Keysight EDU36311A (rant)

[Someone]

A better approach for a company is to make sure the age across all departments averages around 35 to 40 years with an even distribution.

Good Luck With That.  The only way to do that is to force early retirement onto the older engineers, who tend to be the most capable ones you have, or to get rid of your middle-age engineers who are just starting to become as capable as the older engineers.

End result: you compromise the capability of your organization, make it worse than it would be otherwise.

Unless your organization is large enough that you have a steady stream of people who are retiring, and can bring in young people as replacements, and have an even enough distribution of age and capability both within groups and across groups that your remaining older people can step into the roles that were left by the retirees, you're going to have a really tough time pulling anything like this off.

Engineering is one of those disciplines where youthful exuberance does not make up for hard-won experience.  Yes, someone who is relatively new and fresh can occasionally get a flash of brilliance and approach things in a new and insightful way that nobody else thought of before, but that's rare and completely unpredictable, and often comes with its own set of tradeoffs (for instance, making rookie mistakes in the implementation of the new idea).  Ideally you should have a mentorship arrangement where younger engineers are paired with older engineers so that the younger engineers can continually learn from the experience of the older ones.   I don't know how often that happens in practice in hardware-oriented organizations.

There are some very big companies who do manage this, but the scale required to have enough people in any given field (preferably in the same location/site and working on the same products) makes it unusual. One "trick", give the oldest employees part time contracts to taper off their working hours but retain their knowledge and mentoring (while ruthlessly firing anyone along the way who won't make it to such high levels). Not the most comfortable environment to work in when you have to keep showing growth/excellence ahead of your peers, similar problems in finance/accounting practices.

[porter]

One bit of work philosophy I was taught at HP->Agilent was that everyone has five minutes for everyone else. Or something like that.  I guess that way information isn't siloed, and the newer folks won't waste a lot of time...etc.

[Berni]

Indeed training up a good engineer is hard.

You could have two engineering doctorates with top marks, yet are more likely going to be a shitty engineer if you don't have years of experience on real world projects.

It takes time to absorb all the little details and good engineering practices. When starting out on your own you might have to learn a lot of these the hard way when your project blows up or performs poorly. But if you are working next to someone experienced you can siphon off those tricks a lot faster. Like asking the senior engineer to look at your schematic, watch him point out why some parts of it won't work or a bad idea, saving you from building a non functional board just to find out it is a bad design.

Id say this transfer of knowledge is even rarer for software development. The senior guys might help with setting up the dev environment and fixing common problems, but i don't think they really go into the art of programing all that much.  Sure they apply the algorithms they learned in school and all that, but that is the easy part. The more difficult part is architecting the higher level structure. You can really shoot yourself in the foot by just jumping straight in without a general plan of how to split up the functionality. You get spaghetti code that is all over the place, confusing to read, contains confusing names, difficult to maintain, gives more bug opportunities...etc

[Keysight DanielBogdanoff]

I've sent this thread over to the power supply team.

The one thing I've experienced with their design philosophy for supplies is that the whole goal is to not accidentally put in too much power. That's why the knob acceleration is so slow, for example. The general recommended practice is that you shouldn't adjust power while the output is on. But, I do it with live outputs as much as anyone.

[mnementh]

Daniel! We didn't chase you off with all the pushback over the whole "Commercial/Industrial Customers Only" thing.

Hope we can resolve this quickly, get you a win. 

mnem
 

[2N3055]

I've sent this thread over to the power supply team.

 The general recommended practice is that you shouldn't adjust power while the output is on. But, I do it with live outputs as much as anyone.

Good to hear from you!

I assure you that 99% of people do change output voltage and current while output is enabled...
These are lab power supplies and they are meant to be used for experimenting.

[floobydust]

I checked the UI on Ruideng DPS5005 (the RD6006 would be worth looking at too) and you can edit the setpoint digit by digit with the output enabled - but it won't let you change the 10's of volts, only the 1's of volts which can rollover and bump up/down the 10's. You just can't cursor over to the 10's digit. That limits it to 1V jumps max., might be the simplest fix.
Or, you are allowed to change the PSU's setpoint when its output is on, but must confirm/press enter to accept it after editing/changing it.

