DMM controls

[spanakop]

Given the added design overhead in coming up with a mechanical dial with decent feedback and robustness etc, along with the design of all of the PCB tracks, wouldn't it be simpler to have a bunch of buttons to select each function?

Granted there would be a lot of buttons on the front, but it's surely better than having one rotary dial and you physically moving it around. The DER DE-5000 LCR meter and some Aneng meters are examples of where it works well. It seems odd in that we still have mechanical switches when we have decent electronic switches available that have been used for a long long time.

[BeBuLamar]

The rotary dial would last 30 years or more. I doubt the buttons would last that long.

[Fungus]

It's good having a big mechanical thing on the front that always points to the range in use.

Given the added design overhead in coming up with a mechanical dial with decent feedback and robustness etc, along with the design of all of the PCB tracks

It doesn't seem to be a problem in practice.

Buttons have been tried lots of times. I don't think anybody found the right formula yet.

Fluke even makes a meter with just two positions on the dial (well, three if you count "off" as a position).

[Fungus]

The rotary dial would last 30 years or more. I doubt the buttons would last that long.

They could, but they wouldn't be simple rubber-dome buttons.

They'd have to mechanical, so we're back to cost/complexity of mechanical things with good feedback, etc. The dial might start to look attractive compared to that.

[HKJ]

Given the added design overhead in coming up with a mechanical dial with decent feedback and robustness etc, along with the design of all of the PCB tracks, wouldn't it be simpler to have a bunch of buttons to select each function?

They already exist, both in as cheap meters (Mostly automatic) and as expensive meters. They has one or more relays to handle the selection in addition to electronic switching.

[spanakop]

I have old 80's Casio calculators that are still good and most if not all modern DMM's also have buttons on the top row, so it is possible to have buttons perform functions that are expected to last the life of the device.

True the cheaper auto DMM's and expesive units with relays do exist that have all buttons, but I still think designing and manufacturing a rotary dial function over buttons is still "complex". I can't see any reason other than it's traditionally done this way, to not start offering all button devices now.

[joeqsmith]

My bench DMMs use buttons exclusively.   I bought a couple of very old Fluke 97 scope meters which also only have buttons.   

Many of the handheld DMMs I've looked at use buttons to select secondary functions.  One uses a relay to perform the actual switching.  You can change the state of the relay with a simple magnetic hanger.  This can cause the meter to display very low voltages when high voltages are present.   There are a few handheld DMMs I have seen using relays like this.   

[BeBuLamar]

It's good having a big mechanical thing on the front that always points to the range in use.

Given the added design overhead in coming up with a mechanical dial with decent feedback and robustness etc, along with the design of all of the PCB tracks

It doesn't seem to be a problem in practice.

Buttons have been tried lots of times. I don't think anybody found the right formula yet.

Fluke even makes a meter with just two positions on the dial (well, three if you count "off" as a position).

I have one of those. It's all I need most of the time. I don't really need to pull out my 87V or 289.

[BeBuLamar]

I have old 80's Casio calculators that are still good and most if not all modern DMM's also have buttons on the top row, so it is possible to have buttons perform functions that are expected to last the life of the device.

True the cheaper auto DMM's and expesive units with relays do exist that have all buttons, but I still think designing and manufacturing a rotary dial function over buttons is still "complex". I can't see any reason other than it's traditionally done this way, to not start offering all button devices now.

Although it's traditional for the analog VOM to have a dial it's not so for a DMM. In fact early Fluke DMM's only had buttons. But those buttons are not simpler nor cheaper than the dial they use now.

[HKJ]

but I still think designing and manufacturing a rotary dial function over buttons is still "complex".

Not really, the molds for the mechanic already exists, there is no need to design a new switch for each meter. If you look at the backside of a rotary switch you will often see extra slots for more sliders, i.e. the design can be used for more advanced switching without doing any changes to the molds.

Anyway a typical rotary switch do not need to switch much:
1) It need to signal the position to the DMM chip.
2) It need to short the uA shunt and move the sense connection to before it when in mA.
3) It need to switch the current generator onto the V/ohm/etc terminal when a range requires that.
For more advanced meters it may also need to switch a capacitor in on VAC.

