Tests with the LeCroy PMA Power Measurement and Analysis (PMA) option

[Wuerstchenhund]

Based on a request from another thread I've done a few very simple tests with LeCroy's Power Measurement and Analyis (PMA) software options available for their better scopes. There are two PMA variants available:

• PMA1 which runs on the older non-Windows scopes (9300 and LC Series running a proprietary OS; WaveRunner LT, WaveRunner2 LT and WavePro 900 running VxWorks)
• PMA2 which is for LeCroy's X-Stream oscilloscopes running Windows

Since I don't have any real-world sources to test I've used my Rigol DG1062z AWG to generate the input signals. For the first sets of test, I've generated two 7V DC voltages with a 3Vpp 55kHz sine wave on top. The Ch2 phase is offset by 20deg from Ch1. For the second set of tests I've generated two 55Hz sine waves with Ch1 going from 0V to 2.5V and Ch2 from 0V to 3.5V. Again the phase offset between both channels is 20deg.

The PMA1 tests have been done on a LeCroy WaveRunner2 LT264M. The PMA2 tests were run on a LeCroy WavePro 7300A-XL. The screenshots are chronologically arranged where possible to show how to get through the options.

Please note that these are quick & dirty tests that barely explore the capabilities of the PMA software or the scopes it runs on.

[Wuerstchenhund]

PMA1 on the WaveRunner2 LT264M.

Ch1: 7V DC voltages with a 3Vpp 55kHz sine wave on top
Ch2: 7V DC voltages with a 3Vpp 55kHz sine wave on top; 20deg phase shift between Ch1 and Ch2

[Wuerstchenhund]

PMA2 on the WavePro 7300A-XL.

Ch1: 7V DC voltages with a 3Vpp 55kHz sine wave on top
Ch2: 7V DC voltages with a 3Vpp 55kHz sine wave on top; 20deg phase shift between Ch1 and Ch2

[Wuerstchenhund]

PMA1 Line Power tests on the WaveRunner2 LT264M.

Ch1: Sine wave 55Hz with Umin=0V and Umax=2.5V
Ch2: Sine wave 55Hz with Umin=0V and Umax=3.5V; 20deg phase shift between Ch1 and Ch2

[Wuerstchenhund]

PMA2 Line Power tests on the WavePro 7300A-XL.

Ch1: Sine wave 55Hz with Umin=0V and Umax=2.5V
Ch2: Sine wave 55Hz with Umin=0V and Umax=3.5V; 20deg phase shift between Ch1 and Ch2

[electronic_eel]

What do they mean by "safe operating area"?

Isn't this something very much depending on the capabilities of the DUT? One power supply can only handle 5 W and the other 500W.

[TunerSandwich]

This is awesome....thank you so much for doing this. 

I can immediately see the benefits of PMA2 over PMA1.  No question about which one is right for our needs. 

Having the metrics plotted directly on the waveform, in a time correlated index, sells it.....it's amazing how significant the losses are during phase angle crossings....

In your opinion is stepping into 7k or better series LeCroy scopes worth the extra investment and "hunting time"?

I am the type of person who would rather just buy and start using something, than agonize over specifications.
 
Also what is the cheapest windows based scope LeCroy has ever made?  I just don't need the bandwidth, and wouldn't mind a sub 350MHz front end (which I assume will have even lower noise floor).....basically i am trying to figure out the cheapest way to get into PMA2.  Since processing power can be "upgraded" cheaply and easily (via computing hardware) and I certainly don't need anything in the RF ranges, I wonder if there is a way to get the memory depth and statistical power for somewhere under $3kUSD

[Wuerstchenhund]

This is awesome....thank you so much for doing this. 

No problem, you're welcome. If you need something else just let me know.

I can immediately see the benefits of PMA2 over PMA1.  No question about which one is right for our needs.

Obviously PMA2 is more advanced than PMA1, but most of the things PMA2 offers can also be done with PMA1.

A PMA2 datasheet can be found here. It shows screenshots of the current scopes but aside from the different theme (brown instead of the older blue) and that it's widescreen it's exactly the same on the WR6k(A).

In your opinion is stepping into 7k or better series LeCroy scopes worth the extra investment and "hunting time"?

