Agilent MSO 5000/6000/7000 - anyone hacked these scopes?

[nctnico]

I notice the other day thay Keysight are discontinuing the 6000/7000 series on 1 June 2016. Not surprised, they've had a fair run.

In some ways though it's a backwards step in terms of screen resolution and memory depth, the current equivalent models have almost 40% worse screen resolution and half the memory depth. Admittedly some other options and specifications are improved though.

I agree. There is a bit of a void in Keysight's lineup and by today's standards their memory depth is severly lacking especially when taking into account how wildly optimistic the memory depth is (over) specified. I like my 7104A (appearantly I bought it hacked) but I very much would like to have deeper memory.

[Carrington]

I notice the other day thay Keysight are discontinuing the 6000/7000 series on 1 June 2016. Not surprised, they've had a fair run.

Yes, as might be expected. However, these machines are still fully capable these days.

I'm also surprised that over something like 10 years of pretty much daily use the CCFL backlight is still going strong!

I hope that mine also lasts a lot of time.

I wonder if they'll do what Tek did a few years ago and release a final firmware update to enable all options.
Not that we need it now, but it would be a nice gesture!

That would be great.

I'm somewhat surprised they didn't send you one ...

Absolutely, Keysight should do that with one of these new beasts.

... the way you put a 6000 through its paces is one of the best scope demos I've ever seen.

In fact these demos convinced me.  When you see them, you think: Look what you can do with these machines!
Of course, I also had in mind its specifications and price (used one).

... I very much would like to have deeper memory.

Yes, it's the only downside, but when it was designed 8M was a lot.

[mikeselectricstuff]

I'm somewhat surprised they didn't send you one ...

Absolutely, Keysight should do that with one of these new beasts.

I already have have a spare MSO6034A, but if they have a 6054 or 6104 lying around I'd happily accept!

... I very much would like to have deeper memory.

Yes, it's the only downside, but when it was designed 8M was a lot.

... I very much would like to have deeper memory.

Yes, it's the only downside, but when it was designed 8M was a lot.

I find memory size is very rarely a problem, especially when you understand how to optimise it by selecting the  right permutation of acquire mode, channels and single-shot mode.
 And a lot of the situations where memory is an issue can be dealt with using segmented. 

[Carrington]

I already have have a spare MSO6034A, but if they have a 6054 or 6104 lying around I'd happily accept!

I know. 
But I was referring to one of those with an X in its name.

[nctnico]

... I very much would like to have deeper memory.

Yes, it's the only downside, but when it was designed 8M was a lot.

I find memory size is very rarely a problem, especially when you understand how to optimise it by selecting the  right permutation of acquire mode, channels and single-shot mode.
 And a lot of the situations where memory is an issue can be dealt with using segmented.
[/quote]
Ofcourse you can jump through some hoops and make most of it (only using channel 1 & 3 and use single shot mode) but still more memory makes life easier in general. For example: my first logic analyser (Philips PM-something-) had 2kbit per channel. Ofcourse setting the triggering and samplerate would allow for making most of it but for many measurements you had to take several acquisitiongs with different settings and then piece everything together. Fast forward to today: my fourth logic analyser has 32Mbit per channel. If I need to measure something I just let it trigger on a line starting to wiggle and go from there. It can capture long timespans with a lot of details so nowadays I often take just one acquisition which has all the data and analyse offline.

[Howardlong]

My only comment about segmented memory is that it's a bit of a fiddle to set up. My default mode is to just do without, but when I do use it I have to switch down a gear because of the setup, and generally I never get it right first time. It just gets in the way of my workflow a bit, but equally, it's a damn fine answer to "how much deep memory is enough?" because in some very sparse scenarios effectively you have almost unlimited memory.

Given the choice between, say, 4Mpt/ch with segmented memory and 20Mpt/ch without, it's a tough choice because I want it all! I guess in the end, segmented memory realistically gives you more options as long as you have a reasonably repeatable scenario to get your triggering, timebase and segmented memory just right.

[mikeselectricstuff]

As long as you have  a good trigger, most scenarios that need a lot of memory can be dealt with segmented - you very rarely need all of a large acquisition, only a small part of it. If you can capture all the small parts, not only does that save memory, but also gives additional useful info like the number in a sequence, i.e. you know that a particular burst is the <segnum>'th burst, and how many burst have occurred, which is something that can be  hard to do with a single deep acquisition.

[barryforever]

@Mike : I could not find the version sys6000.bin on Keysight Website could you upload it to some file sharing website so that I could download it.
Another question is I have a Agilent MSO6034A oscilloscope ,I have read that the hack you were talking about would only enable the user to activate all other options except for MSO is it true?

[Berni]

@Mike : I could not find the version sys6000.bin on Keysight Website could you upload it to some file sharing website so that I could download it.
Another question is I have a Agilent MSO6034A oscilloscope ,I have read that the hack you were talking about would only enable the user to activate all other options except for MSO is it true?

I tweaked the hack to work on the new firmware file since i could not find the old one anywhere.

As for the MSO i wouldn't know since my scope already was MSO before. But i don't see why it wouldn't work since there is a license code for enabling MSO (Comes as part of a MSO upgrade kit that includes MSO probes and a new sticker that says MSO).

[Howardlong]

Certainly on my ex-DSO7000 it miraculously became an MSO7000. You will, of course, need to find the physical LA probes too!

[Samogon]

Well, that sucks.  I have an MSO7104A that needs a new mainboard too, but the reason is more subtle: it will not run the self cal due to some sort of digitizer issue.  However, I have not noticed any aberrations during normal operation, so I am sitting on it right now because I can't afford the repair cost of a mainboard swap.  This is the main problem of these monolithic designs; it's not possible to replace the failed part individually since the LRU is the whole damn board.

