, the lack of an inexpensive debugger is a showstopper. I think I will look into using a PGA in the phototransistor circuit and go back to the PICs!

Charlie

Not sure which PSoC and SW you are using but PSoC Creator has a debugger in there:

--
Regards, Joerg

http://www.analogconsultants.com/

"gmail" domain blocked because of excessive spam.
Use another domain or send PM.

Yeah, that is what bit me. Creator is only for the new PSoC3 and PSoC5 line, that apparently do include internal debugging. However, they are just coming out, have a HUGE price premium, and are way overkill for what I need to do. The existing PSoC Designer software does have a debugger built in, but what they don't emphasize is that it is only for use with their emulator/debugger, not for their every day chips. Nothing like going to use some feature, and find out it really isn't there!

Charlie

Have you talked with Cypress about that? I'd probably call them in this case instead of emailing.

--
Regards, Joerg

http://www.analogconsultants.com/

"gmail" domain blocked because of excessive spam.
Use another domain or send PM.

Can you setup a fixed filter and run a cal routine on the LED? Grab the resulting data, calculate a correction, and reprogram the PSOC? No debug mode required. This can be done with FPGA's using tables loaded at synthesis time. FPGAs arent really debugable, only simulators are available for them.

Cheers

That is around the range i would consider killing the idiot boss. =20

Just as an alternative approach, buy it yourself and ballyhoo how=20 much time it saves you. You just might humiliate him into getting=20 up to the 1980s, when the transition between tool cost ($XXXXX)=20 versus engineer productivity improvements had begun to turn the=20 corner (for even mid scale companies, one seat shared over several=20 engineers).

t

=A0 =A0...Jim Thompson

It is sort of a shame that the PIC doesn't have DACs but I think this idea will work without them.

The light output of an LED is almost exactly linear with current. You can trim the gain of your system a little by tuning the current into the LED. In the PIC case you can adjust it with a PWM duty cycle setting. It seems to me that this should give you enough control to get within the span of the ADC.

If you use a micro with multiple ADC inputs and a MUX, you can have a fixed gain of something like 64 that you can switch in and out just by selecting that input with the MUX.

You can build your amplifier with a bit of bandwidth limiting to keep the very high frequency noise from being too big. Once you have made the values into numbers, the bulk of the filtering can be done in software. A little wobbling of the brightness of the LEDs would help you here. It will ensure that you are running up and down over several codes on the ADC so that the dif nonlinearity drops out of your results.

Ah, but it doesn't help that I am married to her... ;-)

Charlie

Do like me: Get a serial interface + command line running on the PSoc

*and* write / test most of the software on a PC first. Compile for the target after debugging.

--
Failure does not prove something is impossible, failure simply
indicates you are not using the right tools...
                     "If it doesn't fit, use a bigger hammer!"
--------------------------------------------------------------

That does rather complicate things. Then again it took the two of you=20 to so thoroughly mix the domains. Maybe re-visiting the issue of having=20 mixed the domains. Buying it at the price involved should still be more=20 engineering decision than management decision. Try explaining it in = terms=20 of lost or wasted hours.

I start out a project with a wish list of tools. If management (or SWMBO) doesn't buy into at least a minimal set of tools and a risk statement (whine) regarding those not purchased then the project risk is on their plate. In the case of a micro or FPGA I wouldn't start without a full set of at least the vendor's tools and applicable development board(s).

That problem doesn't arise in a true partnership. Me suspects that Charlie whines, blaming SWMBO, rather than 'fessing up that it's HE who doesn't want to spend the money ;-)

In our household we generally don't ask each other for permission, though SWMBO will occasionally ask, "Can we afford...", when she has a big project in mind. ...Jim Thompson

--
| James E.Thompson, CTO                            |    mens     |
| Analog Innovations, Inc.                         |     et      |
| Analog/Mixed-Signal ASIC's and Discrete Systems  |    manus    |
| Phoenix, Arizona  85048    Skype: Contacts Only  |             |
| Voice:(480)460-2350  Fax: Available upon request |  Brass Rat  |
| E-mail Icon at http://www.analog-innovations.com |    1962     |
             
                    Help save the environment!
              Please dispose of socialism properly!

Yes, but in this case, I already had the development kits on hand, and was caught by the 'fine print' gotcha! Didn't see that, with all the ad copy about debugging, the menu in the IDE for debugging, etc. there was a requirement for a separate debugger. Was too used to chips with on-chip debugging resources!

Charlie

No, it is part of our 'teamwork.' She does the budgets, and takes care of the finances. Simple truth is that we have a budget for development of this product. The goal is to build at least 100 units for preliminary sales this spring, and that means that it will take a certain amount to manufacture them, even if I do a lot of the work. We only have so much in the kitty, especially after taxes and all the other overhead we expect in the next year. Unless we put in some more money, it is going to be very tight... :-(

Charlie

With us it's more of a three stage thing. Small purchases (up to a couple of hundred $$) is just a buy it if it's needed. Intermediate purchases (less than $1000ish) it's more of an informational issue; "I'm planning on buying ___". If there is no argument about timing, then it's no big deal. Larger than $1K things are planned a little more. She's going to have to wait a couple of years for her Mustang ragtop. ;-) I want to get the house paid off first.

That's why you put in contingency toy^h^hools in the budget. ;-)

I appreciate your information here. I'm considering a PSoC for an application but the PSoC3 isn't ready for prime time. If I go down this route I'll make sure the FAE gives me all the tools I need. He's shown himself to be pretty hungry for our business. ;-)

My SWMBO has always lusted over having a Jag ragtop ;-) ...Jim Thompson

--
| James E.Thompson, CTO                            |    mens     |
| Analog Innovations, Inc.                         |     et      |
| Analog/Mixed-Signal ASIC's and Discrete Systems  |    manus    |
| Phoenix, Arizona  85048    Skype: Contacts Only  |             |
| Voice:(480)460-2350  Fax: Available upon request |  Brass Rat  |
| E-mail Icon at http://www.analog-innovations.com |    1962     |
             
                    Help save the environment!
              Please dispose of socialism properly!

You'd need a full time mechanic just to drive one on the weekends.

Now that we're in the South a ragtop makes a lot more sense. I'm sure we'll end up with a 'stang within two years.