Analog vs digital display

[trebejo]

This question is usually addressed inside threads, maybe it is too obvious, but I thought it deserved its own little space. I don't think I saw an explicit thread on this.

I have a nice bench power supply with digital display (it is a Korad KD3005P but that's not important). It will tell me that it is putting out xx.xx volts, x.xxx amps, and according to the multimeter, it's not too bad. If I want 5 volts, it will give me something between 4.9951 to 5.0049 volts--no other way to refine it.

Then the other day someone gave me a power supply as a thank-you gift for buying their scope. It has a many-turn potentiometer for setting the voltage or current level (two separate knobs and a switch to select which one is active). The level is indicated with a needle that swings back and forth. With some patience--and a multimeter--one is able to set far more precise values. As for the value staying put, well...it does, but a little readjustment is required sometimes. I have not tested it enough to say how perfect that is, but it seems OK.

I think I was lazy before, when I assumed that the LED was a better way to go. But apparently the industry has left multiturn knobs behind. Am I just being temporarily retro, or will analog knobs always beat digital LEDs?

For the moment, I'm not scoping out killer deals if the display is digital, and of course it's personal, ymmv and all that jazz. But in general, if you want your multimeter to read 5.0000xx without paying a LOT of money, don't you have to have a good ol' analog potentiometer?

[Benta]

I don't quite understand your post, and I don't understand your question.
But one thing is clear to me:
Your are confusing resolution with precision. That a digital display can show any number of digits does not mean that they are correct.
What "LED was a better way to go" means is undecipherable in this context.

[tggzzz]

I think I was lazy before, when I assumed that the LED was a better way to go. But apparently the industry has left multiturn knobs behind. Am I just being temporarily retro, or will analog knobs always beat digital LEDs?

Welcome to the forum. It is always good to see someone that observes, thinks, and is prepared to change their mind.

Whether analogue or digital displays are more appropriate depends on the use-case.

Digital is usually better when you need more resolution of a static variable. Analogue can be better for quick glances, guick measurements, or where a variable is varying and you want to observe the trend.

There are some cases where analogue is much better. Try measuring charge with a digital meter; it is easy with a ballistic galvanometer. Try understanding an aircraft's orientation from a digital display showing 32, 36, 43, 45, 48, 50, 51 degrees, or an artificial horizon that shows "/".

A note about scopes...

The distinctions are between analogue and sampling/digitising scopes, and conventional vs storage scopes. There are sampling analogue scopes and analogue storage scopes, as well as the common conventional analogue scope.

There are still a few use-cases where sampling analogue scopes are the best technology. Conventional analogue scopes are usable for many purposes. Analogue storage scopes always were a pain in the backside. Digitising storage scopes made in the past decade are pretty good; ones made earlier often had significant disadvantage.

[BeBuLamar]

I would prefer the digital one. You can easily set the output voltage to within 0.01V while the multi turn pot allow you to get finer than that but.... I bet it's not easy and took you some time to get it. And really depending what you use the power supply for. You normally do not need a power supply with precise voltage in order to power something. If you use the power supply as a voltage reference for something (that requires very little current flow) you can put a potentiometer after the power supply output and fine tune to the exact value.

[tggzzz]

I would prefer the digital one. You can easily set the output voltage to within 0.01V while the multi turn pot allow you to get finer than that but.... I bet it's not easy and took you some time to get it. And really depending what you use the power supply for. You normally do not need a power supply with precise voltage in order to power something. If you use the power supply as a voltage reference for something (that requires very little current flow) you can put a potentiometer after the power supply output and fine tune to the exact value.

Don't confuse accuracy precison repeatability and resolution.

If you don't need high resolution, then an analogue control is as good as a digital control.

If you want to change an input and see how an output changes, usually an analogue control is faster and more intuitive.

[bd139]

Big fan of analogue controls for this. But with digital readouts. That’s a good compromise.

Surprisingly most of the time you don’t need that much accuracy though. If you give someone enough zeroes they will want to see 5.00000 volts on their power supply or they won’t sleep. 5.0 is probably close enough.

[CatalinaWOW]

Better requires definition.  Once upon a time analog displays were "better" because they were cheaper and more widely available than digital.  Now the reverse is true.  Earlier posts have mentioned or implied other definitions of better.  Some more possibilities - Digital is generally easier to read across the room.  Accurate readings on analog actually require training and care, usually not true on digital displays.  Only you can decide what is better.

You might also like to look under the hood to find some of the subtle details hidden in your question.  You mention a multi-turn potentiometer for setting voltage.  In many cases such potentiometers involved a 10:1 gear reduction mounted to a wound rheostat.  Since the wiper of a rheostat only touches one side of the coiled resistance wire there are step changes in the setting as the wiper moves between coils.  The "analog" actually has discrete "digital steps".  With a good voltmeter and some care you can observe these steps.

[Doctorandus_P]

Generic power supplies usually are not very precise / accurate. A resolution of 10mV is already considered good.
The Korad is also not built for accuracy. The accuracy you get is already quite amazing for an EUR100 power supply.

I wonder what that other power supply is. Between roughly the '70-ies an 2k  some amazing analog precision power supplies have been made.

If you want both resolution, accuracy and precision in a power supply, then you can also buy them, but this class of power supply has a special name. It's called a SMU (Source Measurement Unit)
The Keithley 2400 is a popular example of this, but you won't like it's price

It's quite nice to have some different kinds of power supplies. That way you can mix and match them to best fit the task you want to use them for.