Everything else can be done electronically and most of it is. I have seen 3) done electronically on a single cheap meter and 2) done electronically on some very expensive DMM's.

[Fungus]

I have old 80's Casio calculators that are still good and most if not all modern DMM's also have buttons on the top row, so it is possible to have buttons perform functions that are expected to last the life of the device.

Multimeters are often used in conditions that would kill a calculator in five minutes.

[HKJ]

Here is a meter with all buttons:



But there are two meters where the range switch is only a signal about selected range and do not switch any measurement voltages, with the two other meters the range switch will switch some measurement voltages.
In usage I prefer the rotary switch.

[Fungus]

I have one of those. It's all I need most of the time. I don't really need to pull out my 87V or 289.

The only problem I see with it is the input impedance - only about 3kOhms.

That image must be old. The 113 on the Fluke web site is CAT IV 600V.

[Fungus]

Here is a meter with all buttons:

Here's another one:

[spanakop]

I have old 80's Casio calculators that are still good and most if not all modern DMM's also have buttons on the top row, so it is possible to have buttons perform functions that are expected to last the life of the device.

Multimeters are often used in conditions that would kill a calculator in five minutes.

True, they wouldn't last very long, not even five seconds . But my point is that using these kinds of buttons can last a very long time. As I said, the top row of buttons are already on DMM's are there, why not expand it to the other functions.

If you're going to design a hand held instrument to do the functions of a DMM, why would you even start with a rotary dial. And as joeqsmith mentioned, I missed the fact that bench DMM's are almost all buttons.

[HKJ]

Here's another one:

The switch handles all the switching required, i.e. switching the current generator to the V/ohm terminal and notifying the chip about the position.

[HKJ]

If you're going to design a hand held instrument to do the functions of a DMM, why would you even start with a rotary dial. And as joeqsmith mentioned, I missed the fact that bench DMM's are almost all buttons.

Bench DMM's are filled with power hungry relays, you can't do that in a handheld DMM.

[BeBuLamar]

I have one of those. It's all I need most of the time. I don't really need to pull out my 87V or 289.

The only problem I see with it is the input impedance - only about 3kOhms.

That image must be old. The 113 on the Fluke web site is CAT IV 600V.

Most of the time the 3KOhms is OK. I wouldn't want to check 480VAC or 600+ VDC DC bus voltage on an inverter drive or servo drive though. Yes mine said CAT IV 600V. The 3K impedance is good for checking SSR output.

[HKJ]

Most of the time the 3KOhms is OK. I wouldn't want to check 480VAC or 600+ VDC DC bus voltage on an inverter drive or servo drive though. Yes mine said CAT IV 600V. The 3K impedance is good for checking SSR output.

The 3kOhm is only at low voltage, it will be much higher at mains voltage.
Here is a curve from a Fluke 289:

[BeBuLamar]

I don't know but when I check 240V and 480V with it it causes a spark both with the 113 and the 289 in low Z.

[HKJ]

I don't know but when I check 240V and 480V with it it causes a spark both with the 113 and the 289 in low Z.

When you connect the meter the resistance will be 3kOhm, but a fraction of a second later it will be much higher. It is a PTC where the resistance increases when it is heated. The steady state power will probably be around 1 Watt.

[Cymaphore]

My personal experience:

It depends on where the device is in use.

On a lab bench, it doesn't really matter so much.

In "the field" it's a whole different story. When crawling arround in a dark cabinet, accidentially pressing one of those many buttons or pressing the wrong one can be a real showstopper.

Rotary switches that can be single hand operated and that tell you about the selected range in a haptic way are superior in such an environment.

So it depends on what you're doing and how you're doing it. Probably other people have good reasons for using those all-button meters (for example Metrix).

