I am new to Arduino and starting to equip myself in electronics parts etc.
For learning electronics ( I know some basics) a friend suggested Boylesdad's electronic devices and circuit theory (have started reading it). I have an engineering and computer science background, so the science and physics behind electronics interests me as well as the design/ tinkering part.
My questions
I would like suggestions on other quality publications to learn electronics: such as other well respect books and magazines, websites.
As far as equipment and parts that are desirable in a hobbyist lab, what is desirable to have?
so far I have stuff like
varied resistors, caps, diodes. Some mosfets, leds, some chips for arduino such as eeproms, i2c stuff, shift registers, arduino starter kit
A bench power supply with old school analog meters is a great addition especially if the current limiting works well. Digital meters are kinda out of it in that you are trapped by the sampling rate of the digital meter. If the current is changing in a cyclic manner and it coincides with the sampling rate you won't see the variations..
An oscilloscope is kinda mandatory for advanced stuff and a logic analyzer is your best friend for any asynchronous or synchronous communication. I bought a Saleae Logic and it has quickly become my go to tool even clock speeds can be measured.
Docedison:
A bench power supply with old school analog meters is a great addition especially if the current limiting works well. Digital meters are kinda out of it in that you are trapped by the sampling rate of the digital meter
Analog meter?...
Sure, they have their place learning institutes love them (stops people blowing the fuses in digital ones) but I can't see how you think that they're better than a digital one, even a $2 one off of ebay.
Cons of an Analog.
You can't accurately read an analog display (it varies just from the Angle) .
The coil (to drive the deflection) will also have some influence and will take some power
With digital, you're going to get a more accurate figure than from digital.... I still don't understand why resettable fuses are not as standard in digital ones just like their old analog counter parts have.
A lot of the so called accuracy of a digital meter is an illusion, just because a voltage is displayed to three decimal places does not mean it is accurate to three places.
As to the angle of read, that is what the mirror is for on the scale.
An analogue meter is not going to give you silly readings if you have oscillation or noise on the signal.
I'd go for that over an analog, esp when it comes to taking current measurements, accuracy is not a great issue (eg, how far could a voltage measurement be out to the real thing?) if all you're doing is learning.....
Once you understand the basics you have the ability to design a voltage divider and turn an Arduino into a voltage reader.
Sure, they have their place learning institutes love them (stops people blowing the fuses in digital ones) but I can't see how you think that they're better than a digital one, even a $2 one off of ebay.
Cons of an Analog.
You can't accurately read an analog display (it varies just from the Angle) .
The coil (to drive the deflection) will also have some influence and will take some power
99% of the time with a PSU you aren't in the slightest interested in accuracy(*), 1% of the time you are.
100% of the time you want to see the value immediately, DMM's take several seconds to stabilise and
use integrating ADCs that take perhaps 0.75s to sample - although the better ones have a bargraph
display as well that runs at a higher sample rate.
Have you any idea of the power consumption of a good analog meter as used to be used in
multimeters - a few microwatts (example AVO meter 50uA meter with 2500 ohms winding).
Also digital display of a rapidly varying value is impossible to read, with analog you can estimate the DC and AC
components of the signal by eye at least.
(*) Typically with a PSU I'm setting the supply voltage to 5V or to 12V (+/-5%) or I'm varying it to see how current
depends on voltage. Accuracy needed is ~5% which is no problem with analog meters - good meters have a mirror
scale to prevent parallax. If accuracy is important I measure the voltage at the load to remove IR losses in the wires
from the measurement.
Wow, so many responses, I almost started an argument it seems.
I have some crappy digital meters, good enough for most stuff, so I did a little reading, so instead of buying a fluke at many $$$, I got a b&k 2703c that I bought from digikey for 55$. Hopefuly it was a good choice. Looks good so far.
For a bench power supply I converted 2 computer power supplies to get either 12,5 or 3.3v, I also use them to supply my battery chargers.
Is it a good idea to make a variable one?
This post has started giving me ideas, perhaps As a first real arduino project, I can make a bench power supply. I will start another post for this in the projects section with the details.
As far as an oscilloscope goes, many people suggest a stand alone one vs a computer usb one. Any comments on this so I can make an informed choice? Brands?
You can probably get an analog dual trace 'scope, 20MHz, for next to nothing or even nothing, depending on where you live. You'll regret it if you waste money on a 1MHz USB 'scope. Get something with real x10 probes.
A good old analog meter is better than a cheap new digital. Digital meters can give you really strange readings when a lot of noise is present. A variable power supply with analog meters is better than one for digital, as pointed out you can tell at a glance what the voltage is set at and what the current draw is.
You can add a variable output to that converted PC supply with an LM317 on a heat sink. Don't tie it to one of the heat sinks in the PC supply. There are sellers on Amazon and eBay that can sell you a 10k 10 turn potentiometer and a couple of meters for current and voltage.
There are some who might say that the analog meter is on its way out, but they would be wrong. As mentioned earlier, the analog meter is almost immune to EMI (electromagnetic interference). In addition to this it is very good for showing changes in electrical quantities.
There are many cases in which an adjustment must be made for maximum or minimum current or voltage. While it is possible to use a digital readout for such an adjustment, an analog meter makes it much easier.
Suppose you are adjusting a control for a minimum current. When using an analog meter you do not actually read the scale of the meter. You watch the pointer moving to the left as you turn the control. When the pointer starts moving to the right, you reverse direction on the control and bring the pointer back to its left-most position. It is a matter of eye-hand coordination.
On the other, hand if you are using a digital readout to make the same adjustment, you do have to read the number on the display. As you make the adjustment you continuously read the number and do a comparison to the previous one. It's no longer a matter of eye-hand coordination; now the mind must remember a number and do calculations of sorts: "Is this number larger than or smaller than the other one?" This remembering and calculating takes more time and requires more mental effort than does eye-hand coordination.
"But wait a minute" I hear some of you saying. "What about bar graph displays?" Bar graph displays usually have ten elements which gives only 10% resolution. In tuning the output circuit of a radio transmitter the capacitor is adjusted for minimum amplifier current. This setting gives maximum power output and maximum efficiency of the amplifier. If a bar graph were used for this purpose the amplifier current would have to change by 10% of full-scale before any change could be detected by the operator. If a transmitter's output stage is operated 10% "off the dip" the output could be down by as much as 30% and the output amplifier could even be damaged.
This is but one example; there are many others in the field of electronics. It can be argued that there is no reason why a bar graph must be limited to ten elements. There is a reason, money. To match the resolution of an analog meter a bar graph would have to have at least 50 elements and 100 would be preferred. At the present state of the art, a 50 or 100 element bar graph readout is so costly as to be unfeasible. And don't forget that matter of EMI. Analog meter readouts will be with us for many years to come.
In service work there are many service adjustments which require making an adjustment for zero, minimum or maximum voltage or current. That is one of the strongest arguments for keeping a VOM on the service bench.
I almost started an argument it seems.