Author Topic: Teledyne LeCroy WaveRunner 64xi, morphed to differential probes... (Read 56484 times)

joeqsmith · « **on:** April 12, 2015, 09:51:45 pm »

I've had this DSO for a while and had it apart once before to see if I could find a problem with a dead channel. I had installed the latest software at that time cleaned up the drive. The system boots really slow and the hard drive was always active. Came with 256M of RAM.

So today it's coming back apart for a little TLC.

This is my forth LeCroy DSO and also my newest. I also have an early 7200 68000 VME from the USAF, a 7200A that came from IOMega, my 8500A.

The 64xi is not even close to the quality of these systems. Not sure why they would want to start marketing such low end equipment. The price for this system is still over 10K, so not seeing it competing with the real low end systems. I would love to see how the top end Rigol DSO compares with my 30 year collection.

joeqsmith · « **Reply #1 on:** April 12, 2015, 09:56:05 pm »

I was walking with the 64xi and some of the knobs fell off.

I started looking and about half of them were super loose. LeCroy sent me some new knobs and these are very tight. The vertical and horizontal knobs are supposed to be larger. Waiting for those two knobs to show up.

My 7200 has an end cap on the knobs that you first remove, then you remove a set screw that holds the knobs on.

joeqsmith · « **Reply #2 on:** April 12, 2015, 09:58:23 pm »

Here are a few from the sides and back.

joeqsmith · « **Reply #3 on:** April 12, 2015, 10:04:32 pm »

I want to replace the hard drive. To do this, you have to pull this cover. Of course it's all plastic. Self tapping screws. There was some damage already to one of the posts. I know this system has been in for service at least twice in its short life.

No shock mounts, just bolted to the case. Maybe this thing is not meant to ever be moved off the bench.

dom0 · « **Reply #4 on:** April 12, 2015, 10:06:56 pm »

Well it's a 2.5" notebook drive (apparently) and those are meant to resist shock and transportation w/o shock dampeners.

joeqsmith · « **Reply #5 on:** April 12, 2015, 10:07:42 pm »

Looking at the main acquisition board, get ready for it.....

You have to remove all those covers and the plate over the BNCs to get to this one screw under the main heatsink.

joeqsmith · « **Reply #6 on:** April 12, 2015, 10:12:35 pm »

Under the cover holding the drive. Upper right plastic standoff is damaged. Looks like the tape has started to lift.

Actually, it's pretty much held together with tape. Maybe their hold stock in 3M.

joeqsmith · « **Reply #7 on:** April 12, 2015, 10:16:14 pm »

I am not sure why, but the two main cables going to the DAQ are hot glued in place. These are the only two. I had remelted these last time I had it apart.

With the DAQ pulled, you can see the channel 3 connector. This has actually been fine since I put it back together last time so I plan to leave this together.

joeqsmith · « **Reply #8 on:** April 12, 2015, 10:18:31 pm »

These were some of the pictures I took last time I had the DAQ apart.

joeqsmith · « **Reply #9 on:** April 12, 2015, 10:21:44 pm »

Once the DAQ is pulled, there is another cover that must be removed to get to the CPU board. That's a lot of screws to add some RAM and change the battery.

tautech · « **Reply #10 on:** April 12, 2015, 10:23:33 pm »

Quote from: joeqsmith on April 12, 2015, 09:53:18 pm

Case is all plastic. Of course, it has some chips and cracks. My old 7200 is all metal.

I don't think anybody in their right mind would want a metal cased scope these days.

Quote from: joeqsmith on April 12, 2015, 09:56:05 pm

My 7200 has an end cap on the knobs that you first remove, then you remove a set screw that holds the knobs on.

And the set screws are a point of contact to the operator should he/her ever decide to "float" the scope.
Assuming metal pot spindle of course.

joeqsmith · « **Reply #11 on:** April 12, 2015, 10:25:44 pm »

There is only one memory slot. I picked up a 1G stick for it. It can't hurt.

Battery replaced, RAM installed, DAQ back in the chassis.

joeqsmith · « **Reply #12 on:** April 12, 2015, 10:29:51 pm »

Had replaced the drive on the 8500A with an SSD. It's been working very well. I had bought two different ones to try. One SATA one PATA. The SATA with adapter went in the 8500A, the PATA is going into this system. Different brand, we will see.

Imaged the original first then did the transfer.

joeqsmith · « **Reply #13 on:** April 12, 2015, 10:31:53 pm »

Put the new SSD into the system and turned it on. Boots fast.

joeqsmith · « **Reply #14 on:** April 12, 2015, 10:33:42 pm »

Left over parts... This is the screw that was in the broken standoff. Used a longer self tapping screw to make sure this one was not going to fall out.

joeqsmith · « **Reply #15 on:** April 12, 2015, 10:38:33 pm »

Back together and powering it up for checkout.

No worse for wear anyway. All four channels survived. I may try and find an original cover for it.

digsys · « **Reply #16 on:** April 12, 2015, 10:57:02 pm »

Thanks for the teardown and pix. I have a collection of old LeCroys as well, but just too lazy to fix the minor faults :-)
Both my 9424 and 9354 have a lot of metal and are damn heavy, easy to get into though. Still, I'd prefer a a much lighter plastic version
next time.

Hydrawerk · « **Reply #17 on:** April 12, 2015, 10:58:14 pm »

These wires were placed in the factory or during a repair??

digsys · « **Reply #18 on:** April 12, 2015, 11:01:42 pm »

Quote from: Hydrawerk

These wires were placed in the factory or during a repair??

I'd say definitely from a repair. Factory mods are usually very neat with much lighter, high grade wire. We used to use wire-wrap wire.

joeqsmith · « **Reply #19 on:** April 13, 2015, 12:55:57 am »

There was another picture of the same wire being used. If it was a repair, it was done at the factory. I have attached a different, larger picture of this area.

This thing now has waveforms being displayed in 1 minute 5 seconds after power up.

It no longer stalls when I am running it. Just very smooth as I would expect it would be. 4X more memory seems to have made a big improvement.

joeqsmith · « **Reply #20 on:** April 13, 2015, 01:00:13 am »

Running interlaced mode (10Gs/s), channel 2 (red) with a 800MHz signal. Channel 3 (blue) with a 500MHz signal.

joeqsmith · « **Reply #21 on:** April 13, 2015, 01:05:35 am »

Showing the two channels looking at the rising edge of an ECL signal. This is the same signal, same coax.

Wuerstchenhund · « **Reply #22 on:** April 13, 2015, 04:28:46 pm »

Quote from: joeqsmith on April 12, 2015, 10:29:51 pm

Had replaced the drive on the 8500A with an SSD. It's been working very well. I had bought two different ones to try. One SATA one PATA. The SATA with adapter went in the 8500A, the PATA is going into this system. Different brand, we will see.

Just be careful with SSDs in that scope, as they tend to suffer from file system corruption (I tried several ones, including the Transcend PATA SSD on your picture), and all suffered from the same problem.

The reason seems to be that the SSDs only really support UDMA modes, which for PATA requires a 80 conductor cable to work reliably. The scope however uses a 44pin cable (there are no 84 conductor cables, at least I'm not aware of them) which is not really suited for UDMA66 or faster modes. Another forum user (Tunersandwich) who has the same scope made the same experience, he also tried a SSHD but that was a no-go either.

In the end I went back to spinning rust, i.e. a modern fast SATA laptop drive connected through a cheap SATA-PATA bridge. Works fine and absolutely reliable.

Quote

$7350 for the Rigol DS6064 or $9660 for the DS6104. I assume this is the market LeCroy is going after.

Not really. The WaveRunner Series starts quite a bit above that.

The WaveSurfer 3000 is the one that competes with the Rigol DS6000 scopes, although no-one really buys them anyways due to their poor value for money.

tautech · « **Reply #23 on:** April 13, 2015, 08:13:54 pm »

Quote from: Wuerstchenhund on April 13, 2015, 04:28:46 pm

Quote from: joeqsmith on April 12, 2015, 10:29:51 pm
Had replaced the drive on the 8500A with an SSD. It's been working very well. I had bought two different ones to try. One SATA one PATA. The SATA with adapter went in the 8500A, the PATA is going into this system. Different brand, we will see.

Just be careful with SSDs in that scope, as they tend to suffer from file system corruption (I tried several ones, including the Transcend PATA SSD on your picture), and all suffered from the same problem.

