Parts recommendations for my first project

[Bert]

I know most people will recommend I start with a kit of some kind for my first electronics project, but this circuit seems like it will be simple enough to just get the individual parts and solder them into a perf board.

I would like to control an audio amplifier electronically, so I'm looking to replace 18 pots, 2 switches, and 3 three-way switches with digital pots and relays. This circuit will always be plugged into my computer, so I'm hoping to connect to it and power it via a USB B connector. I am pretty overwhelmed by the selection of IC's and other parts when I visit DigiKey, so I was hoping to get some recommendations, mostly for the IC.

Here are some more specific q's I'm having trouble finding answers to:

How do I know what clock speed I need?
How can I tell if I need a separate timing crystal?
Memory size and type?

Any other tips are welcome of course

Thanks,
Rob

[Soeren]

Hi,

I know most people will recommend I start with a kit of some kind for my first electronics project, but this circuit seems like it will be simple enough to just get the individual parts and solder them into a perf board.

"Simple" is when you know how.
When you don't, the right word is "Very complicated"

I would like to control [snip very long list]

Absolutely nowhere near a beginners project.
All you'll accomplish is to burn money.

Here are some more specific q's I'm having trouble finding answers to:
How do I know what clock speed I need?
How can I tell if I need a separate timing crystal?
Memory size and type?

If you cannot answer this yourself, you have a zero percent chance of completing - save money and time by buiding a beginners project. With luck and hard work, you may be ready in some years.

Designing specialized electronics from scratch is not like assembling a kit - you need to be fluent in reading datasheets, understand all the basics (that isn't mentioned in datasheets) and understand how a PCB should be made correctly (forget matrix boards for such things), just to mention a few of the hills that are guaranteed to tip your project.

I bet you won't believe me, judging from your fresh attitude towards making a project with zero experience to base it on, one that really takes an engineer, but do try to get a price quote from an engineering firm, or even better - tell them that you wanna make it yourself with no prior - that'll make their day

Save your money for something sensible.

[Bert]

Thanks for the help and encouragement!

[Daanii]

I have to agree with Soeren. The project you describe will be too hard for a beginner to design for himself. I suspect you would just get frustrated instead of learn anything.

But maybe you can find a schematic, with parts list, for something like that to build. That would teach you a lot, and perhaps get you closer to designing and building something yourself. There are a lot of projects for amplifiers and their controls out there, both on the web and in magazines.

[newInRobotics]

This topic got my attention. I'm no pro (or had any detailed study of the matter), hence the question, why WHOULD it be difficult to achieve what Bert wants? Common sense suggests that if one replaces pots by same value digi pots and switches with relays (with few transistors and diodes) all should be fine (as long as soldering does not mess anything). Is there something more to consider?

Two replies above might be true, however they did not answer questions raised. Here it goes:
Clock speed is not that important as long as you have UART/USART and I2C (for digi pots) capability in the microcontroller of Your choice. Processes in Your application will not be very processor intensive so just go with what is the lowest clock speed that can still support serial comms.

External crystal is required if uC does not have internal oscillator or You want more precise clock speed that uC can offer or You need clock speed that is not offered by internal oscillator of uC. If You go for one of the commonly available ATmega uC (like ATmega48), You will most likely get a choice of 1MHz and 8MHz internall oscillator (no need for external crystal).

By memory, do You mean SRAM or FLASH? Anyhow, You shouldn't worry about that unless You are a terrible programmer and don't use optimisers.

[Bert]

Thank you newInRobotics. Here is another question: Do I even need a microcontroller? This circuit will always be plugged into my pc so is it possible to connect a large serial port directly to a digital pot and write a program that can communicate with the serial port?

[Soeren]

Hi,

[...] why WHOULD it be difficult to achieve what Bert wants?

Bert has never build any electronics and have zero experience/knowledge in neither datasheet reading, construction nor design - according to himself.

New Years eve, Bert posted this

I'm totally new to electronics and robotics and after digging into all this for a couple weeks I am realizing what a colossal task it will be for me to build a circuit and control it from the internet. It seems a number of folks have accomplished this on this forum and I am interested to see if anyone would either design my idea for me or walk me through a design for some cash (and how much cash you'd do it for). There actually won't be a whole lot of robotics involved. Basically all I want to do is control a whole bunch of digital potentiometers and relays through a nice looking flash web interface. Please PM me if interested.