Only if you need the extra bandwidth, sample rate, sample memory and the extra features the WavePro Series offers over the WaveRunner. For PMA2 however there shouldn't be a difference feature-wise, and the processing of the WR6kA Series is the same as for the WP7kA, and considering that the WP is huge compared to the WR (which is more of the size of a standard scope) I'd say a WaveRunner scope is the best option for you.

I am the type of person who would rather just buy and start using something, than agonize over specifications.

Well, a tool has to fit, so thinking about the specs to some extend is sensible.
 

Also what is the cheapest windows based scope LeCroy has ever made?  I just don't need the bandwidth, and wouldn't mind a sub 350MHz front end (which I assume will have even lower noise floor).....basically i am trying to figure out the cheapest way to get into PMA2. 

The cheapest Windows scopes are the WaveSurfer Series, the oldest one being the WaveSurfer 400 Series. The WS Series is cut down in features noticeably, and the WS400 doesn't even offer serial decode. As far as I know, none of the WaveSurfers offer PMA2 anyways.

However, another alternative could be buying a WaveRunner Xi (i.e. a 44Xi or 64Xi) which has a bigger screen, faster processing, and a much more compact form factor. With some luck you should be able to get a 44Xi (400MHz 4Ch 10GSa/s 25Mpts 1,250,000 wfms/s) for below $3k. WRXi-PMA2 is if I remember right listed with $1500 but in my experience you should be able to negotiate that closer to $1200, which would give you a great scope. But of course in terms of price/features it's no match to the current listing.

Since processing power can be "upgraded" cheaply and easily (via computing hardware) and I certainly don't need anything in the RF ranges, I wonder if there is a way to get the memory depth and statistical power for somewhere under $3kUSD

Well, processing is very easy to upgrade on the WR6k(A) and WP7k(A) but less so on the other scopes as space and heat requirements are much tighter. In any cases you're limited to the integrated mainboard.

For under $3k I'd say you'd probably be looking at a non-Windows scope with PMA1, i.e. a WaveRunner2 or one of the older LC Series.

[TunerSandwich]

Obviously, you wouldn't buy a 10 mm wrench, for a 12 mm bolt.  i was talking more along the lines of not needing to agonize over that wrench being nickel chrome plate or chrome vanadium plated.

I see a lot of folks get caught up on specs, and by the time they buy something and employ it, they could have finished a dozen projects and financed something even "nicer".  I am an admitted gear whore, but over the years have learned that sometimes it's better to look at needs vs, cost benefit....vs some ideal that is on another planet. 

For example I don't need the front end bandwidth of that 6100A, but I would gladly pay a tad bit more than I need, for features I will never use....than spending 3 months fishing for "exactly the right thing".  In that time frame I could have used the less ideal tool to make some money, to further my endless gear lust. 

One other test I would like to see is a video of those phase shifted waveforms modulating over time.  Which would obviously produce a variety of phase angle crossings.....I am curious as to how quickly the PMA2 on a 6xxx/7xxx series scope can update those metrics, in relation to that real time shift in phase modulation.  That is what we really need to measure on our end....with an update speed fast enough to match our processing core (oscillator if you like).

So really this 6100A is the better value for my needs, considering time vs cost, vs upgrade potential etc....what strikes me about the unit listed is it's exceptional value and availability.  The quicker we get our hands on the right tools, the quicker we can finish our design and worry about other things.  Or find new uses for the scope 

So the MOBO is an integrated unit?  I.E. no PCI or SATA extensions to a discreet/traditional motherboard?  Would love to see the guts of a 6100A.....will try to find that on google. 

The PMA2 is a much easier to read interface, and the overlay of metrics on the waveform (time correlated) makes it "better for our needs" than PMA1.  Part of our need is to migrate the scope into a QC chain, where the operators might have a hard time understanding a less "intuitive" interface.

[Wuerstchenhund]

I see a lot of folks get caught up on specs, and by the time they buy something and employ it, they could have finished a dozen projects and financed something even "nicer".  I am an admitted gear whore, but over the years have learned that sometimes it's better to look at needs vs, cost benefit....vs some ideal that is on another planet. 

I agree. Things change when you have to get tools to earn a living instead of purely for hobby.