The reason seems to be that the SSDs only really support UDMA modes, which for PATA requires a 80 conductor cable to work reliably. The scope however uses a 44pin cable (there are no 84 conductor cables, at least I'm not aware of them) which is not really suited for UDMA66 or faster modes. Another forum user (Tunersandwich) who has the same scope made the same experience, he also tried a SSHD but that was a no-go either.

In the end I went back to spinning rust, i.e. a modern fast SATA laptop drive connected through a cheap SATA-PATA bridge. Works fine and absolutely reliable.

Hmmm, IME the 80 conductor cable has an inactive conductor between each active on to reduce crosstalk and they were quite available in the days of IDE.

SSD management might be the other problem, OS's have "switches" than can be enabled to help lengthen the life of SSD's. Auto Defrag should be off, wear leveling on, etc.

Sevenforums had a lot of info about this when SSD's came out, but whether the scope OS has the updates to fully support SSD's

Wuerstchenhund · « **Reply #24 on:** April 13, 2015, 08:59:13 pm »

Quote from: tautech on April 13, 2015, 08:13:54 pm

Hmmm, IME the 80 conductor cable has an inactive conductor between each active on to reduce crosstalk

Correct.

Quote

and they were quite available in the days of IDE.

Yes, for desktop IDE components (which used a 40pin connector). But not for the much narrower 44pin variant used for laptop drives. The main reason is that laptops rarely use cables to connect to hard disks but direct connectors.

Quote

SSD management might be the other problem, OS's have "switches" than can be enabled to help lengthen the life of SSD's. Auto Defrag should be off, wear leveling on, etc.

Windows XP (which is what the scope runs) doesn't support trim so it is dependent on the SSD's wear leveling (which is always on btw). Over-provisioning (i.e. not using the full capacity of the SSD) helps.

Background defrag should be off as it's useless on a SSD anyways.

Quote

Sevenforums had a lot of info about this when SSD's came out, but whether the scope OS has the updates to fully support SSD's

There are no updates. Windows versions before Win 7 do not support SSDs, period. It doesn't mean they can't run on SSDs but there are some things to look out for.

tautech · « **Reply #25 on:** April 13, 2015, 09:02:16 pm »

Quote from: Wuerstchenhund on April 13, 2015, 08:59:13 pm

Quote from: tautech on April 13, 2015, 08:13:54 pm
Hmmm, IME the 80 conductor cable has an inactive conductor between each active on to reduce crosstalk

Correct.

Quote
and they were quite available in the days of IDE.

Yes, for desktop IDE components (which used a 40pin connector). But not for the much narrower 44pin variant used for laptop drives. The main reason is that laptops rarely use cables to connect to hard disks but direct connectors.

Of course.

joeqsmith · « **Reply #26 on:** April 13, 2015, 10:23:59 pm »

Quote from: Wuerstchenhund on April 13, 2015, 04:28:46 pm

Just be careful with SSDs in that scope, as they tend to suffer from file system corruption (I tried several ones, including the Transcend PATA SSD on your picture), and all suffered from the same problem.

The reason seems to be that the SSDs only really support UDMA modes, which for PATA requires a 80 conductor cable to work reliably. The scope however uses a 44pin cable (there are no 84 conductor cables, at least I'm not aware of them) which is not really suited for UDMA66 or faster modes. Another forum user (Tunersandwich) who has the same scope made the same experience, he also tried a SSHD but that was a no-go either.

In the end I went back to spinning rust, i.e. a modern fast SATA laptop drive connected through a cheap SATA-PATA bridge. Works fine and absolutely reliable.

Quote
$7350 for the Rigol DS6064 or $9660 for the DS6104. I assume this is the market LeCroy is going after.

Not really. The WaveRunner Series starts quite a bit above that.

The WaveSurfer 3000 is the one that competes with the Rigol DS6000 scopes, although no-one really buys them anyways due to their poor value for money.

64xi was $11,500 new. A little higher price than the Rigol. I would like to see how the DS6104 compares.

No problems at all with the 8500A using the PATA to SATA with the SSD. It's been in there about four months now.

I looked up the controller and old drive. Both supported UDMA5. Appears to be running in UDMA2. Guess they just skimped on the cable.

I ran the 64xi at least 10 hours over the weekend. Maybe 5 power ups. I'll post if it starts to have any troubles.

You still have one of these right? Does your DAQ have the same rework?

XFDDesign · « **Reply #27 on:** April 14, 2015, 02:16:36 am »

Out of curiosity, how old is the unit?

Wuerstchenhund · « **Reply #28 on:** April 14, 2015, 05:36:44 am »

Quote from: joeqsmith on April 13, 2015, 10:23:59 pm

Quote
Quote
$7350 for the Rigol DS6064 or $9660 for the DS6104. I assume this is the market LeCroy is going after.

Not really. The WaveRunner Series starts quite a bit above that.

The WaveSurfer 3000 is the one that competes with the Rigol DS6000 scopes, although no-one really buys them anyways due to their poor value for money.

64xi was $11,500 new. A little higher price than the Rigol.

Yes but that was $11,500 in 2006, not in today's dollar. These days (2nd gen successor) WR6zi starts at around $13,900.

Quote

I would like to see how the DS6104 compares.

Not well. The DS6000 is a very simple scope. Only very basic maths, slower sample rate (5GSa/s for the Rigol vs 10GSa/s for the WR64Xi), slower max waveform rate (180k for the Rigol vs 1.2M for the WR64Xi), only simple and basic measurements, no advanced options (i.e. Advanced Math, Jitter Analysis, Power Analysis, Serial Decode for a ton of standards, or one of the ton of other options that exist for the WR64Xi), and because it's essentially the same platform as the smaller Rigol models (embedded Linux) there's no Windows, and no means to install Mathlab, LabView or any other program directly on the scope.

The Rigol does have larger sample memory (140M) but that isn't of much use as there's no search capability whatsoever, and scrolling through the memory is slow on the Rigol.

The DS6000 is completely overpriced for what it offers, not just in comparison with older scopes like the WR64Xi, but also in comparison with current scopes in the same price bracket like the WaveRunner 3000 or the R&S RTM2000/RTE, or even the (aside from the new touch screen functionality pretty dated) Keysight DSOX3000T Series.

Quote

No problems at all with the 8500A using the PATA to SATA with the SSD. It's been in there about four months now.

Yes, but that is a scope with a standard 40pin SATA connection used in desktop PCs, and the 8500A also came with an 80 conductor cable.

Quote

I ran the 64xi at least 10 hours over the weekend. Maybe 5 power ups. I'll post if it starts to have any troubles.

Please do, it would be interesting. I had these problems with a scope from 2006, and the one Tunersandwich has is IIRC from 2008, and he was affected as well. But it took a while to show.

Quote

You still have one of these right? Does your DAQ have the same rework?

Yes, I have. I can't remember about any rework but I see if I can take some pictures.

joeqsmith · « **Reply #29 on:** April 14, 2015, 11:23:54 am »

I was just trying to find the highest end Rigol offers to compare and assumed it would be lower cost.

Was not sure about the features. This one only has the DFP2 enabled. Wonder what they would charge turn turn on all of the features at this point. With the 8500A, they wanted the full price for the JTA2 until they offered it for making that video.

Wonder it the quality is any better on the Rigol. The 64xi is only 8 years old. Been in for service at least once for dead channel. You were saying yours has dead channels as well with the connector problems. Wonder how the Rigols are holding up.

Yes, the 8500A uses the full size 80-pin cable. Not sure what UDMA mode it is using with that crazy setup. The 33MHz mode should be fine on the 64xi. Maybe something else was causing your problems. Any idea how long yours ran for before running into problems? Did it just fail to boot?

Wuerstchenhund · « **Reply #30 on:** April 14, 2015, 11:54:31 am »

Quote from: joeqsmith on April 14, 2015, 11:23:54 am

Was not sure about the features. This one only has the DFP2 enabled. Wonder what they would charge turn turn on all of the features at this point. With the 8500A, they wanted the full price for the JTA2 until they offered it for making that video.

Well, the price for the software options very much depend on if you're a business (and if so, how good you're at negotiating) or a hobbyist (in which case you might even get some enabled for free).

Quote

Wonder it the quality is any better on the Rigol. The 64xi is only 8 years old. Been in for service at least once for dead channel. You were saying yours has dead channels as well with the connector problems. Wonder how the Rigols are holding up.

