Any you you DIYnos who good at electronics? I do not mean; "I

I've designed a few amplifiers, effect pedals, and power supplies, semi-"from scratch," but I sure as hell don't use any calculus. The only math I do is some simple back-of-the-envelope stuff before i throw something into SPICE. Learning how to do real EE math would probably help me put together higher order filters much faster than the roundabout trial-and-error method I tend to use, though. I took calc through differential equations in school but that was a while ago and I'm not really sure how to apply it to circuits.

It's very hard these days to come up with an "original" design in the realm of audio amps and effects, so most of the shit i've tried is usually over-convoluted and doesn't end up working as I imagined. Right now I'm trying to figure out how to get a PLL working with a guitar as input for some synth-y sounding and frequency shifting stuff.

>>1041437

File: DSC_0130.jpg (3MB, 3840x2160px) Image search: [Google]

3MB, 3840x2160px

>>1041423
I would say, calculus on the back of an envelope can bring you pretty far in a project.
But for amplifiers, both linear and switching, I would do a bit more calculus, just to ensure losses and gains are correct. But most importantly, the stability when feedback is applied. Here I'm falling a bit short on experience, though.

Was it just linear power supplies?
I have done a few switching supplies already, some more well designed than others..

I'm actually going to study the PLL my self. Need it at my job and for a spare time project with my old supervisor. They are pretty cool.

I would say it is easy to come up with new designs, but not new topologies.

What are the power ratings of your amplifiers?

>>1041446
Are you talking about calculus specifically, or math in general? I still have done any calc in electronic design

>>1041447
Sorry if I use the term incorrectly.

I mean calculating(estimating) a set of performance indicators, like efficency or bandwidth. Based on the topology, component selection, min and max signals or voltage, and a lot of other stuff..

Or do you just find a schematic online and order component based on a third party design?

>>1041446
All the amps and supplies were fairly low power, under 200W; class B amps and linear or unregulated supplies, with the exception of one switching supply I built to get +-20V out of a car battery/alternator.

The issue with "effect pedals" is that all of the simple "sound effects" have been done into the ground. Even if I come up with a new design, it's going to sound exactly like something originally designed in the 60s, unless I start reaching for far places like using PLLs and other digital chips in otherwise "analog" circuits. My first non-clone pedal was a compressor that used analog signal multiplexers and an array of resistors as the variable gain element, instead of a VCA or something. Naturally, this functions "worse" from an engineering perspective as the transitions are not smooth; but the point isn't to make it sound "good" as the popularity of distortion effects will attest to.

>>1041453
No, I do the math. But that's mostly just +,-,*,/ since I work in DC. AC gets a little more complicated but I still don't think it has calculus, which specifically involves integrals and derivatives.

>>1041456
Sounds like great projects.

Was the +/- 20v isolated or?
Did you use integrated switcher or old style PWM controller?

My impression of many "audiophile" folks, are they are kind of religious.
Arguing that they can hear the difference between main cables, gold vs. Cu cables.

>>1041460
AC and transient analysis has things like Fourier transform, complex poles and integral transformations like Laplace or Z-transform which build on derivatives, integrals and limits.

File: EMG.jpg (250KB, 1726x585px) Image search: [Google]

250KB, 1726x585px

Ah was wondering why you have so many RF connectors being test points.

For me it's my 2nd "big" project, an EMG amplifier which measures 50mm x 30mm. Am currently waiting for the PCB's to arrive from OSHpark, at least the new Deus Ex game will keep me entertained while that happens.

>>1041414
i did just make a pedal kit.

fuck you for belittling me you fucking nigger faggot

i hope you get cancer and die, cunt

>>1041907
The connectors are SMB, they offer quick and reliable connection. The evaluation module are used for third semester EE students..

What are the purpose of the amplifier?. The layout looks cool. Did you use altium?.

>>1041925
Some one seemes to be triggered..

>>1041925
There's absolutely nothing wrong with pedal kits. Everyone has to start somewhere. What's wrong is when people act like they're an experienced EE when all they know how to do is solder together a circuit someone else designed and printed.

