Keysight New instruments

[Bassman59]

I can appreciate the new kind of housing, it saves a lot of space on the workbench (depth).

At home I have a Tek TDS2024. (I bought it new, well before Rilent came out with 2 GHz sampling/200 MHz 'scopes for $500. I think it was really the only option back then!) Despite the limited triggering, low memory depth and small screen, it still works well enough. That it's not very deep is a bonus, I can fit it in tight spots.

At the day job I have a DPO3054. It's also not a very deep 'scope so it fits in packed benches. (Sometimes you work away from your usual bench.)

My HP1660E logic analyzer takes up much more bench space. Actually, it takes up zero bench space as it's been on the top of a bookshelf for years. I have no need for a logic analyzer any more.

Prices are better for hobby purposes but what when Keysight doesn't sell to private persons anymore...

Here in the Colonies I can buy whatever Keysight product I want, up to the limits of my credit cards!

[floobydust]

Deal-breaker on the EDU36311A PSU is the danger when charging a battery. It appears there is an OVP crowbar that trips if the PSU is connected to a battery and power is switched off. When power is switched off, the reference drops and the OVP wrongly thinks O/P voltage is too high, OVP activates and sadness results, as the user manual mentions.
So you must use a series diode to charge batteries. What a hassle and loss of precise charging voltage.
Some techs first switch off a PSU, then disconnect the leads. This is bad if there is a battery or load with huge capacitance connected, the backfeed can kill a pass-transistor and control circuitry.

[SilverSolder]

Deal-breaker on the EDU36311A PSU is the danger when charging a battery. It appears there is an OVP crowbar that trips if the PSU is connected to a battery and power is switched off. When power is switched off, the reference drops and the OVP wrongly thinks O/P voltage is too high, OVP activates and sadness results, as the user manual mentions.
So you must use a series diode to charge batteries. What a hassle and loss of precise charging voltage.
Some techs first switch off a PSU, then disconnect the leads. This is bad if there is a battery or load with huge capacitance connected, the backfeed can kill a pass-transistor and control circuitry.

Couldn't the pass transistor related control circuitry be designed to cope with this kind of real-world use case?  It does seem a bit lame in 2021...  this isn't Keysight's first power supply, there is a long and strong ancestry there.

[25 CPS]

Deal-breaker on the EDU36311A PSU is the danger when charging a battery. It appears there is an OVP crowbar that trips if the PSU is connected to a battery and power is switched off. When power is switched off, the reference drops and the OVP wrongly thinks O/P voltage is too high, OVP activates and sadness results, as the user manual mentions.
So you must use a series diode to charge batteries. What a hassle and loss of precise charging voltage.
Some techs first switch off a PSU, then disconnect the leads. This is bad if there is a battery or load with huge capacitance connected, the backfeed can kill a pass-transistor and control circuitry.

I took a look at the manual too and there's no provision for remote sensing either, so the idea of picking off the feedback voltage externally after the protection diode to regain precise voltage control wouldn't work either although if it did, it'd be even more setup overhead on top of the hassle of adding the diode.

[Kleinstein]

Most lab supplies have a protection diode to protect the output stage from residual charge in a reasonable sized capacitor. So turning the supply off should not directly be a problem. The over voltage protection / crow bar may engage than, but it should be able to absorb quite some energy.


The combination of relatively weak (1A) channels and a 5 A 5 V channel could be a problem: the manual says not to wire the 5 V and another channel in series or parallel (as they are different current rating). However for educational use this does not stop the users to do it. Let's hope the build in diode for reverse protection can withstand the 5 A.
It could be a bit problematic that all 3 sources in sereis give up to 65 volts - this could be a deal breaker in some areas.

[wizard69]

The estate version looks like it'd be doing a wheelie all the way home from a hamfest.

Nah, they have a *huge* engine in the front as a counterwieght for the valuables.

Still don't underestimate the weight of old dinosaur test gear and giant RF amplifiers with real iron transformers.

This brings back memories.   About 40 Years ago I had started work in industrial automation (actually still a young kid).   One of the old techs said he need my help moving an RF amplifier.   This shocked me because the guy was huge, a cross between a biker and a weight lifter, well over 6 feet tall.   That RF amplifier didn't look that heavy when I first saw it, but after a bit of a two man struggle we got it into position.   I was shocked that so much weight could be inside an electronics enclosure.