I'd guess a bit better (they are pretty simple scopes with I guess a single PCB), and build quality wide the WR64Xi is really a few steps down from other LeCroy scopes. The successor (WRXi-A) uses the same chassis, but apparently the build quality has improved a lot in them.

Quote

Yes, the 8500A uses the full size 80-pin cable. Not sure what UDMA mode it is using with that crazy setup.

I'd guess it's running at UMDA/133. I wonder why you used the PATA cable for the SSD, as the WM8kA should have SATA ports on its mainboard which would be easier (and a bit faster, too).

Quote

The 33MHz mode should be fine on the 64xi. Maybe something else was causing your problems.

UMDA33 is fine, however most SSDs seem to want to switch into faster modes, and without the additional conductors in the cable there is some corruption.

Quote

Any idea how long yours ran for before running into problems? Did it just fail to boot?

Difficult to say, in some cases it showed after several boots, in others it ran fine for a few days. Most of the time it booted, but after a while I got strange errors, and when I checked the file system I found some corruption. The scope's hardware is fine, as were the drives. And it never happens with hard drives (SATA or PATA), only with SSDs.

joeqsmith · « **Reply #31 on:** April 14, 2015, 12:28:01 pm »

Quote

I wonder why you used the PATA cable for the SSD, as the WM8kA should have SATA ports on its mainboard which would be easier (and a bit faster, too).

My 8500A has an Intel D845GE motherboard and no support for SATA. I thought about adding a card but then wanted to add the 1G Ethernet board.

Quote

Difficult to say, in some cases it showed after several boots, in others it ran fine for a few days. Most of the time it booted, but after a while I got strange errors, and when I checked the file system I found some corruption. The scope's hardware is fine, as were the drives. And it never happens with hard drives (SATA or PATA), only with SSDs.

Good to know. I'll start keeping a record for it. I would guess it now has at least 20 power ups and 20 hours of operation with the SSD installed. When I store data, I have been using an external USB stick.

XFDDesign · « **Reply #32 on:** April 14, 2015, 01:42:13 pm »

Do you know if all/most of the LeCroys in this series have replaceable PC kit? I've been looking at the WaveJet line, but the WaveSurfer isn't too much more money for a bit better digitizer engine.

Wuerstchenhund · « **Reply #33 on:** April 14, 2015, 06:20:47 pm »

Quote from: XFDDesign on April 14, 2015, 01:42:13 pm

Do you know if all/most of the LeCroys in this series have replaceable PC kit? I've been looking at the WaveJet line, but the WaveSurfer isn't too much more money for a bit better digitizer engine.

Depends on what you mean with "replaceable PC kit". All LeCroy scopes starting with the WaveSurfer Series are Windows (and thus more or less "PC") based. In fact, using "PC" technology (i.e standard computer technology) is one of the reasons LeCroy got to where they are, which is the top of all high end scope vendors.

Scopes below the WaveSurfer Series (like the WaveJet) are just bought in as complete package from Siglent and Iwatsu and rebadged with a LeCroy label.

However, if you ask if you can swap mainboards around in those scopes, then the answer is 'no' aside from the WaveMaster 8k(A), WavePro 7k(A) and WaveRunner 6k(A) (which use standard intel business mainboards), and even on those scopes the choice of mainboards is very limited due to the display controller. The WRXi and most other LeCroy Windows scopes use industrial mainboards in a specific configuration.

The PC part should really be considered as any other part of a modern scope, i.e. manufacturer proprietary. What usually can be swapped however are hard disks, RAM and the processor, which is useful as LeCroy tended to skimp on the processor they put in their older Windows scopes. The WRXi for example comes with a 1.3GHz Celeron M with 512k cache and 400MHz FSB, and upgrading mine to a P-M 1.8GHz with 2MB and 533MHz FSB made a huge difference, simply because a LeCroy X-Stream scope does all calculations in the very fast CPU cache.

XFDDesign · « **Reply #34 on:** April 15, 2015, 08:49:27 pm »

LeCroy was actually at my day job today, showing off their full suite of products.

On the HDO6k scopes, they're using a standard form-factor mainboard, and they have access to that area without a lot of fuss. The WaveSurfer3k series is all embedded stuff, no desktop processors in the thing. Part of the question was in terms of repair capacity but unlike Tek, LeCroy has a support sustaining model that is quite nice.

The space of the question, was if it was possible to buy an identical (chipset & class) mainboard, would it be possible to sustain the use if/when the thing burns out, and if that was common for this line of scopes. With the HDO6k at least, it appears the answer is yes.

XFDDesign · « **Reply #35 on:** April 16, 2015, 02:09:22 am »

On the HDO6ks, it's basically a micro-atx mainboard and periphery set. It appears (I didn't get to take one apart) that it should be removing the back panel to get access to it. Much less hassle than you went through.

I think you mentioned you have used a number of LeCroy systems. How do you like them overall (factoring in this experience)?

joeqsmith · « **Reply #36 on:** April 16, 2015, 04:24:08 am »

As much as I bitch about them, consider this is all I have for scopes at home should tell you what I think of the brand overall.

I would guess the original price for the 7200, 7200A and 8500A were all well over $60K new. These are all very nice systems. I would put my first 7200 up against most of the new DSOs and it was built 1989. Consider it came from the USAF, that thing was put through the ringer and is still rock solid to this day. The performance and features were second to none at that time. You will see I still use the 7200A in most videos I make. That DSO had the SCSI and deep memory options and I would guess it was part of IOMega's production test setup. Again, very reliable.

Video I made showing my old 8500A with the JTA2 enabled. It's all about the phase noise you asked about...

Considering the age of this setup I would like to see what they have now.

Where I currently work, we had been using LeCroy for some time for the higher end DSOs. The software, firmware and FPGA updates for them were a nightmare. They would make changes and it would break something. Interesting enough, when I got the 8500A, I installed all the latest updates and tested against every problem I had found with X-stream and I am VERY happy to say it has been VERY stable and bug free. But these are pretty mature systems now.

Some of the people I have had help me from LeCroy were second to none. Over all I have been very happy with their service, even for my home hobby sort of work they have really done more than I could ask for to help me out.

At work I sent one of the high end systems in for a cal one time. These don't get used very often and when it was returned it sat for a fair period of time before I needed it. Turned out a channel was dead. This DSO was never used by anyone but myself. I called LeCroy, explained the whole thing. Was told it had been too long. $3K and they repaired the damaged channel. For the high end DSOs, this is the only failure that system ever had and I believe it happened during calibration.

For work we starting using these low end systems like the 64xi for general work. I have heard of several failures with them at work. This particular 64xi was scrapped a few years ago when channel 3 failed a second time and they quoted $5K to repair it. At work these scopes get a lot of hours on them as they are used in various test fixtures. They really only get bounced around when they are sent out for cal.

We have since started buying all Agilent DSOs except for the automated tests where the code base is big enough we stay with LeCroy. I have not heard of any failures with the Agilents. It's too early to say how the black ones are going to hold up in a test environment. Ask me in a few years.

Again, I would like to get my hands on a top of the line Rigol and see how it compares. Right now my 64xi has minimal features so wouldn't make for much of a demo. At least it's running good and has all the knobs now.

If I needed a high end scope for work again and was looking to spend $80K, Teledyne would be the first call I made!

XFDDesign · « **Reply #37 on:** April 16, 2015, 01:36:11 pm »

Good to know, thanks!

When LeCroy visited here, I got to fondle the WaveSurfer3054 and found it to be a very nice scope. I particularly liked how they do their "analog persistence" as it "feels" (i.e. is subjective) much closer to the classical CRO display. What has pushed me over the fence towards that scope, on top of its capabilities, is their support system. 3 year warranty over Tek's 1, and 7 year sustaining support after they EOL the line, and after that "as best as we can do." So again, many thanks for the information you provided (even though this particular scope is at their low-end of in-house products).

free_electron · « **Reply #38 on:** April 17, 2015, 08:26:19 pm »

Quote from: joeqsmith on April 12, 2015, 09:56:05 pm

I was walking with the 64xi and some of the knobs fell off.