For example I don't need the front end bandwidth of that 6100A, but I would gladly pay a tad bit more than I need, for features I will never use....than spending 3 months fishing for "exactly the right thing".  In that time frame I could have used the less ideal tool to make some money, to further my endless gear lust. 

Indeed. ANd don't forget, if you want to sell the scope on say 5 years from now it will still be worth some money because of the options.

One other test I would like to see is a video of those phase shifted waveforms modulating over time.  Which would obviously produce a variety of phase angle crossings.....I am curious as to how quickly the PMA2 on a 6xxx/7xxx series scope can update those metrics, in relation to that real time shift in phase modulation.  That is what we really need to measure on our end....with an update speed fast enough to match our processing core (oscillator if you like).

I can try to do that. Any specifics as to what you'd like the input signals to be?

So the MOBO is an integrated unit?  I.E. no PCI or SATA extensions to a discreet/traditional motherboard? 

No, it's a standard intel 865GCL microATX mobo. However, it uses an DVO ADD AGP card which is specific to this board and which connects to the display and parts of the front panel controls. Other mobos with i865G chipset might work, although there's no benefit in doing that. And intel 865GCL boards can be found cheap as chips.

The PMA2 is a much easier to read interface, and the overlay of metrics on the waveform (time correlated) makes it "better for our needs" than PMA1.  Part of our need is to migrate the scope into a QC chain, where the operators might have a hard time understanding a less "intuitive" interface.

Based on your other posting re. power consumption and heat, I wouldn't completely dismiss PMA1.

[TunerSandwich]

You know that is a good point, about the power consumption.  Obviously one of those smaller scopes is going to be a little bit easier on the thermal and usage issues. 

I might be getting suckered by the fancier waveform overlays. 

I suppose any kind of DC with some AC piggybacking on it, below 60Hz would be great.....just as long as we have modulating phase and an offset of below 8 volts.  Even better would be something that amplitude modulates as well, to simulate some noise from a switching cycle.  That is a tough set of metrics to grab and calculate.  Would really show off the capability (or lack of) in the PMA software vs. scope capture rate (I wonder if the PMA software drops some data).

I had also considered that resale factor, although I generally don't look at the tools as "investments" and they do follow a depreciation schedule.  However it would be nice to take the scope home after it's served it's purpose and "play"...

Gotcha on the embedded GPU...I was thinking more along the lines of upgrading to lower TDP computing components and possibly trying to run some more efficient and powerful hardware.  That's really an afterthought though.  If it does what it does the only real concern was power consumption vs. processing power.  I know I just brought a 300 watt server down to 58 watts by moving to some newer/better hardware and "undervolting" everything...was just a whacky idea about how to get away from that 400w consumption (doubt it would be as dramatic as the server).

Thanks again for all of your help. 

[Wuerstchenhund]

What do they mean by "safe operating area"?

Isn't this something very much depending on the capabilities of the DUT? One power supply can only handle 5 W and the other 500W.

This has nothing to do with the Wattage output of power supplies. It's about switching states in power electronics (which can be a switch-mode PSU, a transmitter, or something else) where certain combinations/timings can be deadly to the electronics involved.

[Wuerstchenhund]

You know that is a good point, about the power consumption.  Obviously one of those smaller scopes is going to be a little bit easier on the thermal and usage issues. 

Well, they're not really smaller scopes, they are just the predecessor models that run VxWorks (or some proprietary LeCroy OS in case of the really old scopes like 9300 and LC Series, but I wouldn't go that far back in your case) instead of Windows. The WaveRunner LT and WaveRunner 2 LT are roughly the same size as the WaveRunner 6000(A) Series (which is the direct successor of the WR2), with the latter being a bit deeper, heavier, noiser and much more power hungry. The screen size is 8.4" for all three Series, but of course only the WR6k has touchscreen (but the older scopes have a very good and intuitive menu system).

However, the older scopes do have advantages. Aside from the noise factor, a scope like the WaveRunner 2 LT boots within seconds and is ready to go, while the Windows-based scopes can take a couple of minutes easily until the OS and all the software is loaded. The WR2 is also easier to carry around as it's much lighter than the WR6k(A) Series, and it's a bit more robust, too.