>>1041944
Neat. EMG amplifiers are used to amplify muscle activity, i'm hoping to basically activate something with muscle movement. And yuppo it's Altium 16

>>1041947

Where do you sit when you can make PCBs in kicad that are mostly copies of reference designs modified for what you need, and then getting them made and working?

Feels like that's just the super lazy cheap version of putting together a kit. I mean screw protoboard, PCBs are dirt cheap.

File: IMG_20160509_003623[1].jpg (1MB, 2592x1944px) Image search: [Google]

1MB, 2592x1944px

>>1041414
Definitely most proud of my audio spectrum analyzer so far. I have some bigger projects in the works at the moment but they aren't completed yet.

I designed most of the circuit myself however I did have some help with certain aspects of the design, I used resources from ESP for some of the MFB filter calculations, and I took the peak detector design from the LED driver datasheet with little modification.

The board was entirely laid out by me with orCAD capture and PCB editor and I also populated it.

1/3

File: IMG_20160507_182738.jpg (2MB, 2592x1944px) Image search: [Google]

2MB, 2592x1944px

>>1042266

File: IMG_20160507_111554.jpg (3MB, 2592x1944px) Image search: [Google]

3MB, 2592x1944px

>>1042267
3/3

File: IMG_20160507_111627.jpg (2MB, 2592x1944px) Image search: [Google]

2MB, 2592x1944px

>>1042268
4/3

>>1042266
>>1042267
>>1042268
>>1042269

Neat, now do it all using SMD parts ;^)

>>1042439
Would do it if I had a steady enough hand for SMD soldering.

I would actually prefer to re-engineer the whole thing using DSP instead of analog signal processing but it'd require an mp3/AAC/etc. audio decoder which I'm not too sure how to implement. The DAC and the digital filtering shouldn't be too bad though.

>>1042473
Smd isn't as scary as it seems. Watch some YouTube videos on the tap and drag method. Some chipquik helps a ton for passives. (Great for non smd parts too)

You don't need a audio decoder you feed in a analog signal and use a ADC (or feed in i2s or whatever) unless you're making a player you don't need to worry about the file codecs.

>>1042475
I want it to interface with a portable music player over USB not aux, to do that I would need a decoder and DAC.

>>1042476

Ah okay that's different then

File: 2015-11-09 11.12.18.jpg (3MB, 4465x3353px) Image search: [Google]

3MB, 4465x3353px

>>1042124
Sound like a cool project, I hope it will work without too much trouble!

>>1042266
Nice! But them Through hole components though!..
I try to completely avoid TH components, besides for large caps, and power components.
As mentioned, SMD components are not that bad, especially if you work with 1206 or SOIC packages. 0201 and MSOP10 can be a pain in the neck..

When i do a layout, i usually make a solder mask sencil. The stencil is the cut in regular paper on a laser cutter. (I know steel stencils can be ordered, but a bit expensive for 2-4 prototypes.)
Once i have all components and my stencil, i apply solder paste through the stencil, and add components. Then i bake the PCB in a SMD oven (a regular oven will do fine).

If the pads, clearances and past is some what correct, the tolerances of component placement can be surprisingly poor, and you will still see a good result!.

- Yet again nice PCBs! Hope you manage to solve the DAC and decoder problem.

Pic relevant, SMD paste before soldering

File: 2015-11-09 12.17.43.jpg (2MB, 2653x2241px) Image search: [Google]

2MB, 2653x2241px

>>1042512
with components

how would one start designing circuits, and learning more about building electronics in general? i have a basic grasp, but i don't think i could do much of anything in this thread.

>>1041414
I design ICs to make a living. Currently I'm working on a switched capacitor DC-DC converter which has no external elements. I really enjoy doing these, but I'm not sure if it counts as a hobby.

File: aZMxG59_460s.jpg (52KB, 460x663px) Image search: [Google]

52KB, 460x663px

>>1041414
I've designed a few things myself.

-sensitive photodetector amplifier with capacitor integrator built in
-triac dimmers for stage lights
-complete coil gun including power supply, timers, trigger, and rifle assembly
-constant current laser diode drivers
-rfid antenna
-man wearable virtual reality system
-some custom audio effects including analog chorus effect.