All the newer LeCraps have this problem. i don;t know if it is the plastic of the knobs that becomes weak , or the shaft fo the encoders that shrinks. but every single LeCroy i worked with ( waverunnners, waveace's , 7zi's ) suffer from this problem...

joeqsmith · « **Reply #39 on:** April 17, 2015, 11:35:37 pm »

Quote from: free_electron on April 17, 2015, 08:26:19 pm

Quote from: joeqsmith on April 12, 2015, 09:56:05 pm
I was walking with the 64xi and some of the knobs fell off.

All the newer LeCraps have this problem. i don;t know if it is the plastic of the knobs that becomes weak , or the shaft fo the encoders that shrinks. but every single LeCroy i worked with ( waverunnners, waveace's , 7zi's ) suffer from this problem...

I don't think its the shafts. The new replacement knobs they gave me for it are very tight. On the two larger ones, to the point of needing to hold the back of the circuit board to press fit them into place.

BTW, I have been running the system every day from a cold start and not a single problem with the SSD. Continues to boot up and have waveforms in about a minute.

Gotta say, it's starting to grow on me.

joeqsmith · « **Reply #40 on:** May 08, 2015, 02:21:39 am »

Printing with the WaveRunner and WaveMaster is a snap. Press print and your done. It was not so simple with the 7200. You had to connect your printer/plotter to the Centronics port. Maybe an AB switch or two.

If you wanted to print to an Ethernet server, there were no options. I built this little reverse print server that connected to a standard printer port and then sent the data onto an Ethernet printer.

This used a Motorola MC6811 because I had them laying around still. I modified my old Pentica ICE to support an 8MHz clock rate. An Altera FPGA read data from one of the printer ports and DMA'ed it into DRAM. I wrote the stack in assembler.

It was fairly fast. Printing a screen from the 7200A at 75 DPI with the printer connected directly to the port requires 24 seconds. Over Ethernet it requires 28. At 150 DPI, a direct print was 30 seconds and 40 using Ethernet. Using the serial port at 150 DPI was 2 mins 3 seconds!

joeqsmith · « **Reply #41 on:** May 08, 2015, 02:51:38 am »

This is a block diagram for the reverse print server.

joeqsmith · « **Reply #42 on:** May 08, 2015, 03:24:24 am »

Printing with the two LeCroys to an Ethernet print server.

https://www.youtube.com/watch?v=l0GiC-dw9W8&feature=youtu.be

joeqsmith · « **Reply #43 on:** May 08, 2015, 10:48:49 pm »

No probes came with the 64Xi. I am looking for some general purpose, low cost ones and came across Hantek. I am interested in the T3100 and the PP200. Reviews on Amazon seem good.

http://www.hantek.com/en/ProductDetail_15_74.html
http://www.hantek.com/en/ProductDetail_15_73.html

Specs on the T3100 seem decent. 1KV up to 7MHz. 500V to 100MHz.

Also looking at the Probe Master 4906-1RA and the 4910-2RA. They cost more but they have no specs that I see.

http://www.probemaster.com/product_info.php?products_id=71

Would like to see a scope probe shootout.

Sailor · « **Reply #44 on:** May 09, 2015, 08:23:06 am »

Wow, good job

I haven't built a wire-wrap system for 30-35 years - well, actually I didn't build them, my wife did. I would give her lists of thousands of connections and she would work day and night with only one, maybe two errors per thousand. I couldn't do better than 1 in a 100, my mind would always be thinking about the next step. I've still got the OK mains-powered gun, and a bunch of sockets and Vero Hi-Density Eurocard boards somewhere. If you cross-gridded the power and ground, put plenty of bypasses, and ran clocks and other critical signals by twisted pair, you could easily get 20MHz operation.

And I'm glad to hear that there is someone else who still writes in assembler

joeqsmith · « **Reply #45 on:** May 12, 2015, 11:29:09 pm »

All of the low cost probes arrived today from Amazon. I had ordered two of the Hantek T3100s from here:

http://www.amazon.com/gp/product/B0030L0X58/ref=oh_aui_detailpage_o00_s00?ie=UTF8&psc=1

However, what I received I suspect are Siglent P4100s. What is funny is the package is marked Hantek. The Hantek T3100s are rated for their full 2KV up to 1MHz and 1KV at 7MHz. It is difficult to tell on the data sheet that came with the the P4100 but it appears to be good only to 40KHz at the full 2KV. At 1MHz they derate to 100 volts!

I had made up a test jig to try out the Hantek probe but it will have to wait.

I was not expecting much for a $25 probe except I did order their 100MHz probes as well. These were $17.50 for 2 of them. These actually came in as a Hantek part and wow! I am comparing it with my higher end probes and they are impressive.

tautech · « **Reply #46 on:** May 12, 2015, 11:50:04 pm »

Quote from: joeqsmith on May 12, 2015, 11:29:09 pm

However, what I received I suspect are Siglent P4100s.

I very much doubt they are.
What makes you think they are Siglent probes?

Just checked all my price lists and the don't list a 100:1 probe.
Even the website doesn't show one:
http://www.siglentamerica.com/prodcut-fjxx.aspx?fjid=355&id=27&tid=1&T=2

joeqsmith · « **Reply #47 on:** May 13, 2015, 01:34:43 am »

I just put P4100 into Google and got this for the first hit:

http://www.siglent.eu/p4100-100mhz-100x-high-voltage-probe.html

See attached if you can't view the site for some reason. The probe itself and the manual have no brand markings that I see.

joeqsmith · « **Reply #48 on:** May 13, 2015, 01:40:15 am »

If you go to this site, towards the bottom you will see the data sheet. This is very close to the data sheet that was included with these probes. The derate curve appears to be the same.

http://globaltone.en.alibaba.com/product/301181277-210286086/P4100_High_Voltage_Oscilloscope_Probe_2KV_probe_Oscilloscope_digital.html

tautech · « **Reply #49 on:** May 13, 2015, 03:43:58 am »

Quote from: joeqsmith on May 13, 2015, 01:34:43 am

I just put P4100 into Google and got this for the first hit:

http://www.siglent.eu/p4100-100mhz-100x-high-voltage-probe.html

This is not the first time Siglent EU has caused confusion, why?
From their website:

Quote

Siglent.eu is a trade name and part of JR Special Electronics.

So not all of their products are sourced from Siglent.
You'll also see they have P6100 probes available, also definitely not from Siglent.

joeqsmith · « **Reply #50 on:** May 13, 2015, 10:56:12 am »

More surprised that the seller would advertise a Hantek T3100 and ship some other part. I did get a response from Amazon.

"... I checked and found that you order was placed with "BEST.CHOICES", a seller on our website, because "BEST.CHOICES's" inventory is constantly changing, we can't replace items sold by them that are fulfilled by Amazon.... "

They will allow me to return them free of charge but I may just try one out for fun. Looking for the real Hantek T3100, it seems that most places show it out of stock. Can't see any point in placing another order just to get some random probes.

Still even with that little curve ball, I am very happy with the Hantek 10X probes I did receive. They seem to work fine with the LeCroy 64Xi.

My video comparing the 100MHz Hantek PP150 with a 350 MHz LeCroy PP002 and a 300MHz Probe Master PM 4906A

joeqsmith · « **Reply #51 on:** May 14, 2015, 04:28:47 am »

Decided to go ahead and run a few tests on the P4100. If you ever wondered what one looks like inside, I'll put up a video in a day or so.

Wrote Hantek to see if they have a distributor with the T3100 in stock.

Somewhere in the mess is the 64Xi.

joeqsmith · « **Reply #52 on:** May 14, 2015, 04:35:23 pm »

In this video I evaluate the no name P4100 high voltage probe with the LeCroy 64Xi.

tautech · « **Reply #53 on:** May 14, 2015, 07:30:47 pm »

All the P**** probes are YPioneer brand.

joeqsmith · « **Reply #54 on:** May 17, 2015, 02:34:27 am »

Low voltage testing on the Reynolds Wrap probe.

joeqsmith · « **Reply #55 on:** May 20, 2015, 02:46:43 am »

Making two this time based on all of my prototypes. The cases are made from spent shell cases. Trimmed and drilled them the lathe and just soldered together. I plan to stay with the EC pogo for a tip. Much smaller parts this time around to get the over all size down. Plan to use a Teflon core this time.

tautech · « **Reply #56 on:** May 20, 2015, 03:41:34 am »

Quote from: joeqsmith on May 20, 2015, 02:46:43 am

Making two this time based on all of my prototypes. The cases are made from spent shell cases. Trimmed and drilled them the lathe and just soldered together. I plan to stay with the EC pogo for a tip. Much smaller parts this time around to get the over all size down. Plan to use a Teflon core this time.

