FLUKE 190_204 Scopemeter. Data on USB seem to be lost 1

[Skogsgurra]

Being the nerd I am, I couldn't resist buying the new FLUKE 200 MHz, all insulated four channel colour scope.

I like most of it. Battery life is good - a full day's work on an airport and still some battery left at the end of the day. But the plastic BNC connectors are delicate like they have been for decades. And internal memory is a mere 15 recordings 'deep'.

Not to worry, I thought. There's always the USB port. Just save data on USB sticks and use Flukeview 5 to get it back to the computer and into your report. good thinking - but no luck, I could not get the data back.

Yes, I could get the data back to internal memory in the scope. But not able to transfer that data from internal memory to computer. The 'native' data (data that were stored directly in internal memory) could be tranferred with no problems, but not data that had been stored to USB and then taken back to INT.

I prize my luck that I did save the more important measurements in INT and not USB. All I can do now with the USB data is to get them back as BMP. Coarse and not at all what my customer expected - and not me either.

Anyone knows what to do? Is there a magic trick so I can make USB data behave as well as the internally stored data?



Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[electricpete]

2. I think it is fair to say that I am surprised that Fluke still uses an antique way of doing thing

Is that based on the subject of this thread.... your perception that the Fluke does not save numeric data (which I believe has been refuted) or something else?

=====================================
(2B)+(2B)' ?

 

[electricpete]

By the way, my latest question is just in interest of clarity of communication....I honestly don't see evidence that you have ever read the link that I posted. I apologize for suggesting your motive was sales and I'm sure there is a great deal we all can learn about scopes and recorders from you.

=====================================
(2B)+(2B)' ?

 

[electricpete]

I honestly don't see evidence that you have ever read the link that I posted

OK, I see now that you are not happy with the resolution in ScopeRecord mode. So you've read the link, played with the device, and it doesn't meet your needs. Fair enough.

=====================================
(2B)+(2B)' ?

 

[electricpete]

So, the remaining objection is the resolution of the stored data, stated as 200usec.

But apparently it is a litle better than simple samples spaced as 200 usec.

There is also this:

A: Using the scope in recorder mode, the total time span of a recording is related to the time resolution of the recording, just like it is with a paper recorder. Thanks to the high sample rate, sudden changes such as glitches will still be captured.

The table shows:
Time resolution of recording: 200 usec
Will show glitches as narrow as: 50 nanoseconds.

I have to admit it is not 100% clear what that means. Apparently it is an attempt to provide better resolution for suddent changes without wasting storage by recording the entire trace at the higher esolution

=====================================
(2B)+(2B)' ?

 

[electricpete]

Sory, I see you already referred to this same glitch feature 8 May 11 1:30.

=====================================
(2B)+(2B)' ?

 

[GTstartup]

Gunnar I am interested in your comment about the PicoScope (damage to inputs with transients > 20V) because I have one too. Would always recording with the 10X probe setting and readjusting the scaling resolve this?

 

[Skogsgurra]

Yes, definitely. Use a 10:1 probe to protect the input.

My mistake was that I had a small HF transformer and a high-pass filter to measure noise. I didn't think that the small HF transformer was able to let enough energy through. But it obviously did.

I have used the same transformer with several other scopes and never had a problem. With the PicoScope, I have destroyed two inputs. Didn't understand what had happened the first time. So had to blow another to realize what was causing it.

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[electricpete]

I think if I had bought a product advertised as 200MHz bandwidth, 2.5 GSamples/sec sample rate, and I was unable to extract any waveform finer than 200 microsec spacing (5khz sample rate), I would also be surprised and disappointed.

Apparently their strategy was to process the heck out of the data and offer all varieties of processed data that they believe is more user-friendly and still somehow contains the vital info from the fast sampling. I'm not familiar with all their features (glitch detection etc) and even if these features do get back some of the data lost, I would certainly be annoyed at having to learn them, rather than just getting raw data which requires no extra learning and which can always be processed later.

=====================================
(2B)+(2B)' ?

 

[Skogsgurra]

Pete, your 8 May 11 12:59 post.

No, I meant to say that there are more direct, faster and reliable ways of connecting to a USB port than the antiquated COMxx: technique.

The simulated com ports are a nuisance and it is much better to talk directly to the USB hardware/software instead.

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[electricpete]

Thanks. So what does better USB performance do for us?
Would it be possible to "stream" the data straight to the usb without being constrained by the in-box storage limit?

=====================================
(2B)+(2B)' ?

 

[Skogsgurra]

Yes, but not fast enough for the 2.5 GS/s. The main advantage is that the connection is faster and works everytime. The simulated COMxx: does not always find a free com port and it is sometimes necessary to restart the PC to get a connection at all.

Direct communication with the hardware works everytime and is fast. We have used FTDI components, which makes communication easy. TiePie seems to use a custom solution and that works very well also. The COMxx: method is, in my opinion, a relict from the RS232 era when you had either that or the Centronics interface to chose between.

The main problem, that I have, is still that there is a very limited space for data (15 measurements) and that the USB stick, which I thought would be a good way of saving data, doesn't store anything but bitmaps of the screen.

The other things mentioned in this thread are 'extra nuisances' that do not make me any happier.

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[Tekfan]

I have recently bought a 190-104.
I can store more that 1000 waveforms when I use the USB as extra memory.
No problems with the communication. I had to remove the USB drive when I connect to my PC as told by the manual.
I think you have to send in the unit for service.