The other funny thing here was back when I was taking tech classes I asked about tubes.   The instructor said I didn't have to worry about that old tech.   Hardly two years later I'm moving an RF amp around with the biggest tube I'd ever seen up to that time.

In any event a little off topic, but it does highlight how technology has changed.

[wizard69]

I'm looking or really waiting for, affordable electric cars.   Well trucks or vans are actually my cup of tea.    Tesla with their pickup is getting there.   I actually like some the speculation floating about the net for a van covered in solar cells.   The idea of retirement is not far away and a van to explore the world has some appeal.

As for the Aston Martin, do people really by those ugly cars?

We know nothing of the price point for these new Keysight instruments? The playing field has a lot of competition now.

Go a few pages back, prices and user manuals posted.

Sorry I was distracted at the wheel. To recap the list prices of these new instruments:
EDU33212A Dual Channel Function Generator; USD $915
EDU36311A DC Power Supply; USD $838
EDU34450A Digital Multimeter; USD $696
EDUX1052G Dual Channel Oscilloscope; USD $722

New Aston Martin Lagonda planned production this year...
The old one nice in green, pop up headlights and all digital dash, flat switches. I actually like it when designers can go futuristic - have fun, create/innovate, think outside the corporate box. Steve Jobs did it as well.
Yes it has some non-chicken dinner aspects but these ventures nonetheless are a step forward.

[Berni]

Deal-breaker on the EDU36311A PSU is the danger when charging a battery. It appears there is an OVP crowbar that trips if the PSU is connected to a battery and power is switched off. When power is switched off, the reference drops and the OVP wrongly thinks O/P voltage is too high, OVP activates and sadness results, as the user manual mentions.
So you must use a series diode to charge batteries. What a hassle and loss of precise charging voltage.
Some techs first switch off a PSU, then disconnect the leads. This is bad if there is a battery or load with huge capacitance connected, the backfeed can kill a pass-transistor and control circuitry.

That is a pretty silly design decision in that case. I charge batteries with lab PSUs all the time and none of them gave a toss about having voltage fed back into them even when disabled. Tho connecting a battery across a lab PSU in reverse will do damage, but that case is just user stupidity and a PSU should not be expected to cope with.

But perhaps this works more like the current sinking capable rack mount PSUs that HP/Agilent/Keysight makes. That functionality is a bit of a trap for young players because it is not clearly started on the list of features and its rare for a PSU to do this. Most PSUs will let you drive its output above the set voltage by hooking up a higher voltage source to the output, but some of these rack mount PSU models actually start to fight you back and actively sink current to bring it back down to whatever is set. For something like a Agilent 66332A (Dynamic Measurement DC Source) that is designed for 2 quadrant operation this sinking current is the same as the current limit set on the front panel (this makes it a nice electronic load or battery simulator) while some of the other PSUs ignore the current limit and always use a sinking current that is equal to the max rated current for that channel. So having a battery connected would just drain the battery some while the PSU is unharmed. But even then i think this current sinking only happens when a channel is actually enabled and set to a lower voltage, i think the terminals just go open circuit if the channel is disabled.

This feature is likely present in the rack mount system PSUs to avoid blowing up a DUT on a production line test jig if a 12V rail shorts to a 3.3V rail or something. So perhaps the people designing this EDU power supply seen that feature and thought this was a neat idea to also help students blow up less cirucits. Yet have not looked closely enough at how it works and implemented it wrong to also sink current in disabled mode or forget to limit the sinking current to protect itself.

Dave should test it out what happens when you backfeed current into this thing from another PSU

[HighVoltage]

Deal-breaker on the EDU36311A PSU is the danger when charging a battery. It appears there is an OVP crowbar that trips if the PSU is connected to a battery and power is switched off. When power is switched off, the reference drops and the OVP wrongly thinks O/P voltage is too high, OVP activates and sadness results, as the user manual mentions.
So you must use a series diode to charge batteries. What a hassle and loss of precise charging voltage.
Some techs first switch off a PSU, then disconnect the leads. This is bad if there is a battery or load with huge capacitance connected, the backfeed can kill a pass-transistor and control circuitry.