Nice, will the calibre matter?

Rupunzell · « **Reply #57 on:** May 20, 2015, 07:39:44 am »

The Tek P6015 probe innards. The 100Meg resistor is about 75-100mm long x 8mm diameter. Note the tapered capacitor with two different sized rings on the resistor ends to allow higher break down voltage by increasing the effective gap. The gold plated connection becomes the probe tip. The red plastic ring is a freon level float.

Those puny resistor inside the China made HV probe is very telling about the probe design, construction, users and expectations of what these probes are. That probe IMO is not worth the $ spent and resources used to produce them.

Bernice

joeqsmith · « **Reply #58 on:** May 20, 2015, 10:57:36 am »

That P6015 picture is very helpful. It confirms what I am doing with the copper foil.

Cases are 223, 35 Remington, 25-06.

joeqsmith · « **Reply #59 on:** May 22, 2015, 04:50:51 am »

Made some Teflon inserts for the tip. EC pogo pins.

joeqsmith · « **Reply #60 on:** May 24, 2015, 05:50:35 am »

It's now been 2 months since installing the 1G RAM and Transcend SSD into the 64Xi. The DSO has been ran pretty much every day and I have not seen any problems with it. Even channel 3 continues to work fine. I should mention that on this system, I did not make any changes to the system settings for the SSD. I just mirrored the drive, installed it and used it. It may shorten the life of the drive, but so far everything works great. It just needs the protective cover.

The video shows the end result of attempting to make my own 2KV 100MHz probes. The attached picture shows them next to a few other home made probes to give you an idea on their size.

Wuerstchenhund · « **Reply #61 on:** May 24, 2015, 01:04:46 pm »

Quote from: joeqsmith on May 24, 2015, 05:50:35 am

It's now been 2 months since installing the 1G RAM and Transcend SSD into the 64Xi. The DSO has been ran pretty much every day and I have not seen any problems with it. Even channel 3 continues to work fine. I should mention that on this system, I did not make any changes to the system settings for the SSD. I just mirrored the drive, installed it and used it. It may shorten the life of the drive, but so far everything works great. It just needs the protective cover.

What model of Transcend SSD did you use? It seems Transcend has replaced the PSD320 (which was one of the drives that was suffering from corruption in the WRXi) with the PSD330, and according to the Transcend representative I talked to the PSD330 has improved firmware which handles PIO and UDMA33 modes much better.

Also, the WRXi wasn't the only case where the older 320s ran into problems.

I also wouldn't worry about SSD lifetime. Just leave some space unpartitioned (overprovisioning), and disable auto defrag in XP, and the SSD should work fine for many years to come.

nctnico · « **Reply #62 on:** May 24, 2015, 03:05:24 pm »

I would not be so sure about the lifetime of an SSD. The guaranteed data rentention time of many SSDs is 3 months when powered down. No this is not a typo or a bad joke; it is actually in the specs! I got a big

moment when I found out. From what I understand the cells are refreshed every now and then so it is better to leave an SSD on permanently.

Muxr · « **Reply #63 on:** May 24, 2015, 04:10:21 pm »

Quote from: nctnico on May 24, 2015, 03:05:24 pm

I would not be so sure about the lifetime of an SSD. The guaranteed data retention time of many SSDs is 3 months when powered down. No this is not a typo or a bad joke; it is actually in the specs! I got a big moment when I found out. From what I understand the cells are refreshed every now and then so it is better to leave an SSD on permanently.

There was also a lot of fear mongering about SSD data retention by the media for a short while. Data retention depends heavily on the temperature. Basically the higher the temperature while the SSD is on, and the lower the temperature while it's off the longer the data lasts.

http://www.anandtech.com/show/9248/the-truth-about-ssd-data-retention

In your normal conditions it is at least a year, for a full drive, less than full drives can retain data longer.

joeqsmith · « **Reply #64 on:** May 24, 2015, 05:57:17 pm »

Quote from: Wuerstchenhund on May 24, 2015, 01:04:46 pm

Quote from: joeqsmith on May 24, 2015, 05:50:35 am
It's now been 2 months since installing the 1G RAM and Transcend SSD into the 64Xi. The DSO has been ran pretty much every day and I have not seen any problems with it. Even channel 3 continues to work fine. I should mention that on this system, I did not make any changes to the system settings for the SSD. I just mirrored the drive, installed it and used it. It may shorten the life of the drive, but so far everything works great. It just needs the protective cover.

What model of Transcend SSD did you use? It seems Transcend has replaced the PSD320 (which was one of the drives that was suffering from corruption in the WRXi) with the PSD330, and according to the Transcend representative I talked to the PSD330 has improved firmware which handles PIO and UDMA33 modes much better.

Also, the WRXi wasn't the only case where the older 320s ran into problems.

I also wouldn't worry about SSD lifetime. Just leave some space unpartitioned (overprovisioning), and disable auto defrag in XP, and the SSD should work fine for many years to come.

This is the TS64GPSD330. Like the one in my 8500A, it also has built in wear leveling. I have left defrag enabled and partitioned the entire drive. Really, just mirrored it over and plugged it in.

I plan to keep the scope and will let you know when it starts to have problems. It was 2 months to the date when I posted, and I thought you would like to know how it was going as you had such a different experience.

On my 8500A, I also partitioned the entire drive but in that case I had made several setting changes to try and increase the drive's life.

joeqsmith · « **Reply #65 on:** June 27, 2015, 07:36:44 pm »

Month three, DSO in use daily still. Full drive was formatted, defrag active, no changes to the OS at all.

The 8500 has not had near the use but continues to work fine as well along with the G Ethernet interface installed.

joeqsmith · « **Reply #66 on:** July 23, 2015, 02:57:22 am »

If you watched my hand held meter testing, you know I took the home made shell case probes a little over 4KV before switching to my Tektronix P6013 probe. I am pleased with the results from the home made probes.

The P6013 is rated to 12KV and I ran it a little over 13KV when testing the Fluke 101 and 87V. This is the highest voltage probe I have at home. If I had the cash, I would pick up a Northstar but I think we paid around $3 for that last one. They are top quality products.

My plan is to attempt to construct a new probe for home use, similar to the shell case probes.

Looking at DC - 50MHz, 40KVDC/80KV (pulse at 50% duty) min with a max load of 200uA, 1000X, uses 1M input on DSO. I plan to stay with standard coax with no compensation at the DSO. With 80KV on the high end at 50% duty cycle, the life of the film resistors should be fairly long.

Quote from: joeqsmith on May 24, 2015, 05:50:35 am

The video shows the end result of attempting to make my own 2KV 100MHz probes. The attached picture shows them next to a few other home made probes to give you an idea on their size.

joeqsmith · « **Reply #67 on:** July 24, 2015, 01:55:18 am »

This probe will be a little harder to prototype. Bored the phenlic rod to accept the Super MOX parts. This will become the dielectric for the capacitor. The copper plates will solder to the brass end cap. There are a couple of Teflon inserts that I need to make and then can try out this part.

Hope to keep the size down to about the same as the 6013. May look a little strange...

joeqsmith · « **Reply #68 on:** July 25, 2015, 03:01:12 am »

Cleaned, soldered, doped and assembled the first parts. I was not sure about using the phenolic for a dielectric but I soldered some coax on it to try it out and it looks useable. I'll post some sort of video.

As shown, end to end is about 10".

Mark_O · « **Reply #69 on:** July 25, 2015, 04:45:50 am »

Joe, that's certainly some nice looking workmanship. I'll be interested in hearing how well it works, when you test it.

Mark_O · « **Reply #70 on:** July 25, 2015, 04:52:23 am »

Quote from: joeqsmith on July 23, 2015, 02:57:22 am

If I had the cash, I would pick up a Northstar but I think we paid around $3 for that last one. They are top quality products.

At 3 bucks a piece, I would have grabbed a bunch of them.

joeqsmith · « **Reply #71 on:** July 25, 2015, 05:30:50 am »

Quote from: Mark_O on July 25, 2015, 04:45:50 am

Joe, that's certainly some nice looking workmanship. I'll be interested in hearing how well it works, when you test it.