Hi,
I have the same error on my DSO5054A on ch1
And ch1 and ch2 showing sine 1Ghz and 200V with no imput on them. This error prevent self calibration which prevents using auto set functionality. Did you find anything?

[HighVoltage]

Can someone here please provide me the older firmware version for the 7000B series
I am looking for the FW file:
"sys7000A.bin 06.16.0001"

It is no longer available on the Keysight website.

Thank you.

[mikeselectricstuff]

Check out this hack

[plesa]

Check out this hack

Best hack of Agilent/Keysight scope ever.
Congratulation!

[HighVoltage]

WOW that is a real FW update out of the Enterprise !

[Berni]

Dang i need to try that firmware on my MSO6034A. But then again it would ruin then nice quick boot time.

[KE5FX]

That's the fabled "HP Way" in action.     Where HP = Harry Potter.

[berk98]

The hack which Mike detailed before also seems to work with firmware version v6.20.0000, at least on my DSO 6000 scope.  Only difference was the offset where the edit is made.  Same data and all.  In short,

sys6000.bin 6.20.0000:
002E80F8: 94 21 F8 E0 7C 08 02 A6 -> change to -> 38 60 00 00 4E 80 00 20

I think the same is true for the 7000 series, since the magic sequence appears at offset 002E81D8, but I haven't tried it.

[TheSteve]

Howard, thanks for the detailed probe info, exactly what I needed.  I just bought two 1160A probe sets on eBay which I'm pretty sure are the equivalent early version of the 1165A you recommended - they come in neat zipper pouches with a load of different probe accessories - got them for $80 each.  The guy had them listed on eBay as 10073C, probably because nobody was going to search for '1160A'.  I guess if they don't work, I can return them as technically they are not the advertised item.

I'm still working on the LA probe stuff.

The more I read bout the MSO7104B, the more confident I get that this was a great choice.  I thought I might be having buyers remorse right now but I'm not.  A scope is one of the basic tools I use to make my living and, to date, I've been struggling with a Rigol MSO2202A and had access to a Tektronix MDO3204 but they are both dogs when it comes to I2C / SPI serial bus decode - to the point of being unusable, looks like the MSO7104 is the dogs bollocks in this regard with full speed hardware decode at 100,000 Wf/sec.

I am not certain how well the 1160A will compensate, the difference between the 1160A and the 1165A in terms of compensation range is 6-9pF and 12-14pF respectively, with the scope input spec'd at 14pF.

Also be aware that SPI decoding fails at a clock speed beyond about 35MHz (it's spec'd to 25MHz).

The best thing about the 5000/6000/7000 series is the UI in terms of ease of use and responsiveness. Everything is obvious. In contrast, the UI of the MDO3000 you mentioned isn't any of those things!

I recently purchased four 1161A's for my DSOX3014A, the seller sent me three 1161A's and one 1160A by mistake. The seller sent a replacement 1161A but also told me to keep the 1160A. I really didn't think it would work as its compensation range is listed as 6-9 pF. I gave it a try and it compensated fine on my 3014A which has a specified input capacitance of 14 pF. I wouldn't recommend buying a probe that is not in the range of the scope but perhaps the 1160A's have a larger range then specified.

edit - one other thing I noticed. A full 1160A/1161A kit includes a replacement probe tip. If you swap the tips between probe types it looks to switch the official compensation range.
Also on the DSOX3014A the 1161A is detected as a 10x probe and disables the 50 ohm menu, the 1160A is also detected as a 10X probe but does not disable the 50 ohm termination menu. I don't believe they are officially supported probes on my scope and this thread is for the 5000/6000/7000 but I thought I'd include the info for the sake of completeness.

[Neganur]

the 1160-series is quite nice, I got a couple 1161A and one 1163A for cheap. I did not imagine them this small and I think they're nicer than the newer 2.5mm probes that come with the 4000X series and above.

[barryforever]

@berk98

I could not find this key 0x002E80F8 for my Sys6000.bin version 6.20 . Mine is a MSOx6000 series oscilloscope.
Can you find this key under your sys6000.bin.

Has anybody got his one?

[Berni]

@berk98

I could not find this key 0x002E80F8 for my Sys6000.bin version 6.20 . Mine is a MSOx6000 series oscilloscope.
Can you find this key under your sys6000.bin.

Has anybody got his one?

Looks like Keysight was quick to patch that one. Rigol already gave the sales of there x3000 series scopes quite a blow, image what people turning there 6000 to x3000 would do to there sales

[HighVoltage]

Is there any known hack for a DSOX6004A with FW 6.10 installed?
Mainly to upgrade the BW to 6 GHz and make a MSO out of the DSO?
Thanks for any feedback.

[Berni]

Is there any known hack for a DSOX6004A with FW 6.10 installed?
Mainly to upgrade the BW to 6 GHz and make a MSO out of the DSO?
Thanks for any feedback.

Considering how rare and new these scopes are there proabobly is none yet. But seeing the software has an awful lot in common with the x3000 the way to do it is likely similar. If you have a X6000 to hand try poking around with the existing X3000 hack methods.

Even the fairly old by now 9000 series had no hack out there, I had to develop one on my own. But from what I saw it poroabobly works on all modern windows based scopes, including the S series that is software upgradable to 8GHz (but I don't have access to one in order to test it out.

[memset]

Is there any known hack for a DSOX6004A with FW 6.10 installed?

Should be all the same as 3000X-series.
BTW, it's interesting what people think about 6000X vs. DSO-S series.