Basic must-have ICs

[Sa1tama]

I am planning to buy some components for my very basic lab and i was wondering what ICs shoul i get. (NE555 is already in my cart). I know i should get some OP Amps, but i don't know which model i should get.

Also i am building a DAC and i was wondering if there was any OP Amps that can drive 1.5w speakers.

Thanks in advance!

[straight_stewie]

1 hour ago, _SAITAMA_ said:

am planning to buy some components for my very basic lab and i was wondering what ICs shoul i get

That depends on what stuff you want to work on in your very basic lab.

I keep a couple of FPGA boards and a few cheaper microcontrollers around, along with an assortment of basic logic chips, and even some radio and GPS equipment. Of course, my interests are processor design and vehicle guidance.

What are your interests in the field?

[Sa1tama]

2 minutes ago, straight_stewie said:

That depends on what stuff you want to work on in your very basic lab.

I keep a couple of FPGA boards and a few cheaper microcontrollers around, along with an assortment of basic logic chips, and even some radio and GPS equipment. Of course, my interests are processor design and vehicle guidance.

What are your interests in the field?

 

I like to build some amplifier circuits for some reason.I am currently trying to build a Class D amp for 80+80Watt stereo 6 ohm speakers (It's the most complicated thing i have ever tried to do)

[Unimportant]

4 hours ago, _SAITAMA_ said:

I am planning to buy some components for my very basic lab and i was wondering what ICs shoul i get. (NE555 is already in my cart). I know i should get some OP Amps, but i don't know which model i should get.

Also i am building a DAC and i was wondering if there was any OP Amps that can drive 1.5w speakers. 

Thanks in advance!

The most known op-amp for audio purposes is the NE5532. It's a somewhat older chip and better ones are available by now but it performs great and is cheap as chips, pun intended. Such an op-amp cannot drive a 1.5W load directly - off course - but when buffered with a BD139/BD140 pair it should be able to easily reach 10W RMS with a +12V/-12V power supply. Something along these lines:

 

2 hours ago, _SAITAMA_ said:

I am currently trying to build a Class D amp for 80+80Watt stereo 6 ohm speakers

Class D amplifiers are a whole different ball game. 80 Watt RMS with a 6 ohm load will require a +35V/-35V power supply for a half bridge or a single 35V supply for a full bridge. (Theoretically that is 100W RMS but losses and the fact that duty cycle cannot be to close to 100% for the bootstrapping make that you cannot swing all the way to the rails). Peak currents will be about 6 Amps and you want your MOSFET's to transition extremely fast to keep efficiency high and distortion low. Transition times of 50nS are not uncommon. Switching 6 Amps in 50nS makes dI/dt go off the charts so make sure you realize you can't be mucking around with the power stage on some experiment board. You'll need to design a proper PCB that minimizes parasitic inductances (that means carefull layout and SMD components) from the get-go. I'm not kidding, if you cobble this together on a experiment board even a measly 50nH of stray inductance in the wrong place can create negative voltage spikes of 6V on the switch node which will kill your gate driver. (due to latch-up and overcharging the bootstrap capacitor). Other important topics you can study are snubbers across the MOSFET's, A Zobel network on the output (especially important for a class D as it needs to act as filter damping in the frequency ranges where the speaker's impedance goes up and stops damping), and avoiding resonance in the output LC filter creating high voltage spikes that can cause damage. Proper power supply bypassing is important as well and probably a field of study in and off itself.

 

Furthermore you will need decent equipment, a oscilloscope in particular because obviously you need to be able to see what is happening. We're talking a real oscilloscope here, not one of those toy kits on ebay. You need to be switching at about 400-500 Khz at least in order to reach the common -40db attenuation target in your LC output filter. You want to capture a lot of harmonics to see the actual square wave and not some scope-bandwidth-limited approximation so I'd say a 50Mhz bandwidth scope should be the minimum. Don't skimp out on the amount of channels either because you'll be making differential measurements to minimize noise pickup. (A class D amplifier is a great project to learn proper scope probing techniques - otherwise you'll see all kinds of crap that is not really there but simply pickup from the probes.) To top it off you'll probably want a good FFT in there so it can act as a poor man's spectrum analyzer. Rigol and Siglent make some superb scopes at affordable prices. Expect something between 500$ and 1000$ tough.  

 

As for parts. You'll obviously need MOSFET's. Study them first, there's a lot to know about these things (although you might want to take a look at the IRFI4212, which is especially designed for class-D amplifiers and combines both MOSFET's in a single package, thereby minimizing stray inductance in the critical S1/D2 switch node). Then you'll need half bridge gate drivers. These provide the high dI/dt, high current gate drive signals for the MOSFET's which allows them to transition so fast. They also provide the bootstrapping circuitry which allows using a superior N-Channel as the top MOSFET in stead of a P-Channel. Some gate drivers also provide either programmable dead-time or adaptive dead-time. I'm particularly fond of the Micrel/Microchip MIC4605 which provides automatic dead time and is very cheap but the commonly used hobbyist part seems to be the IR2110, which provides neither programmable nor adaptive dead time (requires external dead-time circuitry).

 

As for the other parts, it depends on the topology you're aiming for. A self oscillating design or the trusty old triangle wave + comparator approach ?  The latter can work open-loop (without negative feedback), so you can try to make things work without feedback first and try adding feedback later which I would recommend (Stable feedback is the real rabbit hole here).

[Hackentosher]

If it were me, I'd get some ATTiny85 boards and some TP4056 boards

[ShredBird]

I would highly recommend getting "The Art of Electronics" by Horrowitz and Hill.  This book covers how to build very practical analog circuits and it identifies plenty of off the shelf parts that are very well rounded.

As for OP-AMPs, the cheapest go-to amp for basic use is the LM741.  If you need something a bit better go for an LF411.  If you need something precise, but still inexpensive, go for an OP27.

[Pixel5]

as other have already mentioned it really depends on what you wanna do.

 

generally i stocked up with many random parts like 7805 Voltage regulators, assortments of capacitors, resistors, buttons LEDs and some micro controllers.

most of the stuff i got is just here in case i need it especially these basic components like capacitors and resistors.

[corrado33]

10 minutes ago, James Evens said:

old AVR chips? STM offers much better chips in this price range with ARM M architecture.

 

Beside the standard passives I mostly have left over from projects. Sure  555, 817b, 7400 series ICs, transistor (2N3904, 2N3906 and some more), linear voltage regulators, comperator and some opamps are part of it. For audio there are amplifier ICs.

 

Nice to have is a collection of various µC  boards and your favorite FPGA if  µC don't work good enough.

µC i use at the moment: tiny STM8, STM32F103, ESP8266, ESP32 What is missing is a very lower power STM32L and a ARM H7 µC.

left overs: Atmel 328p, 32u4, attiny 85, Mega 2560, tube of tiny AVR i don't remember the part number of

I mean... it's not like the Atmel chips are expensive... (I used to play around with ATMega8)

 

As for some necessary ICs or the like, definitely some linear and/or switching voltage regulators. You can find switching regulators that accept up to like 30V and output down to 3V without much need for a heatsink. Linear regulators, on the other hand, no chance. 

[Hackentosher]

3 hours ago, James Evens said:

old AVR chips? STM offers much better chips in this price range with ARM M architecture.

 

Oh absolutely there are better chips, but the attiny85 boards such as the adafruit trinket, or even the ol' ebay special are great. They're cheap as beans, and tiny. Great for small projects. If more power or io is needed, I love the adafruit feather as well.