[alm]

I checked the UI on Ruideng DPS5005 (the RD6006 would be worth looking at too) and you can edit the setpoint digit by digit with the output enabled - but it won't let you change the 10's of volts, only the 1's of volts which can rollover and bump up/down the 10's. You just can't cursor over to the 10's digit. That limits it to 1V jumps max., might be the simplest fix.
Or, you are allowed to change the PSU's setpoint when its output is on, but must confirm/press enter to accept it after editing/changing it.

I think you described the UI of the Atten power supply Dave hated so much. I would hate both changes. I should be able to change the power supply by 10 volts if I want to, but it should not jump to larger increments when switching ranges. And I definitely should be able to change the supply voltage live, for example when watching when a linear regulator will drop out, without having to press 'enter' all the time.

[mnementh]

I concur. Sooner or later, you will want some mode which mimics the organic flow of analog as much as possible, in which my preferred arrangement is coarse/fine adjustment knobs on both CV and CC.

Why is it so hard to make this happen on digital PSUs? Why is it always either pushbuttons or a single encoder, and neither of which with an ideal UI but rather something which fails miserably due to trying trying to make it serve as both? The kind of money these tools cost, a numeric keypad and a few additional encoders is utterly trivial.

mnem

[floobydust]

EDU36311A already has a keypad, two rotary encoders, arrow keys - how many more methods of UI do we want here?  Wait until it gets a touch screen   
I think the issue is any of those methods making an unexpected, dangerous leap by an order of magnitude i.e. 10x is something unwelcome.

[TurboTom]

I concur. Sooner or later, you will want some mode which mimics the organic flow of analog as much as possible, in which my preferred arrangement is coarse/fine adjustment knobs on both CV and CC.
...
mnem

For what purpose? Coarse and fine regulators are a makeshift solution on analog PSUs when decent multiple-turn pots didn't fit the BOM for economical reasons. The four knobs are only confusing things and are prone to cause errors when having (had) to manually control several parameters at a time. Any properly implemented incremental encoder with a digit selector provides several such coarse / fine gradations. I agree that in some cases separate encoders for voltage and current can be useful. But a decimal keypad is even more important on a digitally controlled PSU.

All the manual adjustments became less important with the advent of computer control of instruments. But in some situations it's still convenient to be able to "turn a knob" to change a value, and this ususally happens (in case of a PSU) with the output active. Hence, "automatically" changing the digit to be adjusted is more than a no-go, as many have pointed out already!

[capt bullshot]

I concur. Sooner or later, you will want some mode which mimics the organic flow of analog as much as possible, in which my preferred arrangement is coarse/fine adjustment knobs on both CV and CC.
...
mnem

For what purpose? Coarse and fine regulators are a makeshift solution on analog PSUs when decent multiple-turn pots didn't fit the BOM for economical reasons. The four knobs are only confusing things and are prone to cause errors when having (had) to manually control several parameters at a time.

+1 on proper multiturn pots. Fine / Coarse is cheap crap.
Though IMO encoder with selectable digit is the best digital way, and a keypad should be available too, nothing beats the experience and ease of use of a proper multiturn pot.
I've got various lab PSU, here's my personal order of entry method preference:

1. multiturn pot for voltage and current
2. multiturn pot for voltage, standard pot for current
3. select parameter by single keypress, enter numeric value through keypad, press Enter
4. select parameter by menu or cursor keys, enter numeric value through keypad, press Enter
5. separate digital encoders for voltage / current, with single keypress selectable and consistently acting digit
6. anything else

7. Coarse / Fine knobs

And another one: Order of output terminals

Minus bottom / Plus above
or
Minus left / Plus right
or
Minus left / GND middle / Plus right


everything else is anti-intuitive and annoying.