I suppose any kind of DC with some AC piggybacking on it, below 60Hz would be great.....just as long as we have modulating phase and an offset of below 8 volts.  Even better would be something that amplitude modulates as well, to simulate some noise from a switching cycle.  That is a tough set of metrics to grab and calculate.  Would really show off the capability (or lack of) in the PMA software vs. scope capture rate

I'll see with what I can come up in the next days.

I tried to measure the power consumption of my scopes but the cheap crap power meter has given up. I'll see if I can fix it.

(I wonder if the PMA software drops some data).

Improbable. LeCroy's design mantra says that a scope shall not drop any sampled data.

Gotcha on the embedded GPU...I was thinking more along the lines of upgrading to lower TDP computing components and possibly trying to run some more efficient and powerful hardware.  That's really an afterthought though. 

Well, it's a Pentium4 so there simply are no power efficient processors. The Celerons have a lower TDP but they are much slower and don't support SpeedStep which means they run full throttle all the time. And this means that the most efficient processor for this scope is probably the later 3.2GHz Pentium4's with HT and 1MB cache

If it does what it does the only real concern was power consumption vs. processing power.  I know I just brought a 300 watt server down to 58 watts by moving to some newer/better hardware and "undervolting" everything...was just a whacky idea about how to get away from that 400w consumption (doubt it would be as dramatic as the server).

Well, the majority of the power consumption probably comes from the Front End and the Aquisition System, so even if you could replace the "PC side" with some ultra-efficient components you'd still have a noisy and power hungry heater on your bench.

There simply is a trade off between capability, performance and price on one side and energy efficiency on the other. One of the current HDO4000 scopes will take a lot less power but with PMA2 even for a low bandwidth scope you'd probably be looking at close to or in excess of $10k.

If you want PMA2 at below $5k then something like the WR6100A on ebay (or a similar scope) is probably the only option. If you want low power consumption and minimal heat, and if you can live with PMA1, then you should really consider a WaveRunner2 LT or the older WaveRunner LT. PMA1 doesn't look as nice but it should still get the job done. The simpler presentation hides a lot of what it can do behind.

[blacknoise]

And this means that the most efficient processor for this scope is probably the later 3.2GHz Pentium4's with HT and 1MB cache

If you are willing to exchange more than only the CPU, there is a way to tune the WR6xxx a bit above the Pentium 4 / 3,2 GHz speed limits - and get a cooler and slightly less power hungry system.

My simple recipe for that (working on my WR6100A):

1) an Asus P5PE-VM Motherboard - Lecroy's AGP Video Card is running fine on this platform.

2) an Intel Core 2 Duo E6700 CPU

Unfortunally i couldn't run the E6700 at it's full speed (2.66 GHz) because therefore this CPU needs a FSB of 1066 MHz - while using the Intel-865G-based internal graphics of the Asus-P5PE limits the FSB to only 800 MHz.

So the E6700 runs only with about 2.00 GHz. But even throttled so much, this dual core outperformances Pentium 4 CPUs theoretically by about 2.5 to 3.5 times - with noticeably less heat dissipation compared to the P4 hotheads.

Instead of the E6700 (or E6300, E6320, E6400, E6420, E6600) you also may use E4300, E4400, E4500, E4600 or E4700 Core 2 Duo CPUs in the Asus MB. Then E4xxx-CPU are natively designed for 800 MHz FSB, and so you'll get the fully 2.6 GHz out of an E4700 instead of only 2.00 GHz from the E6700)!

I also tested this configuration - to get astonished by the results: Good a halfe dozen benchmark utilities and also the real FFT calculations on the WR6100A proved me that the E6700 @2.0 GHz was about 20 percent faster than the E4700 @2.6 GHz. Should this be the admirable effect of the E6700 doubled L2 cache (4 MB istead of 2 MB for the E4700)? (*)

Fortunally LeCroys 6.3.0.5 firmware already uses both CPU cores. Updating XP to MicroSofts final state was the only mentionable care on the OS software.

The picture shows an FFT over the 50...150 MHz range, with the local german FM band (88..108 MHz) in the center. The speed is somewhat over 40 FFTs per second (for the upper waveform, the lower shows an averaging by 10 of the upper wave). And please notice, that i made the screen shot in VNCs remote window of the WR6100A, which also comsumps some CPU power.

