[Nebby]

The one feature I'd like to implement in my built-sometime-in-the-future amp would be a power LED that changes color to reflect the warm-up state. With the temp probe ability, it sure sounds like it will be do-able

 

[linuxworks]

if you use an rgb led, you can 'burn' 3 PWM lines on the arduino and get exact color (within 255 pwm-based levels of r,g,b) and intensity

sure, its do-able.

 

[pixeljedi]

Slightly off-topic / slightly related*

If anyone wants to try their hand with Arduino programming, Sparkfun Electronics is having a free day on Jan 7th. A good opportunity if you've been holding off due to lack of fundage.

(*related in the sense that if more people on the forum can program Arduino's, more people can contribute)

 

[manaox2]

Quote:

Originally Posted by pixeljedi /img/forum/go_quote.gif Slightly off-topic / slightly related* If anyone wants to try their hand with Arduino programming, Sparkfun Electronics is having a free day on Jan 7th. A good opportunity if you've been holding off due to lack of fundage. (*related in the sense that if more people on the forum can program Arduino's, more people can contribute)

Seriously great opportunity, thanks! I'm going to pick up a pint glass while I'm at it. http://www.sparkfun.com/commerce/pro...oducts_id=8405

 

[ShinyFalcon]

This is awesome! Something to do over the holidays. Though I'm not sure what to look for there... I guess I'll take a look around.

Arduino is C/C++ centered, yes? All I have worked on is Java, but I suppose it won't take me long to adapt to C++.

 

[linuxworks]

its c/c++ based.

not a huge amount of c++ (thankfully). I'm converting a lot of chunks of code over to c++ 'classes' but that's about all the c++ I use.

if you know java, you'll be able to follow along well enough.

the compiler (error line # stuff) is useless, though. it points you to wild places that aren't even close. solution: don't write code that won't compile (lol).

 

[m1abrams]

Quote:

Originally Posted by linuxworks /img/forum/go_quote.gif if you know java, you'll be able to follow along well enough.

Except if you have never dealt with pointers and dereferencing, and pass by reference versus by value. Those items tend to give most Java guys the hardest time (that and the many symbols required for it).

On the plus side you can overwrite your operators

 

[pixeljedi]

Quote:

Originally Posted by m1abrams /img/forum/go_quote.gif Except if you have never dealt with pointers and dereferencing, and pass by reference versus by value. Those items tend to give most Java guys the hardest time (that and the many symbols required for it). On the plus side you can overwrite your operators

I'm just glad it's not assembly! Anything based off C is fine by me

 

[linuxworks]

assembly turns me off. I started with it, some 30+ yrs ago but I'm MUCH happier with a higher level language like C.

with the several thousand lines of C in volumaster, for example, there's no way I could write this much if I had to do it only in assembler!

getting away from the low level programming languages let you concentrate on the apps functionality, itself, where your time is better spent, anyway.

 

[GeWa]

Awfully quiet here?

Regards

 

[linuxworks]

the software (firmware) continues to improve and I'm modularizing a lot of it so that the main routines are smaller and contain mostly just the parts that make them special (if its an spdif switch or analog switch or vol control or MPD controller or even espresso machine controller)

by moving a lot of the other parts into 'class libraries' it keeps the main parts smaller and easier to read. so, I'm working on that, in the background - some software re-org and cleanup.

also, turning my attention to relay attenuation control circuits. I have a sample that I built on perf board; but the first version had some issues that needed a redesign to fix them. I hacked my perf board for the redesign and its almost ready for me to test with. after that, the code needs another major update (talking to a bunch of latching relays is quite quite different than talking to a solid state PGA vol control chip, for example). with the PGA case, you talk to the chip and you get what you want pretty much immediately. with relays, I have to setup a 'work queue' and pulse some relays on and off at various times and do that at various time during the whole main polling loop. its a very different model than the PGA chip case. and add to that a motorized pot that keeps trying to play catch-up with the almost instantaneous lcd display and you get some interesting timing loops going on

so, some work on the lcd unit, itself, but more work actually on the functional modules or 'things that the lcduino controls'. we're thinking that the relay attenuator (for vol control) and an input/output selector (spdif for digital and relay based for analog) would be the first 2 'personality modules' that would follow the lcduino-1.

 

[GeWa]

Thanks for the update.
I'm really impressed with al the work you are doing here.

Regards

 

[TimmyMac]

Would this thing be able to control a soft-start circuit for high power amps too? I'm thinking of having a small transformer to run the brains of my amp, and then using that to open/close relays so that it would first connect the big transformer through a resistor and then a few hundred ms later remove the resistor. Sort of a 'standby', where only the low-power stuff is on, vs 'on' mode, where going to 'on' activates the soft start routine. Possible?

 

[linuxworks]

controllers are a nice alternative to discrete logic. it may be cheaper (time and parts) to write software to do some jobs than to hard-connect gates, etc

what you are asking is time-based, right? or do you want to measure something and trigger the relay closure on *that* ? you could - you would not have to play 'guess the time constant' games, then. that also helps if you turn things off then on again - the processor could do something smart then, that discretes might not be able to easily do.

another thing to think about (maybe) is a SSR instead of mech relay. they don't go down to zero but they are softer than a 'harsh' relay closure (maybe).

 

[TimmyMac]

Quote:

Originally Posted by linuxworks /img/forum/go_quote.gif controllers are a nice alternative to discrete logic. it may be cheaper (time and parts) to write software to do some jobs than to hard-connect gates, etc what you are asking is time-based, right? or do you want to measure something and trigger the relay closure on *that* ? you could - you would not have to play 'guess the time constant' games, then. that also helps if you turn things off then on again - the processor could do something smart then, that discretes might not be able to easily do. another thing to think about (maybe) is a SSR instead of mech relay. they don't go down to zero but they are softer than a 'harsh' relay closure (maybe).

Time-based should be fine - I just need something to stop the lights from dimming when I turn the amp on. I mostly want an alternative to this:

Something like this:
press on button (remote or physical button)
controller closes relay to connect power through 50R resistor
wait (200ms or so) to charge a good fraction of a farad of power caps
controller closes relay to connect power directly, bypassing resistor
opens both relays when you press off button