 

[Skogsgurra]

Could you store waveforms or BMP? I have no problems with BMP. But I need to store waveforms if want to import to FlukeView5 and do measurements there. That is important for my applications.

I am talking to Fluke directly. They have not suggested that the instrument needs service.

Question: Why would a Tek fan use Fluke?

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[Tekfan]

Gunnar,
I save the waveform on the USB drive and with the file options menu I can retrieve them and download them into FlukeView (demoversion).
Maybe you are wondering why i have bought a Fluke?
I was using the Tek THS720 but that unit last ony a few years in a rough environment. The Fluke is more rugged and can be used upto 1000V cat III. I used the ScopeMetre for motor speed drives and the output voltage can be very high.
Safety is what counts for me.
I like the View&Connect and Replay features which makes my work as maintenance engineer much easy.

Guenther

 

[itsmoked]

Gunnar did you look at this scope? It may do everything you want.

http://www2.tek.com/cmswpt/psdetails.lotr?ct=PS&cs=psu&ci=13304&lc=EN

Keith Cress
kcress -

http://www.flaminsystems.com

 

[LionelHutz]

We have an old (10 years or so) 43B meter and it does that save the screen as a bitmap to record, which is next to useless. Interesting to hear they haven't managed to improve on that in 10 years.

 

[ScottyUK]

Keith,

I had a long good look at that range of scopes three or four years ago. It's a reasonable scope but its capability is now lagging far behind Tek's mains-powered scopes, especially in memory capacity and communication options (no ethernet or USB), which it didn't when it was first introduced. I asked Tek if they had any plans to update it, but none were forthcoming. It's a shame because it is basically a decent scope but has been left behind in recent years. Had I not changed jobs I'd likely have bought a mains-powered Agilent scope with some decent isolation probes, giving better overall capbility and increased protection for the scope: better to blow an isolation amp than the scope. There aren't many places where I want a scope and can't get either a mains supply or generator given a little effort.


----------------------------------

If we learn from our mistakes I'm getting a great education!

 

[itsmoked]

Ah.. Nice to know. Thanks Scotty. I have the same problem in spades with my 720(?) Which I'm now presuming was replaced by the TPS2012B.

Seems Tek is just lusting after all those damn mega-scopes. At least once a month I see ads and product briefs on scopes costing more than $50k. The other day I saw one that listed for $230k!

Keith Cress
kcress -

http://www.flaminsystems.com

 

[Skogsgurra]

Keith! Don't buy that $230k scope without checking it out first!

Günther, I have tried to transfer all those waveforms stored on the USB stick to Flukeview5 (licensed) and the only way it works is to first manually download each waveform to INT memory (as you know, that takes quite some time for each waveform). The maximum number of waveforms that can be stored in INT memory is 15. No more. So, if I have a modest day's work, say a hundred or so, waveforms and want to transfer them to FV5/PC, I have to manually copy each waveform from USB to INT until I have a set of 15 waveforms available in INT. Then I *can* (if I am lucky) transfer those waveforms - one by one - to FV5 and then store them on disk in the PC.

Usually, the USB stick gets autistic when the USB cable is connected. So I have to disconnect the cable to be able to read the USB. Then, the FV5 doesn't notice that the cable has been pulled and if I try to transfer the next set of 15 waveforms, it starts reading setups and God knows what. It 'reads' happily and after six or seven minutes, it sys that it is rerady to transfer waveforms. Only problem is that it has no information on what waveforms are available. And it still doesn't hace a clue that the cable was pulled.

The situation is aggravated by the fact that there is no information stored with the waveform. No date, no time of day. The scopemeter 'recycles' waveform names. I can have several 'scope1', 'scope2' etcetera waveforms after one day's work. It is extremely awkward to try and keep trace of which is which. Date and time would have helped immensely and is something you get on every simple camera today. But not on this scope.

I can not use a scope like that when doing field work. It is OK for trouble-shooting, yes. But if your task is to verify bearing currents, motor voltage transients on a paper machine or a steel mill where you have hundreds and hundreds of readings that you need to store for reports and future reference, it is not possible.

There is a deeper discussion going on in a 'parallel universe' and we are a bit more outspoken there. All problems haven't been mentioned here. This thread started not as a whining thread but as a way to solve a problem that I got when writing a report during the week-end.

All other problems have popped up after that and I intend to keep that discussion in the 'parallel' and more closed 'universe'.

BTW. Your name is German/Dutch. Your excellent command of English points to Dutch - the little mistake with 'much easy' instead of 'very easy' makes me believe firmly that you are actually Dutch. Have seen that before when talking to Dutch guys. It is kind of a giveaway.

Then, the natural question to ask is: Do you have any connection with the Fluke people?


Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[Skogsgurra]

OK, while I am at it.

The 'problem' with the maximum voltage between channel references (cold side of the probe) and channel reference to ground may be a simple mistake in the derating diagram. If the horizontal scale is in MHz instead of kHz, it makes much more sense. See attachment.

I find it totally unnecessary that such simple errors shall caus a discussion like what we have had here and also that I have had with the guys in Holland. A customer should not be the first to spot it. Especially not when the question has been put directly to top officers at the headquarters.

As several posters have noticed before - the company doesn't give the competent and solid impression it used to any more. I feel sorry for that.

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.