Thanks. Not holding my breath for this one.

Here is the video from today's initial tests.

joeqsmith · « **Reply #72 on:** July 26, 2015, 06:08:34 pm »

Pulled out the Reynolds Wrap, tape and cardboard to try some things out. Not planning to win a beauty contest with it....

joeqsmith · « **Reply #73 on:** July 27, 2015, 04:53:00 am »

Today is 4 months of the 64xi using the SSD with no problems. Flipped on the 8500A a few weeks ago and it would not boot. Not the SSD, but the CMOS battery. Should have changed it when I had it apart.

After playing with the cardboard and tape and getting some fair results with the home made probe, time to start working on the real parts.

Here you can see the Teflon insulator that will fit between the handle and main body. To the right is a ceramic tube that will replace the cardboard to form the shield.

joeqsmith · « **Reply #74 on:** July 27, 2015, 06:14:37 am »

The ceramic tube is wrapped in copper foil to form the shield. I mentioned I planned to insulate the shield. Last picture shows the cover. It's starting to look like something.

joeqsmith · « **Reply #75 on:** July 28, 2015, 09:49:34 pm »

Anyone who has ever made a wide band (DC to >1MHz), 10,000:1, >30KV probe and actually gotten it to work, my hats off to you!

What a nightmare of a project.

tautech · « **Reply #76 on:** July 28, 2015, 09:55:19 pm »

Quote from: joeqsmith on July 28, 2015, 09:49:34 pm

What a nightmare of a project.

And we were so sure you had it sorted.

Keep up the great work Joe.

Mark_O · « **Reply #77 on:** July 30, 2015, 12:19:58 am »

I'm interested in the development of the new probe, but fail to see what it has to do with the Waverunner.

I just wish this had been split off as a separate topic, when the discussion branched. Of course, since the guy doing the thread-diversion was the OP, I guess that's his right.

But no one interested in wide-band HV probe construction will know it is buried in a wr64xi thread.

And others interested in the Waverunner, may not be interested in the probe. So it doesn't seem to be serving either audience very well.

joeqsmith · « **Reply #78 on:** July 30, 2015, 02:32:46 am »

Finally, I have added the protection to the probe in case of a breakdown and dared to hook it to the DSO.... 2KV 100us per division. The transient that took out the 87V.

The neon sign transformer leaks too much use it with the grounded scope.

Mark_O · « **Reply #79 on:** July 30, 2015, 03:43:35 am »

Maybe it's just me, but I'm getting a 0 byte file in Tran1.jpg.

tautech · « **Reply #80 on:** July 30, 2015, 04:03:51 am »

Quote from: joeqsmith on July 30, 2015, 02:32:46 am

Finally, I have added the protection to the probe in case of a breakdown and dared to hook it to the DSO.... 2KV 100us per division. The transient that took out the 87V.

The Red Devil is back.

joeqsmith · « **Reply #81 on:** July 30, 2015, 04:07:48 am »

Quote from: tautech on July 30, 2015, 04:03:51 am

Quote from: joeqsmith on July 30, 2015, 02:32:46 am
Finally, I have added the protection to the probe in case of a breakdown and dared to hook it to the DSO.... 2KV 100us per division. The transient that took out the 87V.
The Red Devil is back.

Too bad Fluke doesn't have a friend for her to play with

joeqsmith · « **Reply #82 on:** August 05, 2015, 08:45:30 am »

I've made some progress on my home made 10,000X high voltage probe. This video shows how the probe is compensated. It's then compared with my Tektronix P6013A. Like the Northstar probes, all of the compensation networks are built into the probe. I ended up shortening the coax cable to about 16 feet. Also shown is some more details on the construction.

During the video you may notice the MOVs and GDTs. There is also a faster set of clamps as well to protect the DSO in case the probe were to arc over.

https://www.youtube.com/watch?v=2SZyn2IMn9g&feature=youtu.be

joeqsmith · « **Reply #83 on:** August 06, 2015, 06:30:43 pm »

Waiting for the camera to charge up.....

The last test I want to run on the probe is to check the breakdown voltage. So far for high voltage testing I have had it connected to the homemade transient generator and the neon sign transformer. These don't put out any where near the what the probe should handle.

A friend of mine who is a mechanic gave me a car ignition coil that has an isolated output. He claims this is a hotter version of the coil. I am not sure how much we can expect to get out of it. I plan to put the probe on the output and just drive the thing into a spark gap and increase the gap until the probe breaks down or the coil does.

I'll sketch up a little schematic for it and go over that as well for those interested.

tautech · « **Reply #84 on:** August 06, 2015, 08:39:38 pm »

Quote from: joeqsmith on August 06, 2015, 06:30:43 pm

I'll sketch up a little schematic for it and go over that as well for those interested.

I've got my hand up.

joeqsmith · « **Reply #85 on:** August 07, 2015, 07:21:35 am »

Played with spark gaps and corona dope most of the day. The LeCroy still lives but it sure didn't like what I was putting it through today. When your scope's LCD flickers and the mouse won't stay running, your doing something right.

The video starts out comparing some basic schematics for probes then I go over the design for this probe. After this, I show the car ignition coil setup which is very crude and then proceed to attempt to damage the probe with it.

Keep in mind that the stub is critical just like it was with shell case probes. I did not provide values in the schematics as I assume anyone attempting to replicate it would need to come up with their own compensation.

tautech · « **Reply #86 on:** August 07, 2015, 02:11:39 pm »

Thanks Joe, very entertaining.

65 KV is a great effort, but being able to safely measure it, mind blowing.

joeqsmith · « **Reply #87 on:** August 07, 2015, 03:18:33 pm »

Quote from: tautech on August 07, 2015, 02:11:39 pm

Thanks Joe, very entertaining.
65 KV is a great effort, but being able to safely measure it, mind blowing.

Thanks! Glad you enjoyed them.

The ignition coils don't produce enough energy that I get too concerned when using them. The transorb I am currently using clamps around 7 volts, limiting the probe to 70KV. The old LeCroy was having problems and I was already near the 70KV limit so I stopped.

Thinking about the handle, I am considering making it all from metal and having a bell similar to the North Star PVM-3.

http://www.highvoltageprobes.com/

Currently everything is out in the open and even without a shield over this area it still works pretty good. It may end up being plastic. Need to give it some more thought.

joeqsmith · « **Reply #88 on:** August 10, 2015, 02:50:44 am »

After some thought, here's what I came up with...

I stole the ground strap idea from the Tektronix P6013A and machined a slot for a small press contact that slides in the outside insulator. This keeps the return path away from handle side.

Next is the new handle next to the prototype one. The handle is held on with a large nut that threads on the back of the BNC. It's only ground is to the BNC.

Next is the Teflon insulator that holds the handle and the compensation network.

joeqsmith · « **Reply #89 on:** August 10, 2015, 03:12:01 am »

Here is the finished probe. The dot in the insulator is one of three set screws that holds the handle in place. The LeCroy is showing a PP002 probe (red) against the 10,000X probe (yellow) looking at the rising edge from my HP33120A ARB at 20nS/div. I also ran some tests with a high voltage DC power supply today using the Fluke 101 (just in case something went wrong).

joeqsmith · « **Reply #90 on:** August 12, 2015, 04:00:05 am »

The last video I made showing the completed high voltage probe on the low frequency network analyzer and then compared with a LeCroy PP002A 350MHz 10X probe. I finish up with the probe on a high voltage Spellman supply.

The parts to build it were in the order of $100.

joeqsmith · « **Reply #91 on:** November 26, 2015, 04:46:00 pm »

I was cleaning out some of my crap and noticed a transformer from Austin transformers. It's in a cylindrical metal can, sealed with solder and ceramic insulators. It was rated for 50/60Hz and is made for isolation. Playing around with it and it seems to have a BW in the 200KHz range.

No idea what this thing was for but thinking to use it to make some sort of AC line power monitor probe for the old LeCroy. Really don't have much of a need for something like this at home but seems like a simple project. Noticed Teledyne/LeCroy offers some sort of package but I have played with it.

http://cdn.teledynelecroy.com/files/pdf/power_analyzer_datasheet.pdf

My plan is to use this isolation transformer and a LEM sensor (also good to about 200KHz) and just post post process the data with a PC.

rsjsouza · « **Reply #92 on:** November 27, 2015, 10:33:56 am »

Joe, excellent job on this probe. I was wondering about the voltage spikes messing with the oscilloscope and seeing your Picture I thought about a thread that mentioned BNC caps (or dust covers). The ones you are using seem to be yellow plastic, but perhaps an all-metal one that physically grounds the input may be worth trying, what do you think?