[james_s]

Either one is fine but generally I prefer analog for this sort of thing. It's very rare that I need an extremely precise voltage or current from a power supply, usually I just need something close and analog is better for noticing quick changes at a glance.

[trebejo]

I should have chosen the subject title better, specially considering that I thought about it a bit.

I think I got the metaphor now, don't know why I missed it since it applies to everything else in the universe. Use the big thing to get somewhat close to the end, then finish it off with a smaller thing, possibly home-cooked, probably already posted MANY TIMES already on this very forum... I can watch Dave's jellybean regulator video (again), read the datasheets, read the metrology forum here, experiment...demanding that it all be finished by the knob on the power supply just misses the point(s).

And oh yes, I had to be patient and "creative" to manually get the voltage to get a 10.0000x reading. Getting an LED to say "10.00" is much easier and, in its own way, more reliable.

And then...it turns out that if I turn a knob seeking the analog experience, perhaps a contact might just be hopping from place to place?

To do all with just one box is expensive. Keithley SMU expensive. Ah. Nice instrument!

So love the knob, but not for the many 0s! Love it because it provides continuous manipulation of the voltage level. Makes sense.

Just for completeness' sake, the old power supply was made by Power Mate Corp. (Hackensack NJ), couldn't find a datasheet or manual. It has a Zilog ownership sticker, a calibration sticker from 1985, and two other stickers too worn to read. 0-10V, 0-5A, more or less the same size and weight as the Korad. No model number. Does not look fragile.

[ledtester]

If I want 5 volts, it will give me something between 4.9951 to 5.0049 volts--no other way to refine it.

Bear in mind that the voltage that gets to your application will depend on the current drawn from the supply due to drop along the connecting wires.

If you really need a precise voltage you probably should use a voltage reference IC. They can give stable voltages but can't source a lot of current. They are usually connected to high-impedance inputs within your application and not used to provide power.

... But in general, if you want your multimeter to read 5.0000xx without paying a LOT of money, don't you have to have a good ol' analog potentiometer?

In a way, yes. Voltage references are often tuned to precise voltages using "trim pots" -- small potentiometers that you adjust with a screwdriver.

Re: analog vs. digital displays

As others have mentioned, analog displays are better for capturing trends, flickers, etc. In fact, some multimeters have bar graphs in addition to a digital readout to give the kind of feedback that an analog meter gives.
 

[David Hess]

Then the other day someone gave me a power supply as a thank-you gift for buying their scope. It has a many-turn potentiometer for setting the voltage or current level (two separate knobs and a switch to select which one is active). The level is indicated with a needle that swings back and forth. With some patience--and a multimeter--one is able to set far more precise values. As for the value staying put, well...it does, but a little readjustment is required sometimes. I have not tested it enough to say how perfect that is, but it seems OK.

I think I was lazy before, when I assumed that the LED was a better way to go. But apparently the industry has left multiturn knobs behind. Am I just being temporarily retro, or will analog knobs always beat digital LEDs?

Power supply applications often do not require better settability than a single turn potentiometer will provide, about 1 in 200, and a single turn potentiometer is very convenient.  10-turn wirewound potentiometer provide better settability and drift but are not as convenient.  Some old power supplies used 3-turn wirewound potentiometers.  My preferred configuration is separate single turn potentiometers for coarse and fine.  Nothing prevents using analog potentiometers for control with digital readouts.

[Doctorandus_P]

Are you aware you are in danger of becoming a volt nut?

there is more in the metrology department.
https://www.eevblog.com/forum/metrology/

Tread carefully.

[Gyro]

I remember when TTI moved from analogue meters to their first generation with digital displays. Their accuracy was way off in comparison. I'm sure they fixed it in subsequent models though.

Analogue control knobs for me too. They're just easier than typing in digits.

[sahko123]

Analog controls allow for simpler regulator circuitry but can drift due to heat and other effects quite easily so need to be readjusted for accurate values. If you want the trifecta of Precise, Accurate and Resolving you would need digital control with digital readout and a good stable internal reference (not an ltz1000 something basic is fine but a reference none the less).

An analog readout is great for very basic measurements and you know that your fairly close but they arent linear so if you need precision beyond about 0.25V (based on a generic 30V analog scale) than you'll need digital readout. Multiturn pots a great for gettting that resolution with a digital readout. But again you have the issue of drift. In digital control the main limitation is the dac that is used to create the output reference (based on desired value and DAC reference).

[Zero999]

It's sometimes user preference. It's also possible to have a digital system, with an analogue display. This can be a PWM output of a microcontroller driving a moving coil ameter via a suitable resistor, or a stepper motor driven dial. On occasion it might be desirable to have both. The analogue quantity is digitised, displayed on a 7 segment display, with the analogue meter either measuring the value directly, or driven by the microcontroller.

[Terry Bites]

A power supply is not usually designed to be a precsison device. That is the job of a voltage /current reference. Or an SMU if your a bit of a voltnut.  So you only need low res metering.
If your circuit needs ultra precise power supplies its a weak design and will fail in the real world. That's one reason people don't fuss about it too much.
You dont have to settle on one or the other of course. You can have both if you like. On a simple supply that lets you get good idea of trending. Moving coil meters are great at averaging small pulsatile currents without having to add any electonics.