New Agilent scopes

[mikeselectricstuff]

Why would they change it? The scope works, nobody's going to avoid buying it because of its standby power draw.

[Rufus]

Why would they change it? The scope works, nobody's going to avoid buying it because of its standby power draw.

Not a valid argument. Buying decisions are based on the sum of many features and this feature (if the buyer knows about it) would have some small bearing on the decision (especially in this age of eco green saving the planet bollocks). If Agilent offered an otherwise identical model at the same price which didn't stay warm and pull a few watts when turned off which would you buy?

I the vblog Dave praised it for having what he thought was a real mains on/off switch. I remember it and thought it was a feature I was buying.

[_Sin]

I'd buy whichever was cheaper, and if it was the high standby power version, I'd stick a postit note on it to remind me to turn off the mains.

[mikeselectricstuff]

Why would they change it? The scope works, nobody's going to avoid buying it because of its standby power draw.

Not a valid argument. Buying decisions are based on the sum of many features and this feature (if the buyer knows about it) would have some small bearing on the decision (especially in this age of eco green saving the planet bollocks).

In this case it is entirely valid, as for something like a scope, standby power is so far down the list of possible purchasing criteria as to be negligible.
I highly doubt anyone would consider it unless they work in a company with some dumb bullshit policy on all purchases having to comply with low-energy standards, and even then, being engineers they would probably find a way around it.

If Agilent offered an otherwise identical model at the same price which didn't stay warm and pull a few watts when turned off which would you buy?

Irrelevant as that wouldn't happen. They would either revise it at some point or not.
It would cost them money to change it, so unless it brought enough benefit to increase sales to cover it, or legislation chnaged to require it, it wouldn't happen.

[hacklordsniper]

Why would they change it? The scope works, nobody's going to avoid buying it because of its standby power draw.

I was not thinking about the standby power draw, even if im on solar power most of the time i dont care about it. I was PM-ing with one owner of this scope and his scope died few days ago and now is returned to Agilent. I'm thinking maybe its better waiting for few hardware revisions to happen? Basically it died similar death as Daves scope but he did not have any LAN/VGA module installed...

[EEVblog]

Photo of the board, sorry don't have the bottom side.
The switch didn't look too wimpy, and it was on the mains side of the board, and of course they went to the trouble to do the traditional mechanical rod interface, so when I did the review I just assumed it was a proper mains switch!
Didn't occur to me to actually test that or look further at all 

Dave.

[Rufus]

The switch didn't look too wimpy

Looks wimpy to me for something that has to take the PSU inrush current. I doubt that switch was originally used for mains, there is space for something bigger.

Mains switching is a real PIA especially for big switchers and big transformers, but, you would expect someone the size of Agilent would have more than enough experience.

[mikeselectricstuff]

The switch didn't look too wimpy

Looks wimpy to me for something that has to take the PSU inrush current. I doubt that switch was originally used for mains, there is space for something bigger.

Mains switching is a real PIA especially for big switchers and big transformers, but, you would expect someone the size of Agilent would have more than enough experience.

yeah - inrush current, especially if not actively managed, can be a bitch - can easily 20-30A for a cycle or two, which can be hard on switches.
This arrangement, although less efficient is probably better for long-term reliability, assuming the switch mechanics are suitably robust.

[Dad-O-Rama]

A reviewer (at EE Times, I think) mentioned they had a pre-production unit that took over 20 seconds at power up before any activity was noted. He was told production units would have much shorter boot times.

Perhaps the original design did switch the mains, but to reduce boot time they keep the unit running in some idle mode? Like tube based TV's back in the day...

[Leo Bodnar]

Here is some performance data from sub-nanosecond pulse generator attached to my MSO-X 3024A.
Slew rate seems to be in the order of 4.7V/nsec
Cheers
Leo

[Rufus]

Perhaps the original design did switch the mains, but to reduce boot time they keep the unit running in some idle mode?

Nope. There is a single mains to 13vdc OEM PSU. All other supplies are generated from the 13v by switchers on the filter PCB. The power switch enables the 13v output and this PSU alone is responsible for the 'off' power consumption and heat.

Photo is another PCB mounted, plastic rod actuated mains switch in an Instek signal generator and all that is switching is mains to a 20VA transformer.

[Leo Bodnar]

Nope. There is a single mains to 13vdc OEM PSU. All other supplies are generated from the 13v by switchers on the filter PCB. The power switch enables the 13v output and this PSU alone is responsible for the 'off' power consumption and heat.

Is it a meant to be car battery powered when necessary?

[Rufus]

Is it a meant to be car battery powered when necessary?

Didn't think of that, but, a single mains generated supply and a bunch of derived ones does support operation from a battery pack.

I think the 13v supply is also used inside the scope but maybe it has a wide tolerance.