Many old Agilent and HP PSUs had this "feature" and many years ago I killed a brand new Agilent PSU, by charging a battery and the crow bar was activated.
The magic smoke appeared really fast and dark.
Since that time I am careful, every time I hook up a battery or large capacitor to a PSU.

[nctnico]

Deal-breaker on the EDU36311A PSU is the danger when charging a battery. It appears there is an OVP crowbar that trips if the PSU is connected to a battery and power is switched off. When power is switched off, the reference drops and the OVP wrongly thinks O/P voltage is too high, OVP activates and sadness results, as the user manual mentions.
So you must use a series diode to charge batteries. What a hassle and loss of precise charging voltage.
Some techs first switch off a PSU, then disconnect the leads. This is bad if there is a battery or load with huge capacitance connected, the backfeed can kill a pass-transistor and control circuitry.

Many old Agilent and HP PSUs had this "feature" and many years ago I killed a brand new Agilent PSU, by charging a battery and the crow bar was activated.
The magic smoke appeared really fast and dark.
Since that time I am careful, every time I hook up a battery or large capacitor to a PSU.

Many years ago I bought a nice HP PSU cheap because the owner had used it to charge a battery. The owner had slowed the fan down so the PSU overheated which also activated the crowbar. In the end the damage was limited to a few burned PCB traces so it was an easy fix. While at it I made the fan temperature controlled.

[Fungus]

The old one nice in green, pop up headlights and all digital dash, flat switches. I actually like it when designers can go futuristic - have fun, create/innovate, think outside the corporate box. Steve Jobs did it as well.

Yep. Gotta love that interior. It's pure James Bond.

The outside looks like a very long Fiat X1/9:


(The Fiat was a few years earlier...)

[Berni]

Many old Agilent and HP PSUs had this "feature" and many years ago I killed a brand new Agilent PSU, by charging a battery and the crow bar was activated.
The magic smoke appeared really fast and dark.
Since that time I am careful, every time I hook up a battery or large capacitor to a PSU.

Ouch so they indeed have a full on crowbar circuit that throws a dead short across the output once upset.

I will definitely be more careful with PSUs from now on then.

[nfmax]

Many old Agilent and HP PSUs had this "feature" and many years ago I killed a brand new Agilent PSU, by charging a battery and the crow bar was activated.
The magic smoke appeared really fast and dark.
Since that time I am careful, every time I hook up a battery or large capacitor to a PSU.

Ouch so they indeed have a full on crowbar circuit that throws a dead short across the output once upset.

I will definitely be more careful with PSUs from now on then.

Most of the time (i.e. when not trying to charge a battery) such a crowbar is exactly what is needed!

[SilverSolder]

Yes, a crowbar is a good thing (see Dave's video about how unpredictable fuses behave.  Crowbars make them predictable!).

The problem is, why not make the output circuits immune to external voltages?  Difference between men and boys! 

[floobydust]

I think a decent PSU doesn't need OVP, it's supposed to be reliable lol. Put the money into a beefier pass-transistor and heatsink or backfeed-diode.
But here we have firmware and memory leaks which the OVP might need to cover.

I can't easily tell if the new PSU can source/sink (push-pull) output. It's likely just a source-only.
Is it an SCR crowbar? Must be.
It's actually a bomb because charging a battery and power glitching or going out, the OVP trips and the battery then cooks the PSU with no fuses etc. on the O/P side. Easily start a fire in a facility.

[bd139]

Disagree entirely. Crowbar is important when the following equation is true:

cost_of_load > cost_of_power_supply

A year or so ago I had what i consider to be a fairly decent quality supply (HP) blow the reference zener. This caused the output to bang itself at the collector supply voltage. You should plan for such failures.

In such circumstances it doesn’t matter if the supply blows up. That’s what a warranty is for.

[Fungus]

Moral: Use batteries with protection circuits on them.

(...and (b) car battery chargers cost about $20)

[bd139]

Or just stuff a diode in series. 

Farnell E series supplies used to blow up if the load drove the supply. This happened when you were charging an SLA off one and the mains went out. Blew the driver transistor and the pass transistor.