The post that references this is below:
https://www.eevblog.com/forum/testgear/siglent-they-filed-a-'wrongful-trademark-claim'/msg786287/#msg786287

Also, someone else sent a link to an eBay listing for a more reasonable quantity:
https://www.eevblog.com/forum/testgear/siglent-they-filed-a-'wrongful-trademark-claim'/msg786701/#msg786701

(edit: fixed link)

joeqsmith · « **Reply #93 on:** November 27, 2015, 04:47:00 pm »

Thanks. I have been pretty happy with all the home made probes.

The Caplugs covers are there to prevent mechanical damage to the connectors. Like keeping my grubby oily hands off them.

I would imagine that most scopes inputs are fairly protected electrically. Maybe not the bottom basement ones for cost reasons.

Those ignition systems can sure make a lot of noise. I would guess my PC was about 6 feet from that jig and the USB mouse on the PC was hung after test and needed to be power cycled. This is pretty common problem with USB. It's not a robust bus. I suspect on the LeCroy that the noise was causing the LCD interface to get corrupt. Beyond the LCD and the mouse shutting down, there were no other problems.

joeqsmith · « **Reply #94 on:** November 28, 2015, 06:09:56 am »

The completed power monitor adapter. The CB has a 10A trip current and the LEM is set for about 8 Amps. Software is mostly done. Calculates the basic stuff.

I'll put together some sort of demo for it.

rsjsouza · « **Reply #95 on:** November 28, 2015, 11:53:16 am »

Quote from: joeqsmith on November 27, 2015, 04:47:00 pm

Thanks. I have been pretty happy with all the home made probes.

The Caplugs covers are there to prevent mechanical damage to the connectors. Like keeping my grubby oily hands off them. I would imagine that most scopes inputs are fairly protected electrically. Maybe not the bottom basement ones for cost reasons.

Those ignition systems can sure make a lot of noise. I would guess my PC was about 6 feet from that jig and the USB mouse on the PC was hung after test and needed to be power cycled. This is pretty common problem with USB. It's not a robust bus. I suspect on the LeCroy that the noise was causing the LCD interface to get corrupt. Beyond the LCD and the mouse shutting down, there were no other problems.

I see. Having blown or crippled inputs in the past, I tend to go a bit overboard in input protection/shielding.

The LCD is the largest non-shielded part of the oscilloscope, therefore it is probably the most affected by EMI surges. Under certain circumstances I see this happening in my Rigol DS4014 as well.

I think the USB (and perhaps a keyboard) are your PC's "antennas" that capture all EMI spikes. Tie these with thinner shielded cables (the norm these days) and you probably get the most susceptible part of your PC.

joeqsmith · « **Reply #96 on:** November 28, 2015, 05:32:30 pm »

Quote from: rsjsouza on November 28, 2015, 11:53:16 am

I see. Having blown or crippled inputs in the past, I tend to go a bit overboard in input protection/shielding. The LCD is the largest non-shielded part of the oscilloscope, therefore it is probably the most affected by EMI surges. Under certain circumstances I see this happening in my Rigol DS4014 as well.

I think the USB (and perhaps a keyboard) are your PC's "antennas" that capture all EMI spikes. Tie these with thinner shielded cables (the norm these days) and you probably get the most susceptible part of your PC.

There's nothing wrong with being overly safe. You may have seen my spark gaps and diode clamps that I use from time to time when I play with high voltages. I have been lucky that I have never damaged the front end of a scope. Not like I haven't put them at risk!

My old LeCroy front end actually has a battery in it that is used for part of the input protection when the scope is powered down. The WR will open the connector and ground the ADC, which is more what I would expect .

They make a transparent film for EMI that can be placed over something like an LCD. I wonder if the more modern industrial displays don't have some sort of layer on the glass.

Line cord and Ethernet cables as well. The unshielded Ethernet is fairly robust. That USB with it's common mode problems, not so much. Most of the USB devices I have will reset or hang when I play with the ignitions.

joeqsmith · « **Reply #97 on:** November 29, 2015, 07:08:06 am »

I put together a video showing the power measurement software with some different loads. Labview sets up the LeCroy and then downloads several cycles. I used the Brymen to calibrate the current and voltage. It would be fun to try a commercial product and see how they compare with some different waveforms.

joeqsmith · « **Reply #98 on:** December 21, 2015, 05:48:27 am »

After making the first power measurement video, I posted here about my attempting to improve the measurement.

https://www.eevblog.com/forum/testgear/making-accurate-low-frequency-measurements-with-your-scope/30/

In this video I put the oversampling into practice.

joeqsmith · « **Reply #99 on:** January 11, 2016, 04:50:34 am »

I picked up one of those in-line AC power meters and compared it with the DSO. It seems to work pretty good. Keep in mind that the meter is plugged in downstream from the DSO's monitor. So the DSO is reading both the load and the power meter.

joeqsmith · « **Reply #100 on:** June 20, 2018, 10:35:01 am »

Quote from: joeqsmith on April 12, 2015, 10:31:53 pm

Put the new SSD into the system and turned it on. Boots fast.

The long term SSD test continues. For something that was to be problematic, we are now well past the three year mark with the one SSD in the 64xi and a whopping four years on the 8500. I have yet to see a file get corrupt. Just a recap for the 64xi, I made no changes to the OS to try an increase the life of the drive. Just mirrored the drive over and called it a day. This is still my main scope.

Also, the knobs are still tight! lol.

rsjsouza · « **Reply #101 on:** June 20, 2018, 12:49:19 pm »

That is a great report, Joe. These SSDs are surely smart so they don't wear out that easily. Although I would inagine that a regular PC would be much more data intensive than the oscilloscope, thus being the worst case scenario for these drives.

joeqsmith · « **Reply #102 on:** June 20, 2018, 04:47:20 pm »

Good point. I installed a SSD in my laptop maybe 5 years ago but it's not the primary drive. When I finally retired my XP based P4 machine in 2016, I used a 200G M.2 drive as the primary in the new PC. It's too early to say but so far I have not seen any problems with it.

rsjsouza · « **Reply #103 on:** June 20, 2018, 09:05:21 pm »

I have five year-old Samsung 840 SSDs running on work laptops that do not show signs of wear - even with intensive embedded Linux and u-boot projects being built on them quite regularly for about three out of the five years.

Not SSDs, but I love the statistics put by these guys at:
https://www.backblaze.com/blog/hard-drive-stats-for-2017/

joeqsmith · « **Reply #104 on:** June 20, 2018, 11:22:03 pm »

Quote from: rsjsouza on June 20, 2018, 09:05:21 pm

I have five year-old Samsung 840 SSDs running on work laptops that do not show signs of wear - even with intensive embedded Linux and u-boot projects being built on them quite regularly for about three out of the five years.

Not SSDs, but I love the statistics put by these guys at:
https://www.backblaze.com/blog/hard-drive-stats-for-2017/

I am not aware of anyone I know in person running SSDs that has had problems. They only seem to exist on the internet. Larger sampling or typical internet regurgitating, hard to say. I'm sure I am setting myself up for a massive SSD failure at home.

rsjsouza · « **Reply #105 on:** June 21, 2018, 02:45:05 am »

Quote from: joeqsmith on June 20, 2018, 11:22:03 pm

Quote from: rsjsouza on June 20, 2018, 09:05:21 pm
I have five year-old Samsung 840 SSDs running on work laptops that do not show signs of wear - even with intensive embedded Linux and u-boot projects being built on them quite regularly for about three out of the five years.

Not SSDs, but I love the statistics put by these guys at:
https://www.backblaze.com/blog/hard-drive-stats-for-2017/

I am not aware of anyone I know in person running SSDs that has had problems. They only seem to exist on the internet. Larger sampling or typical internet regurgitating, hard to say. I'm sure I am setting myself up for a massive SSD failure at home.

Well, if it's on the Internet it must be true, right?