Most of the math I use is simple algebra, although a little calculus has come up from time to time. When designing functions from a chip in to a system, it's hard to say that you really designed it, since the function is already on a chip and an example circuit is already in the datasheet. Most of the work is in tying everything together in to a system with a reasonable human interface. The thing that really makes it your circuit is your custom solutions for noise and power and user interface, and a big part of that is expressed in packaging rather than in circuit design.

Ultimately, it's hard to say that you really designed anything new unless you are designing silicone wafers.

>>1042627
Read, follow tutorials, read, build stuff on breadboards, take things apart, read, build things, watch eevblog videos and other youtube stuff , read, build.

Best way to learn is try and fail. I bet everyone here had hundreds if not thousands of failed projects and experiments for every successful one

>>1042628

Sounds both boring and super interesting. Probably is both.

What do you do it with to prototype and what not?

>>1042767
It's both. Coming up with interesting things pays well and is fun, but solving issues and debugging layout problems is super boring.

Currently I'm designing a fully integrated charge pump with a voltage control loop I designed. It's a digital control loop (almost everyone uses analog integrators) and it uses a hybrid control method to regulate the output voltage.

This is the first prototype of this kind, I'm not sure what will happen in the future though. But there's definitely a publication coming. I don't care about what the company thinks.

But as I said, since I'm doing this professionally I'm not sure if it counts as a hobby. Income aside I do enjoy it a lot. I get a nice feeling in my gut when things work as I want them to work.

>>1041414
Tell me more about that 3 phase motor driver, I've been looking into using some of those 3 phase bldc motors from quad-copters for general robotics shit. Mostly research right now but I have found a ton of interesting reading on things like anti-cogging algorithms and driving them at low speeds/high torque, closed loop control and all that good stuff.

>>1042775
Sounds pretty cool dude. If I could ask you a meta question, what's your outlook on starting an IC design career these days, given the waves of consolidation in the industry?

Good designers are always in demand. A lot of companies are doing alright in revenue and profit but currently, as far as I've observed, they're a bit cutting on hiring. I don't know about the future though.

I meant this >>1042847 as a response to this >>1042807

File: hr8MsXr.jpg (191KB, 1154x1184px) Image search: [Google]

191KB, 1154x1184px

>>1042847
Yea it seems these days companies would rather absorb other entities to expand their portfolio

>>1042851

Tower of power!

>>1042627
YouTube have a lot of good information. Dave from eevblog might talk a lot, but he often talk about interesting stuff.

As mentioned earlier, you will read quite a bit, i havr used books, wiki and application note a lot.. You will also have to invest a lot of time and some cash on projects.. The first project can be taking stuff apart, then do some simple stuff on bread boards..
Once you have got a hang on component types, the look at YouTube for university lectures, then you learn some of the important theory..

>>1042628
At uni or at work?

I have always been quite impressed with IC design. At my department at uni, some of the guys made integrated Dc/DC converts operating at Ghz.. That blew my mind..

>>1042634
Thats a lot of cool projects.

>>1042790
This is only the power stage.
It is basically just 3 half bridges.

The bridge is rated for 48v, the current is low, primarily due to thermal limitations and the output filter inductor.

The minimum switching frequency is around 2kHz.
It is determined by the bootstrap circuit.

The halfbridge is two N channel mosfets. The driver has controllable dead time, and complementary output. So only a single input per half bridge.

A halfbridge is not that difficult to construct, you just have to decouple the bridge with low inductance loops, then ensure low inductance between the gate driver and the mosfet.

The rest is mentioned in the dataset of a halfbridge mosfet driver or in application notes..

Dead time is an important parameter, as it prevents shoot through, but too much gives distortion, losses and perhaps EMI problems..

>>1042851
Nice stack..

Have you ever looked at the de-rating of capacitance versus voltage of MLCC?..
Even X7R can have as little as 30% initial capacitance at rated voltage..

>>1043113
>>1043894
Not mine, and yea read that Maxim article a while back, am going to stick with C0G's for all precision designs

>>1045552
I haven't seen that article.
But as long a you do not use type two ceramics for timing or signal filters, then you should be home safe..

File: RooseveltsBlueTooth.jpg (73KB, 1082x852px) Image search: [Google]

73KB, 1082x852px

>>1042512
>>1042517
if your bag o' smd components didn't come with an inhalation warning you're doing it wrong