[Cubdriver]

If you look at most modern meters, you'll see that the 'rotary switch' is mostly part of the main PCB.  It has a carrier with contact(s), and the switch detents are molded into the case itself, with a plastic spring detent mechanism.  Minimal parts count for a very functional switch that is both east to actuate and at the same time protected fairly well from accidental movement.  It physically switches connections, and they then stay connected.  The pushbuttons on the top row of such meters are momentary, and activate logic level functions - they don't/cant do switching like the purely mechanical rotary switch can - you'd need them to turn on/off internal circuits that could then do the switching, and likely with more loss than a purely mechanical contact that's just there, not to mention they'd draw power (not good in a handheld, battery powered device) while the mechanical switch's contacts are purely passive in that regard.

If you look at the pushbuttons on bench meters, they're either logic level and drive power consuming components, or are mechanical, in which case they are likely considerably more complex than the simple rotary switch currently used.  Look at the early Fluke handhelds that had the buttons along the left side - those are a mechanical nightmare compared to the low parts count rotary currently in use - they each consist of many components (individual contacts staked into a base, a slider, moving contacts in the slider, a spring to force the slider to its 'home' position, a mechanism to lock the slider in the desired location, for ganged switches some means to interconnect them to each other to disengage one when another is pushed.

Look at these from a Fluke 8200 bench meter, for instance:

Before even adding the keycaps and the stuff to hold them together and couple the mechanism, I can count AT LEAST 12 parts in EACH switch - from the front, the plunger, the spring retainer, the spring, the switch body, six fixed contacts and two moving ones.  There may well be more, too - the plunger may fit into an internal carrier for the moving contacts, and they may be spring loaded themselves.  It's a pretty complex little piece of hardware.  now add six more of them and interconnect them to work properly in unison - more fun.

In short, manufacturers largely stick with the rotary switch because it's simple and it works.

-Pat

[tooki]

If you're going to design a hand held instrument to do the functions of a DMM, why would you even start with a rotary dial. And as joeqsmith mentioned, I missed the fact that bench DMM's are almost all buttons.

Bench DMM's are filled with power hungry relays, you can't do that in a handheld DMM.

Many handheld DMMs use relays. For example, the Keysight U1252B.

[Black Phoenix]

If you look at most modern meters, you'll see that the 'rotary switch' is mostly part of the main PCB.  It has a carrier with contact(s), and the switch detents are molded into the case itself, with a plastic spring detent mechanism.  Minimal parts count for a very functional switch that is both east to actuate and at the same time protected fairly well from accidental movement.  It physically switches connections, and they then stay connected.  The pushbuttons on the top row of such meters are momentary, and activate logic level functions - they don't/cant do switching like the purely mechanical rotary switch can - you'd need them to turn on/off internal circuits that could then do the switching, and likely with more loss than a purely mechanical contact that's just there, not to mention they'd draw power (not good in a handheld, battery powered device) while the mechanical switch's contacts are purely passive in that regard.

If you look at the pushbuttons on bench meters, they're either logic level and drive power consuming components, or are mechanical, in which case they are likely considerably more complex than the simple rotary switch currently used.  Look at the early Fluke handhelds that had the buttons along the left side - those are a mechanical nightmare compared to the low parts count rotary currently in use - they each consist of many components (individual contacts staked into a base, a slider, moving contacts in the slider, a spring to force the slider to its 'home' position, a mechanism to lock the slider in the desired location, for ganged switches some means to interconnect them to each other to disengage one when another is pushed.

Look at these from a Fluke 8200 bench meter, for instance:

Before even adding the keycaps and the stuff to hold them together and couple the mechanism, I can count AT LEAST 12 parts in EACH switch - from the front, the plunger, the spring retainer, the spring, the switch body, six fixed contacts and two moving ones.  There may well be more, too - the plunger may fit into an internal carrier for the moving contacts, and they may be spring loaded themselves.  It's a pretty complex little piece of hardware.  now add six more of them and interconnect them to work properly in unison - more fun.

In short, manufacturers largely stick with the rotary switch because it's simple and it works.

-Pat

But you have to admit that a rotary switch may be simpler to implement and can be a joy to use when the parts don't have almost any play and are solid, but a good old school mechanical switch well maintained is "orgasm" for the fingers in terms of feeling.

It's the same as rotary encoders, simple and work but a genuine Alps is a different beast.