It was a 22 or 23 inch HP CRT monitor. It also had a built in switch to use it with two computers. I could have four separate areas on the screen for doing graphics, or web design. That was a lot more convenient than looking for something on the task bar, which was usually several screens long, so you had to scroll through each line to find something. Just move the mouse to that area and click. :)
Another one is the separate code/data memory spaces. In practice this means that the compilers cannot store constant data structures in flash without using special macros everywhere. Doesn't matter in small projects, but eventually you want to have e.g. a lcd or remote terminal with menu structures, settings, multiple languages...
"Jim Thompson" wrote in message news: snipped-for-privacy@4ax.com...
Your desk isn't wide enough for two widescreens?
The traditional 4:3 aspect ratio monitors are still available, though... but rather less common and more expensive. At Newegg there, they list 289 widescreens and 57 non-widescreens. All the 19" ones are 1280x1024, though. Hmm... it was only some 2-3 years ago when you could still get a 21" 1600x1200 LCD for something under $300... gone now...
Ah, eBay still has them:
-- 20" 1600x1200 ~$250 "manufacturer refurbished"... (I've purchased monitors from those Legend Micro guys in the past; they're on the up-and-up).
---Joel
ributions
mber -
n enough
the
in
uI gave you the part number, but you looked up something else.
Any and all of those parts, however, more than made the point that power consumption by the uC was not a significant factor for this app., and that we needn't exclude uCs on those grounds.
-- Cheers, James Arthur
Thanks for the feedback--#1, 2, and 6 are big concerns. I can imagine re-writing the boot loader (6) to make it more reliable, but if you can't get parts (1,2) that's game over.
Since I looked at AVRs, Microchip has come out with lower power devices, but the AVR was already low enough. Against PICs were the assembly language and the 4-clocks-per-instruction, all things I could live with, but would rather not.
What do you like? MSP430? ARM? (overkill) Another?
-- Cheers, James Arthur
Also note that there's "special support" for a bootloader section of memory in AVRs: There are fuse bits to re-direct the part to begin executing code there (rather than at 0x0000) as well as to protect it (or not) from readback.
Personally I think #6 is a non-issue... if you have a part that's somehow accidentally, e.g., jumping into the exact right subroutine that's part of the bootloader that could, e.g., erase a chunk of flash memory, I'd say the design isn't reliable in the first place.
I've used PICs, AVRs, and MSP430s at the assembly level, and of those I think the MSP430 is one of the "nicest" all-around microcontrollers. However, PICs tend to win on price -- if I were going to make 100,000+ of something, I'd probably go with them.
Even when ARMs are overkill, these days they're often no more expensive than the higher-end 8-bit CPUs. (I agree with Vladimar that Atmel -- and I also think TI, for that matter -- tend to charge a rather high premium when you want the 32-64kB parts...)
---Joel
Bill Sloman
t I seem
? This
w it co=3D
d)..
he
nly
between
ships,
hbig
simple fitler.
h channel.
... summing
d, else none.
lable use R2R ladders.
10us, a little longer when picking a new value from the sine table could be much shorter but doing clever arithmetic with carry's etc. isn't C's strong side-Lasse
agreed, the only reason to use an avr is that is available in DIP so it iseasy to breadboard for small stupid one-offs. for anything serious I'd use ARM, life is too short to mess around with 8 bit devices and strange memory limitations
-Lasse
Jim Thompson Inscribed thus:
I'd be pulling apart and re-capping it. :-)
Atmel's stability in the business world may be open to question... they've been through some substantial turmoil in recent years. They appear to be shedding portions of the company.
I just got a call yesterday morning from a guy at Inside Contactless, who informed me of something I hadn't known: the whole division of Atmel that I'd been dealing with, some years ago (the Secure Microcontroller Solutions business) was sold to Inside Contactless earlier this year.
I too appreciate the AVR architecture and instruction set... it's one of the most C-friendly instruction architectures I've run into, and one in which it's actually quite pleasant to write optimized assembly "helpers" for C programs. However, I also would be reluctant to pin my hopes (or a future project) on this architecture, for concern that it would Go Foom if Atmel doesn't manage to survive as a real competitor on the market.
It may be the CCFLs failing.
On a sunny day (Wed, 10 Nov 2010 12:41:00 -0800 (PST)) it happened " snipped-for-privacy@fonz.dk" wrote in :
Here is a fun PWM project, a PIC used as a class D audio amplifier:
Must be shocking for Sloman to see 1000 Hz modulated PWM. Of course it goes much higher, the PWM freq is 63 kHz and the audio sampling freq 160 kHz. The supply current depends mainly on the load of the speaker or headphones.
messagenews: snipped-for-privacy@30g2000yql.googlegroups.com...
this:
Except that the latter is not.
happenedBillSloman
:and doesn't i=
out -
be. JA can
out-think you wit=
when you say
was delighted
was.
you can't
sci.electronics.design, you
10Hzcells."
that.
"a
about
at
at
IOW, you're saying that it can't be done, liar. You're *obviously* wrong, but you were still defending the point. You are now an admitted liar.
More goalpost shifting. Why don't you stick to something no one knows anything about. You can honestly be wrong, there.
No strawman, liar. You claimed a uC couldn't solve the OP's problem. Stupid!
More lies.
I didn't have to waste the time. You just admitted to being a liar.
(PST),BillSloman
happenedBillSloman
:and doesn't i=
point out -
be. JA can
out-think you wit=
when you say
was delighted
was.
If you can't
sci.electronics.design, you
10Hzcells."
that.
"a
about
mA at
at
He is unquestionably an idiot. However, he is also a liar. Not my fault that Slowman is what he is.
happenedBillSloman
happenedBillSlom=
doesn'=
you =
say
deligh=
can't
you
Did I make such a claim?
Actually, I have done microwatt digital designs, and 100 milliwatts isn't bad for the measurement, filtering, protection, and PID loops that the VME module does. We took the power from unacceptable to fine by making the 100 MHz, 32-bit processor sleep until ADC data was available, and ripping through an optimized ISR when it was. Which was the technique that you questioned.
If I needed a 100 uW uP design, I'd do one.
You don't need one. You don't design electronics.
John
Have something to add? Share your thoughts — no account required.
Ask the community — no account required