[mikeselectricstuff]

The 6000 does have a factory-fit battery pack option - I assume it fit in the probe container - seems sensible for them to carry on offering this option as this range replaces the 6000. I'd guess they replace the mains PSU with a DC/DC and battery management board.
Of course standby draw could be improved with a different PSU - with standards for this in various stages of coming in at the moment, it's entiirely possible they already plan to change the PSU as & when lower standby PSUs become readily avaiable at the right price.

[EEVblog]

Here is some performance data from sub-nanosecond pulse generator attached to my MSO-X 3024A.
Slew rate seems to be in the order of 4.7V/nsec

Nice board. Got any build details for that?

Dave.

[Leo Bodnar]

Nice board. Got any build details for that?
Dave.

Hi Dave,

The board is from Janne Ahonen http://koti.mbnet.fi/jahonen/Electronics/AvalanchePulser/
I have started laying out my own but then got one blank PCB from him and modified it slightly.  It is really simple, most of the parts belong to the 90V step up power supply.
Avalanche starts at 60V so maybe DC-DC voltage can be dropped to 70V or so.  As soon as you somehow get 70-90V DC then the pulser itself is just 5-6 components.
The pulse rise time is in the range of 300-600 ps.

Here are the links related to the pulser design:

http://cds.linear.com/docs/Application%20Note/an72f.pdf (Appendix B / p.40) basic diagram for self-oscillating mode
http://cds.linear.com/docs/Application%20Note/an94f.pdf externally triggered, controlled pulse width and damping
http://www.holmea.demon.co.uk/Avalanche/Avalanche.htm quick practical implementation
http://webpages.charter.net/dawill/tmoranwms/Elec_Pulse.html ditto
http://www.i9t.net/fast-pulse/fast-pulse.html
https://www.eevblog.com/forum/index.php?topic=553.msg21587#msg21587 local post
https://www.eevblog.com/forum/index.php?topic=588.msg10635#msg10635 local post

Cheers
Leo

[Bored@Work]

The pulse generator was also published in AN47 (Page 93). The whole app note from Jim Williams is a classic and worth to be read a few times http://www.linear.com/docs/4138

I build one a few years ago in a kind of split design similar to AN47. Mine has the 90 V power supply together with battery and on/off switch in a small plastic enclosure, and the transistor with BNC plug and the B, E, C resistors and capacitor in a separate tiny tin enclosure. Between them 1 m of cable to connect the power. This gives me a small separate head, not much larger than the BNC plug.

The power supply is build on a strip board, and - sorry Linear - I substituted the LT1073 with another boost converter IC I just happened to have in the drawer.

The "head" is build similar to what Jim Williams shows in AN47, although I didn't use a second BNC connector to connect the power. To keep the head small the cable is fixed attached to the head in my version.

If you have the parts, this is a weekend project. Just be careful, if you mess up the power supply you can easily get more then the desired 90 V out of it, and 90 V is already beyond ELV, so officially in the "it can kill you" range.

[EEVblog]

Thanks.
I thought it might have been the Jim Williams design.

Dave.

[tnt]

@Dave: So what about those hi-res photo of the boards ? :p

[Leo Bodnar]

Here is a screen capture of MSO-X 3024A that has direct short (few ns) pulse firing into Ch1 that is also sampled by Ch3 with standard 300MHz Agilent probe N2863B and Agilent digital logic analyser weaved cable pod.

Digital threshold has been set at 200mV as values over 500mV make digital pulse disappear.

You can see that the time delay between the probe and the digital pod is well matched - to within a 1ns which is the maximum resolution of digital sampling anyway.

[tnt]

Very nice, I can't wait to get my own pulse generator.

Why do threshold > 500 mV make the pulse disappear ? Any theory ?
Even 1V should still show a at least 2 sample wide (2ns) digital pulse.

[Leo Bodnar]

Why do threshold > 500 mV make the pulse disappear ? Any theory ?
Even 1V should still show a at least 2 sample wide (2ns) digital pulse.

I think it's on the edge of its capabilities. Agilent specifies minimum detectable pulse width of 5 nanosec. (corrected)

The test pulse is very short and I suspect that due to lower bandwidth and slew rate in the digital system pulse Vpp is much lower then 3.5V showing on analogue channels.

Code: [Select]

Maximum sample rate                1 GSa/s
Maximum record length              2 Mpts (digital only), 500 kpts (analog and digital channels)
Minimum detectable pulse width     5 ns
Channel-to-channel skew            2 ns (typical); 3 ns (maximum)

[tekfan]

Thanks.
I thought it might have been the Jim Williams design.

Dave.

Haha. I got mine too. Yeah, Jim Williams is just cool like that.

[tnt]

Argh ... the 3000-X refuses the 1152A active probe ...

It's not listed as "supported" probe (it's an obsolete model), but the Agilent rep told me it'd work ... now I have two useless probes. Damnit.

I was hoping to watch more scope pr0n tonight, not gonna happen now.

[Leo Bodnar]

I am very interested to learn the difference between the 100/200MHz and 350/500MHz mainboards.

Is it a radical redesign of the front-end and tighter acquisition block or just a few software selection jumpers here and there?