I still hope some day my final "pimping up" of the Waverunner will be unchaining all the options in the firmware which are currently linked to various option codes. ;-)

But after some (quite short) investigation i currently think that this seems to be much more challanging than unlocking Agilent Windows based scopes...

Kind regards

(*) Please not: all E4xxx and the E6300 to E6400 have 2MB L2 cache - only the E6420, E6600 and E6700 CPUs include the 4MB L2.

[Wuerstchenhund]

And this means that the most efficient processor for this scope is probably the later 3.2GHz Pentium4's with HT and 1MB cache

If you are willing to exchange more than only the CPU, there is a way to tune the WR6xxx a bit above the Pentium 4 / 3,2 GHz speed limits - and get a cooler and slightly less power hungry system.

My simple recipe for that (working on my WR6100A):

1) an Asus P5PE-VM Motherboard - Lecroy's AGP Video Card is running fine on this platform.

2) an Intel Core 2 Duo E6700 CPU

Unfortunally i couldn't run the E6700 at it's full speed (2.66 GHz) because therefore this CPU needs a FSB of 1066 MHz - while using the Intel-865G-based internal graphics of the Asus-P5PE limits the FSB to only 800 MHz.

In theory any AGP mobo that supports ADD cards should work. But that is not the issue.

So the E6700 runs only with about 2.00 GHz. But even throttled so much, this dual core outperformances Pentium 4 CPUs theoretically by about 2.5 to 3.5 times - with noticeably less heat dissipation compared to the P4 hotheads.

Well, yes and no. The Core2 architecture performs roughly around 1.6x as fast as a Pentium4, and a E6700 is roughly equivalent with a Pentium-D 940 (which essentially is a dual core 3.2GHz Pentium4 Prescott). The E6700 has a lower TDP (65W vs 95/130W). So yes, it's more efficient, but there's a problem because

LeCroy's 1st generation X-Stream scopes do not support multi-core.

Simple as that. The software is multi-threaded, and Windows does off-load processes onto multiple cores which makes it look like its fine, but this will inevitably lead to more or less serious side effects (i.e. front panel malfunction and various other niggles). It's a well known problem as multi-core disturbs the delicate timing of the various processes. This means multi-core is out of the question, as is Hyperthreading. If you don't believe me ask LeCroy.

The later XStream II scopes (which are based on a PCIe architecture) do support multi-core. XStream I (which is PCI-based) doesn't. The WR6100(A) is X-Stream I.

What that means is that in order to have your scope working and producing reliable measurements you have to disable multi-core or HT in the BIOS if you upgraded it to multi-core or a CPU with HT, which leaves your dual core processor with only one working core. Even worse, if you have multi-core or HT enabled, simply disabling won't fix the problem. You have to replace the Windows multi-processor HAL with the uniprocessor HAL for it to work correctly again.

The other thing is that once you go over a P4 2.8GHz with 1MB cache the real-world performance increase diminishes a lot. The high CPU load you see in taskman is not because of the math calculations, it's mainly because of the polling of the PCI interface to the aquisition board. The Celerons that came standard with these scopes got saturated, but faster P4's don't, and anything beyond does bring only marginal improvements.

These things are much more complex than they seem to be on a first look. "Pimping" old Agilent Infiniium scopes is much easier as they have a much more primitive architecture. Not so with these LeCroys.

[Mark_O]

One other test I would like to see is a video of those phase shifted waveforms modulating over time.  Which would obviously produce a variety of phase angle crossings.....I am curious as to how quickly the PMA2 on a 6xxx/7xxx series scope can update those metrics, in relation to that real time shift in phase modulation.  That is what we really need to measure on our end....with an update speed fast enough to match our processing core (oscillator if you like).

Do you really need to monitor those measurements in real-time?  I.e., if the scope is producing/updating a batch of stats (max, min, mean, stdDev...), do you also need to capture individual updates to match your core cycle?  Would you even want to be eye-balling it for any significant amount of time anyway?

[blacknoise]

LeCroy's 1st generation X-Stream scopes do not support multi-core.

MANY thanks for your VERY interesting insights!

I didn't know about or even anticipate this problems - and so i never saw a reason to investigate the instruments reliability after my MB- and CPU-exchange.