I have the same feel sometimes... When things go too well... Murphy is at the corner waiting to strike!

joeqsmith · « **Reply #106 on:** June 26, 2018, 12:43:46 am »

Quote from: rsjsouza on June 21, 2018, 02:45:05 am

Quote from: joeqsmith on June 20, 2018, 11:22:03 pm
Quote from: rsjsouza on June 20, 2018, 09:05:21 pm
I have five year-old Samsung 840 SSDs running on work laptops that do not show signs of wear - even with intensive embedded Linux and u-boot projects being built on them quite regularly for about three out of the five years.

Not SSDs, but I love the statistics put by these guys at:
https://www.backblaze.com/blog/hard-drive-stats-for-2017/

I am not aware of anyone I know in person running SSDs that has had problems. They only seem to exist on the internet. Larger sampling or typical internet regurgitating, hard to say. I'm sure I am setting myself up for a massive SSD failure at home.
Well, if it's on the Internet it must be true, right?

I have the same feel sometimes... When things go too well... Murphy is at the corner waiting to strike!

I think sometimes people repost so much garbage that eventually everyone starts to believe it, only because of the shier number of posts.

Because I have posted so many of my home made probes in this thread, I may as well continue....

tautech · « **Reply #107 on:** June 26, 2018, 01:10:03 am »

Just watched the diff probe vid Joe, great work and thanks for sharing.

joeqsmith · « **Reply #108 on:** June 26, 2018, 12:16:02 pm »

Glad you enjoyed it. If you do a search on this site, there are a few threads where people have attempted to make their own higher frequency higher voltage differential probes.

During the video, I mentioned used two circuits. One for the high frequency, the other for the DC and mixing the two. This is not a new idea and may have merit for this design.

bson · « **Reply #109 on:** July 01, 2018, 12:42:34 pm »

Quote from: joeqsmith on June 20, 2018, 11:22:03 pm

I am not aware of anyone I know in person running SSDs that has had problems. They only seem to exist on the internet. Larger sampling or typical internet regurgitating, hard to say.

It's just people who don't understand what the data retention spec refers to; it's the guaranteed minimum retention, at the end of a device's useful life. It the retention guaranteed when it starts issuing SMART warnings.

https://www.micron.com/about/blogs/2015/may/addressing-data-retention-in-ssds

However, data retention specs are not given for a new device. It is very important to note that published specifications for data retention describe the behavior of the device only at the end of its prescribed service life.

The JEDEC specification for data retention tells us that for enterprise storage devices, data retention at the end of the service life shall be at least three months (stored at 40°C). For SSDs in the client computing market, data retention shall be at least one year after the drive’s service life (assuming it’s stored at 30°C).

joeqsmith · « **Reply #110 on:** July 01, 2018, 01:09:38 pm »

Quote from: bson on July 01, 2018, 12:42:34 pm

Quote from: joeqsmith on June 20, 2018, 11:22:03 pm
I am not aware of anyone I know in person running SSDs that has had problems. They only seem to exist on the internet. Larger sampling or typical internet regurgitating, hard to say.
It's just people who don't understand what the data retention spec refers to; it's the guaranteed minimum retention, at the end of a device's useful life. It the retention guaranteed when it starts issuing SMART warnings.

Could be. If you take the time to read some of the early comments, one person was talking about a couple of weeks was all they were seeing. But again, it's the internet.

joeqsmith · « **Reply #111 on:** July 05, 2018, 09:27:54 pm »

The parts for my next version of a differential probe arrived. However, I ran into a small setback and it will be a while until I can try building something up. My plan is to still go ahead with the project in the next month or so. Sorry for the delay.

rsjsouza · « **Reply #112 on:** July 05, 2018, 09:37:47 pm »

You disappoint me, Joe. You talk about minor setbacks as if your house was struck by lightning...
https://www.eevblog.com/forum/testgear/hear-kitty-kitty-kitty-nope-not-that-kind-of-cat/msg1646627/#msg1646627

joeqsmith · « **Reply #113 on:** July 18, 2018, 11:31:39 pm »

It's been a mini marathon here the last few weeks. I knew this new probe would not be very good starting out. Its mostly made from parts I salvaged, which made for some poor choices in the design. I assume the $130 MICSIG diff probe would out perform it in any test.

https://youtu.be/0thOfk4I3qs

joeqsmith · « **Reply #114 on:** June 12, 2020, 06:52:43 pm »

Necroposting my own thread....

Someone was asking about this old scope so I dug up this thread and thought, it's been five years now that the SSD has been running in both the 64Xi and the 8500A. Time for an update.

I don't use the 8500A often but the 64Xi continues to be my primary scope and sees a fair amount of use. I have not had a single problem with the two drives.

I recently went to use the 8500A and it came up with some musical song and an error about how it had overheated the last time it was used. I was concerned that maybe the SSD was the problem but it was the battery for the NVRAM was all. Installed a new battery, reset the BIOS and it was back in business.

I picked up an old VNA from around this same era and ended up buying a couple more of the Transcend drives for it as well. That PC is running Windows 2000, compared with XP in the two LeCroys. It's been in service for a few months now and is seeing a fair amount of use as well with no problems. I wonder if these drives are starting to become harder to find so I bought a few of them. They were shipped direct from the factory.

On the 8500A, I had installed a Gb Ethernet card which also seemed to raise some questions if it was going to cause problems. After five years, again it seems to be fine.

Quote from: Wuerstchenhund on April 13, 2015, 04:28:46 pm

Quote from: joeqsmith on April 12, 2015, 10:29:51 pm
Had replaced the drive on the 8500A with an SSD. It's been working very well. I had bought two different ones to try. One SATA one PATA. The SATA with adapter went in the 8500A, the PATA is going into this system. Different brand, we will see.

Just be careful with SSDs in that scope, as they tend to suffer from file system corruption (I tried several ones, including the Transcend PATA SSD on your picture), and all suffered from the same problem.

The reason seems to be that the SSDs only really support UDMA modes, which for PATA requires a 80 conductor cable to work reliably. The scope however uses a 44pin cable (there are no 84 conductor cables, at least I'm not aware of them) which is not really suited for UDMA66 or faster modes. Another forum user (Tunersandwich) who has the same scope made the same experience, he also tried a SSHD but that was a no-go either.

In the end I went back to spinning rust, i.e. a modern fast SATA laptop drive connected through a cheap SATA-PATA bridge. Works fine and absolutely reliable.

Quote
$7350 for the Rigol DS6064 or $9660 for the DS6104. I assume this is the market LeCroy is going after.

Not really. The WaveRunner Series starts quite a bit above that.

The WaveSurfer 3000 is the one that competes with the Rigol DS6000 scopes, although no-one really buys them anyways due to their poor value for money.

gbenyhe · « **Reply #115 on:** June 13, 2020, 10:27:11 am »

Hi Joe,
Thanks for your pics and all, they are incredibly useful for me during the refurbishment of my own 64Xi. I am just repairing the broken standoffs on the front cover's backside. First I gently fix the broken parts together by carefully melting them with a soldering iron, then give it a prosthetic machined from textile bakelite glued on it with a two components resin.

joeqsmith · « **Reply #116 on:** June 13, 2020, 04:02:40 pm »

Quote from: gbenyhe on June 13, 2020, 10:27:11 am

Hi Joe,
Thanks for your pics and all, they are incredibly useful for me during the refurbishment of my own 64Xi. I am just repairing the broken standoffs on the front cover's backside. First I gently fix the broken parts together by carefully melting them with a soldering iron, then give it a prosthetic machined from textile bakelite glued on it with a two components resin.

I highly suspect the same standoffs on mine are also broken as the whole front cover is dangling. The thin strip of plastic along the outside edge continues to brake away. I have thought about rebuilding my cover with fiberglass and filler but the scope has other mechanical problems as well. I wonder if their new products are this bad. Wife may not be to happy with me if I buy another scope.

joeqsmith · « **Reply #117 on:** October 03, 2021, 07:56:14 pm »

Today's job, the CMOS battery died on the old scope. To change it, you have pull the whole scope down.

The old SSD is still working fine after six years now. Not bad for something that I was told would not hold up.

Decided to build up the two corners of the cheap plastic case. They did not use a thermoset, so it was easy enough to melt in new gray plastic, a quick sanding and some paint. Most of the bosses had also failed that hold the front bezel in place, I glued all of them and built up the area around them. Then drilled new holes and used longer screws.


EEVblog Main Site	EEVblog on Youtube	EEVblog on Twitter	EEVblog on Facebook	EEVblog on Odysee