Bert simply asks questions that tells me, that he have no chance whatsoever of doing this on his own. A couple of years of building lesser projects might help some, but Bert doesn't just want to build this project, he wants to design it as well.

I know very talented constructors with many years of experience that goes into self destruct mode if asked to design anything, even a one transistor mic amplifier.

All in all, I think the best advice is to drop it, rather than be disappointed and shy away from electronics completely, after burning a lot of cash on a project that will never fly.

If anybody have time and patience for designing Bert his system, drag him through every step of construction and trouble shooting of the prototype and generally spoon feed him, he may be able to get it done, given enough time.

I know I haven't - I have plenty of larger not-for-myself-nor-for-money projects on the shelf, like an electric miniature 1967 Corvette ("go-cart" like) for my grandson and a kick-behind dolls house for my grand daughter - and to quote the Norwegian poet Bjørnstjerne Bjørnson (which btw. translates to Bear-star Bear-son): "Hvad du evner, kast af i det nærmeste krav"
Which means something like: What free and available resources you have, spend them on what/who your main priorities are" and family, my little gold nuggets in particular, wins hands down, over an all too optimistic newbie stranger thousands of miles away, in getting me to do stuff


Or he can pay his way, but I don't think he can find an engineer that will design and build this for a price that he's ready to pay - research, design, prototyping, layout, construction, boxing up etc. will easily run into the 30..40 hours range and while I have no idea what engineers charge in the US, my time is around $200/hour, if I'm not doing charity, depending somewhat on the exact project.

Beginners should begin with beginners projects (hence the name), then they learn something, rather than end up with a load of expensive junk that they don't understand, but unfortunately, it seems that the less they know, the larger projects they think they can muster - it's crawl, walk, run, no skipping

Common sense suggests that if one replaces pots by same value digi pots and switches with relays (with few transistors and diodes) all should be fine (as long as soldering does not mess anything). Is there something more to consider?

You've been in this long enough to know that this kind of "common sense" is a bit too simple to hold true - how many of your own designs have worked at first try?

Clock speed is not that important as long as you have

Might not be the controller clock speeds that Bert means.

External crystal is required if uC does not have internal oscillator [...]

An external X-tal on the controller will be a very good idea, both for getting reliable comms and for synch'ing purposes - if it goes into audio circuits, even more so.


BTW. The lack of info on the end purpose makes it impossible to find the best solution, which could be something simpler than what Bert figured out.

[Bert]

Soeren, if your time is worth so much, why are you wasting so much of it here. You've posted on two of my threads now and you've done nothing but insult others. If you don't have any real help to offer, then don't post.

[Soeren]

Soeren, if your time is worth so much, why are you wasting so much of it here. You've posted on two of my threads now and you've done nothing but insult others. If you don't have any real help to offer, then don't post.

It's not really any of your business what I do on my time (but the more you earn, the less you have to work ) and I usually don't find it a waste of time helping out here, as most people asking for help appreciate answers bound in reality and actually learn from starting small.

Well, I thought I was helping you save a bundle of money and disappointment, but all I hear now is:
*buhuhu* don't post in MY thread unless you tell me what I want to hear *buhuu*

Why ask if you won't accept any answer not mirroring your own convictions?


Oh, I'll play along on your fantasy then: Sure you can do it and you don't need any prior knowledge - just saddle up sonny-boy - Yehaaa and all that jazz...
I ain't gonna be the your personal tutor for the next couple of years though, but good luck finding one - and have a biscuit to sooth the terrible run in with reality I bestowed upon you.

[Daanii]

As one who has benefited greatly from Soeren's advice over the years, I want to add my two cents. Sometimes people like to be polite and diplomatic. Other people like to be blunt and direct.

On forums like this, I prefer the latter. Even though I too have been on the receiving end of blunt and direct comments. Better to get candid criticism than to get some answer that sounds better but whose message is unclear.

That said, I hope you do stick with your project, or one like it. It will be hard, but you will learn a lot. If you think Soeren is wrong, prove him wrong. Better to do that than just to complain about a person who has provided invaluable advice over years to hundreds of people.