But i am already anxious to fetch that later. I have a pair of two still "original" WR6050s - so next time i'll modify only one of this machines... ...before i start longer sessions comparing both scopes measurements of identical signals. After that i'll report here about my observations.

Unfortunately "later" stands for something like later next year - currently my capabilities are "painful" limited - my amateur laboratory is beeing renovated / reconstructed from the ground up.

"Pimping" old Agilent Infiniium scopes is much easier as they have a much more primitive architecture.

At least persuading the "agilent.exe" of the xSO548xx and xSO9xxx to liberate all options was a "child's play"... ;-)


Regarding multi-core CPUs i noticed the following:

After replacing the single core CPU of an MSO9254A, running the older XP-Firmware 4.20.0008, with a dual core CPU, Agilents scope firmware still uses only a single core!

Using a processor core affinity utility i could force Agilents firmware to use both cores (to be precise: i had to run the processor core affinity utility somewhat delayed to do it's job - possibly not earlier than some initialisations in Agilents firmware).

But again - without your very interesting knowledge -  i didn't ever check if forcing the firmware to use both cores had unwanted side effects... (until now - that's my next date for next year...;-) )

On a MSO9404A running Agilents second last Windows 7 firmware (05.10.0005) both cores of the dual core CPU are used by default.

Before i switched this MSO9404A to the 05.xx.xxxx firmware, it used the the older Windows 7 firmware up to the version 04.60.0024.

From some "dark" memorization i belive to remember that this older Windows 7 firmware did NOT use the built in (originally, by Agilent!) dual-core CPU, and that this firmware also needed the "assist" of the processor core affinity utility to make use of both cores. Next year i hope to confirm this - or not...

Kind regards

[Wuerstchenhund]

I didn't know about or even anticipate this problems - and so i never saw a reason to investigate the instruments reliability after my MB- and CPU-exchange.

Well, I got cought out by it myself, when I put a Pentium4 3.2Ghz processor in my WP7300A and enabled HyperThreading. The scope felt a bit faster as Windows could now spread system tasks to two virtual CPU cores. But after a short while I ran into the same issues. And then I found out that this is a general issue with the first X-Stream generation

But i am already anxious to fetch that later. I have a pair of two still "original" WR6050s - so next time i'll modify only one of this machines... ...before i start longer sessions comparing both scopes measurements of identical signals. After that i'll report here about my observations.

Just to let you know LeCroy did talk about more serious consequences, i.e. disturbing the hardware communication between Aquisition system and the PC mainboard. This can potentially also affect the key management system (the software keys are stored in the Aquisition system), i.e. corrupting the configuration of your scope or rendering the keys unservicable. And you need them to be ok even for the basic scope functionality, as even this is controlled by a software key (if you loose them you'd end up with a very large door stop). If that happens you'd probably have to send the scope in for repair.

"Pimping" old Agilent Infiniium scopes is much easier as they have a much more primitive architecture.

At least persuading the "agilent.exe" of the xSO548xx and xSO9xxx to liberate all options was a "child's play"... ;-)

The older Agilent Windows scopes use Windows merely as a display and UI engine, with all the important things happening in their ASICs. It's a much more simple approach than LeCroy who uses the Windows part as processing engine.

Regarding multi-core CPUs i noticed the following:

After replacing the single core CPU of an MSO9254A, running the older XP-Firmware 4.20.0008, with a dual core CPU, Agilents scope firmware still uses only a single core!

Using a processor core affinity utility i could force Agilents firmware to use both cores (to be precise: i had to run the processor core affinity utility somewhat delayed to do it's job - possibly not earlier than some initialisations in Agilents firmware).

But again - without your very interesting knowledge -  i didn't ever check if forcing the firmware to use both cores had unwanted side effects... (until now - that's my next date for next year...;-) )

I doubt it will lead to similar issues on the Agilent scopes, as their architecture is different.

On a MSO9404A running Agilents second last Windows 7 firmware (05.10.0005) both cores of the dual core CPU are used by default.

Before i switched this MSO9404A to the 05.xx.xxxx firmware, it used the the older Windows 7 firmware up to the version 04.60.0024.