[bdunham7]

Moral: Use batteries with protection circuits on them.

(...and (b) car battery chargers cost about $20)

I can't agree when we are talking about general purpose lab PSU's, especially those that might be used in an educational setting.  The one I have can be connected to a battery with the power on or off and in forward or reverse polarity without damage.  If I hook up the battery backwards, turn the PSU on and turn up the voltage and current to max, it will simply discharge the battery at its rated current.  No smoke and no fans.  Of course it weighs and costs more than a Kaiweets model and doesn't put out as many amps.  A proper, robust lab PSU shouldn't blow up, period.  For more protection of an expensive load device, a separately adjustable crowbar with adequate fusing before and after it would be OK.

[Kleinstein]

Some supplies have a fuse for the crow bar. So the fuse hopefully blows before the rest of the circuit gives smoke signs.  As the crow bar is often a really beefy SCR this may actually work, as SCRs are relatively robust.

Quite a few supplies also have a kind of weak sink capability - so if the voltage is higher than it is supposed to be, they would first sink a moderate current of a few 10 mA, usually with not much harm. This way the ouput capacitor would drain if you turn down the voltage with no load.
Many designs have a diode, so that if backpowered, the source from the input would just provide power to the supply circuit. Unless to high nothing bad happens - draining the battry meant to charge.

A reversed battery is usually problem, because there usually is a diode to prevent significant reverse voltage. This is especially needed with 2 supplies for a +- supply and than the current lmit engaging on one side. Usually the same fuse could engage as with the crow bar (if present). For the KS supply the point may be that the 1 A channels shoud have reverse diode good for at least 6 A (the other 2 channels combined)  - not every 1 A supply has this.

[EEVblog]

Or just stuff a diode in series. 

That's what the manual tells you to do.
Unfortunately no external sense to compensate for the drop.

[EEVblog]

I think a decent PSU doesn't need OVP, it's supposed to be reliable lol. Put the money into a beefier pass-transistor and heatsink or backfeed-diode.
But here we have firmware and memory leaks which the OVP might need to cover.
I can't easily tell if the new PSU can source/sink (push-pull) output. It's likely just a source-only.
Is it an SCR crowbar? Must be.

The manual says is "program the output to zero", that implies software control of theo utput, not a real hardware crowbar.

[nctnico]

I think a decent PSU doesn't need OVP, it's supposed to be reliable lol. Put the money into a beefier pass-transistor and heatsink or backfeed-diode.
But here we have firmware and memory leaks which the OVP might need to cover.
I can't easily tell if the new PSU can source/sink (push-pull) output. It's likely just a source-only.
Is it an SCR crowbar? Must be.

The manual says is "program the output to zero", that implies software control of theo utput, not a real hardware crowbar.

No. Even on analog controlled PSUs (the HP 6012A for example) HPAK calles a current sink circuit a 'down programmer'. This circuit is intended to discharge the output capacitors in order to ramp the output voltage down in a controlled manner. Kind of a 2 quadrant operation.

If you look at the manual for the HP 6012A you'll notice that HP calls applying analog voltages to set the output voltage / current 'programming'. There is no software involved; it is all analog.

[bdunham7]

The manual says is "program the output to zero", that implies software control of theo utput, not a real hardware crowbar.

What is "the overvoltage condition"?  Is it the output exceeding the set point by a certain amount or is it exceeding a fixed limit above the maximum output of the PSU? 

If the former, then it must be at least partly software controlled, if not it could be a standalone circuit that triggers it.  I assume there is no separate OVP control.

If 'program the output to zero' means the same thing as shorting it out (which is what they are warning about) then that implies that there is either a crowbar or that the main PSU outputs can sink current.

I'd sure like to know what the explanation would be for setting up a lab PSU in this manner. 

[SilverSolder]

I just looked at the manual for a supply designed by our ancestors (6200B) and it claims that back-powering it leads to "loss of regulation and possible damage to the output capacitor", presumably if you back-power it at a voltage that exceeds the rating of the capacitor...

So, a battery would just cause "loss of regulation", presumably.   I guess our ancestors weren't so dumb?