[HighVoltage]

I think the issue is any of those methods making an unexpected, dangerous leap by an order of magnitude i.e. 10x is something unwelcome.

That is not acceptable in any PSU.
The good thing is, that this can be solved in Software and a new FW update will probably fix this.

Many years ago, the E36300 Series PSU had many more problems and the hardware had to be fixed and instruments were exchanged. At the end it was a very good PSU.

The surprise for me is the fact that a company like Keysight would not detect these problems internally before they release such a PSU. Its easy to fix before it is released and a nightmare afterwards.

[nctnico]

I concur. Sooner or later, you will want some mode which mimics the organic flow of analog as much as possible, in which my preferred arrangement is coarse/fine adjustment knobs on both CV and CC.
...
mnem

For what purpose? Coarse and fine regulators are a makeshift solution on analog PSUs when decent multiple-turn pots didn't fit the BOM for economical reasons. The four knobs are only confusing things and are prone to cause errors when having (had) to manually control several parameters at a time. Any properly implemented incremental encoder with a digit selector provides several such coarse / fine gradations. I agree that in some cases separate encoders for voltage and current can be useful. But a decimal keypad is even more important on a digitally controlled PSU.

It would be nice though when the knobs on a digitally controlled PSU would emulate a 10 turn pot. Best of both worlds. It would take a very fine pitched encoder though.

[Sighound36]

I shall be watching this thread closely as we have just taken delivery of four of these units for intern's 
Thanks for Capt Bullshot for finding this issue out excellnt work and glad the DUT didn't take a one way trip to Mars

Hope the Danny Bodegnaoff show can swing into to action and re address this issue in short time spell all our love from the MXR appreciation society first tester guild of the UK

[Berni]

Multiturn pots can be annoying too. Going from say 3.3V to 12V involves a whole lot of knob twisting.

For analong supplies i actually don't mind having a corse and fine adjust control. Just leave the fine control in the center most of the time so that you can turn it up or down when you need to fine tune it. (At least when the PSU doesn't do what this Keysight EDU does)

I do prefer modern digital PSUs. They are quick to set and i can have a encoder knob that is exactly as fine as i need it. Can step in volts when i need to quickly get somewhere, or i can step it in milivolts when fine tuning something. Having a keypad is a nice timesaver bonus too since most of the time you need nice round number voltages.

[mnementh]

I concur. Sooner or later, you will want some mode which mimics the organic flow of analog as much as possible, in which my preferred arrangement is coarse/fine adjustment knobs on both CV and CC.
...
mnem

For what purpose? Coarse and fine regulators are a makeshift solution on analog PSUs when decent multiple-turn pots didn't fit the BOM for economical reasons. The four knobs are only confusing things and are prone to cause errors when having (had) to manually control several parameters at a time. Any properly implemented incremental encoder with a digit selector provides several such coarse / fine gradations. I agree that in some cases separate encoders for voltage and current can be useful. But a decimal keypad is even more important on a digitally controlled PSU.

All the manual adjustments became less important with the advent of computer control of instruments. But in some situations it's still convenient to be able to "turn a knob" to change a value, and this ususally happens (in case of a PSU) with the output active. Hence, "automatically" changing the digit to be adjusted is more than a no-go, as many have pointed out already!

I disagree with this assessment. Multiturn pots are not assache-free in every situation; they are just a different flavor of assache in certain other situations. 

I said this was my preferred arrangement. I like a multiturn pot for the fine adjustment; I just hate having to turn a pot a million times to get from say 5V to 35V. We're talking about what we want; everybody wants different things. I also do not want to give up the ability to direct-enter via a keypad. As I said earlier; at the price of these tools, adding a few encoders along with a keypad is utterly trivial.

Even dual encoders to emulate a single multiturn pot on each of CV and CC would still be a vast improvement over only having single-digit incrementation with only pushbuttons. 