From some "dark" memorization i belive to remember that this older Windows 7 firmware did NOT use the built in (originally, by Agilent!) dual-core CPU, and that this firmware also needed the "assist" of the processor core affinity utility to make use of both cores. Next year i hope to confirm this - or not...

Speaking of firmware: you know that you can run the current XStream 7.4.0.5 on your WaveRunner 6000A Series scopes? The latest firmware that has official support for the WR6kA is 6.5.0.5 which is ancient and comes with several bugs that have been fixed in later releases. But you can install 7.4.0.5 on top of it (just download the package for the WR Xi-A). Should you do a fresh Windows installation, install 6.5.0.5 first and then 7.3.0.5 (or whatever will be the latest by then) on top. Just installing 7.4.0.5 won't work as it does no longer contain the necessary drivers for the WR6k(A).

The WR6000 (non-A) comes with Windows 2000 and needs to be upgraded to Windows XP first.


An Update:
I talked to a LeCroy Engineer today, and he told me that the situation with the WaveRunner 6000A is slightly different to the WavePro 7000A Series as the WR6kA scopes use a slower aquisition system with a simpler interface, which makes them less susceptible to interference when using multi-core/HT CPUs than the WavePro 7kA. HyperThreading and multi-core are still not supported on these scopes and can cause various issues, but the side effects should be less serious than on the WP7kA, and there should be no risk in hardware damage. However, as always take this with a grain of salt!

On a different note: hey blacknoise, why don't you write a review of your WaveRunner 6000 Series scopes? There's little out there for these scopes so I guess a review would see quite some interest.

[blacknoise]

Speaking of firmware: you know that you can run the current XStream 7.4.0.5 on your WaveRunner 6000A Series scopes?

NO, not at all! But as "first maintenance", still last night i mirrored all scope firmware from LeCroys site on my server... ;-)

Again many thanks for your extremely interesting insights! Including your update! Awsome!

During last day i caught myself cursing the apparently somewhat unfavorable scheduled time for the conversion of my lab. ;-)

If my lab "renaissances" i shurely will try out what i learned from you - and will also be happy to publisch reviews.

But in the moment, while rebuilding the lab, i not only miss the time and relaxation for that - also nearly all items of my collection are curently put down and stored. E.g. all signal sources. So yesterday displaying the FFT over the local FM band was a real compromise solution to show something more exiting than 50 Hz humming...

I left only some scopes (like the Waverunner and the yesterday named Infiniiums) in my (software) workroom to teach them some new tricks (e.g.: "options") during my "labless time"... ;-)

Kind regards

[TunerSandwich]

One other test I would like to see is a video of those phase shifted waveforms modulating over time.  Which would obviously produce a variety of phase angle crossings.....I am curious as to how quickly the PMA2 on a 6xxx/7xxx series scope can update those metrics, in relation to that real time shift in phase modulation.  That is what we really need to measure on our end....with an update speed fast enough to match our processing core (oscillator if you like).

Do you really need to monitor those measurements in real-time?  I.e., if the scope is producing/updating a batch of stats (max, min, mean, stdDev...), do you also need to capture individual updates to match your core cycle?  Would you even want to be eye-balling it for any significant amount of time anyway?

Yes, because of the proprietary nature of our devices.  I need to see the changes in phase angle in real-time to tune the variable caps, and then log that metric and translate into a delay offset algorithm.  I can't say much more about WHY we need to see it this way, as that would give up the "big secrets" behind our technology.....but we defiantly do need to see the phase response and total losses in real time, to hit our engineering targets, in a timely way.  This is why i was so curious about the PMA software dropping data or not. 

Our current solution relies on a lot of external clocking and compensations in software, and even then I don't wholly trust the results.  We can't define the granularity of each switching cycle in the dual bi-polar array without an accurate assessment of phase angle crossing, in a time correlated metric.....with greater resolution than our core switching frequency. 

If I said much more I would need an NDA 

Think of it as hybridized control of inverting / non inverting amplifiers.....the amplifier class and state changes at double the oscillator rate, and the polling/sense loop has to read at double that....and all of those signals need to have minimal losses at phase angle crossings, to equal our efficiency targets and accuracy targets....

The traditional ways of doing this won't work, because of footprint, thermal, and efficiency issues.....this stuff doesn't plug into the wall