Maybe make the encoders so that when you press and turn, you can select a level of granularity, which is clearly shown on the display? Like being able to select from 10V, 1V, .1V, .01V, .001V etc per increment? Then maybe, just for safety's sake, another push to confirm that choice, and nothing actually changes until you start incrementing AFTER pushing the button?

mnem
 

[mnementh]

Multiturn pots can be annoying too. Going from say 3.3V to 12V involves a whole lot of knob twisting.

For analong supplies i actually don't mind having a corse and fine adjust control. Just leave the fine control in the center most of the time so that you can turn it up or down when you need to fine tune it. (At least when the PSU doesn't do what this Keysight EDU does)

I do prefer modern digital PSUs. They are quick to set and i can have a encoder knob that is exactly as fine as i need it. Can step in volts when i need to quickly get somewhere, or i can step it in milivolts when fine tuning something. Having a keypad is a nice timesaver bonus too since most of the time you need nice round number voltages.

Okay... which models of digital PSU are you talking about? Having both, and having a usable UI, is precisely what I'm talking about. This is hardly a universal thing... or even a common thing... with digital PSUs. 

mnem

[porter]

I think a foot pedal is needed. Maybe more than one. 

[mnementh]

I need one. For my battery-pack spot-welder.

mnem
*tzzzt*

[2N3055]

I think a foot pedal is needed. Maybe more than one. 

Maybe a MIDI so you could use digital drums or any other MIDI controller.....

Or maybe support for full HOTAS ... 

[Berni]

Okay... which models of digital PSU are you talking about? Having both, and having a usable UI, is precisely what I'm talking about. This is hardly a universal thing... or even a common thing... with digital PSUs. 

mnem

Yeah some digital PSUs get it wrong.

I use the Rigol DP832 the most and would recommend it. It is a good feature rich PSU for a good price. Sure the circular number pad is kinda weird but you get used to it and it does save a lot of front panel space. Some of the advanced menus are a bit confusing but the basic UI of setting voltages and currents and limits and all is nicely done.

Sometimes even cheep ones get it right like the MulticompPro/Tenma/FarnellStoreBrand model no. 72-2690. It looks like your classical cheep two knob analog supply, but it has encoders instead of pots, you can move the active digit up and down and adjust it. It remebers the setting on power down. Simple yet effective

Some of the larger Agilent made supplies (Like the rack mount ones) offten have both a keypad and encoder for setting them. They work well tho the UI takes a bit of getting used to.

But there are certainly plenty of poorly executed ones. The Korad (the one dave blew up, i also blew up a similar one) is just plain anoying. The cheep switchmode bench supplies from BK Precision have a rather confusing interface. The TDK Lambda rack mount supply i got does have a simple and clear UI but it is encoder knob only with no velocity control (So you end up with the multiturn pot issue of having to turn and turn and turn and turn and turn to actually get to the voltage you want)

But it is not only limited to PSUs. There are some instruments that i just find confusing to use for some reason. Like the menus of the Keithley 2100 DMM or the classical BK Precision electronic loads.

[Thomas]

TTi MX series is the best I have used.
Setting 5V on a channel is 3 key presses:
<Vset softbutton> <5> <OK/Enter>

They are, however, not perfect.
The rotary encoder is accelerated (in a good-ish way), but I would have preferred left-right buttons below it to select the digit to adjust.

[Berni]

That is a very well designed front panel.

Controls very clear and spread out, minimized button presses for setting things etc. Also like the small sense wire connections (using bananas for that ends up way too bulky really quick on a multichannel supply.

The one downside i see is no mili button. The Rigol does the setting in reverse where if you want 3A you press [3] then [A]. But if you want 10mA you can press [1] [0 ] [mA]. Admitedly it is pretty rare, i mostly use it when setting the current under 1A for the initial power up of a prototype board.

[Thomas]

That sounds nice. Many instruments do it this way, entering numbers with softkeys for units. I like it.

On the TTi, 10mA is 5 key presses:
<Iset> <.> <0> <1> <OK/Enter>
Not too bad IMO, and it is intuitive.