MDO3000 have some hardware bug in large signal test, waveform update at 7 floor

[cody198879]

I come form china. My english is no good. But I just want to share my experience to use TEK's MDO3000.

Our company just buy 4 MDO3014. They look so cool with many LED on the panel. But it is the start of many trouble

The probe TPP0250 is a new type probe and it is have no a general probe compensation. It is compensation by the soft. Most time, it seems good. But if you try to test some PWM signal, like your power IC's LX pin PWM, you can't get a true result. The PWM's waveforms top and bottom will distort. You can see it  when you zooming in 100 or 200 mV.  This bug no on TDO3000 series.

MDO3000 series have another bug on input amplifier. The input amplifier commonmode output just small than 3.2V. It means that if you test a 20V PWM and set the fine scale to 500mV(Use 10X probe), you can see the waveforms of the PWM's bottom no on 0V. It will offset by the input voltage.   

But all this bug no on TDO3000 series. But at this time TDO 3000 series is off production. We ask TEK to return this MDO3000 but they no answer. We also ask TEK to change the proble TPP0250 to P6139B(use on TDO3000 series), they also no allow. We ask TEK engineer about how to fix this bug. They say:" This is no a bug, 95% costomers  don't use like this."  What the fuck TEK. 
 
So this 4 MDO3014 just put on in one corner of our lab. 3000 dollar just like a rubbish. No engineer want to use them because we don't know how many bug they have!

[cody198879]

By the way, our supplier says that this bug is on all new MDO 2000/3000/4000. They are the same.

[tautech]

Welcome to the forum Cody.
Thank you for sharing this with us.
There may be some other owners of these Tek models that can confirm your reported bug.

[w2aew]

The probe TPP0250 is a new type probe and it is have no a general probe compensation. It is compensation by the soft. Most time, it seems good. But if you try to test some PWM signal, like your power IC's LX pin PWM, you can't get a true result. The PWM's waveforms top and bottom will distort. You can see it  when you zooming in 100 or 200 mV.  This bug no on TDO3000 series.

Note that you need to compensate EACH probe one EACH channel.  You only have to do this once.  When you've done this, the scope remembers the compensation factors for each probe when you connect it.  The compensation coefficients are stored in the scope, not in the probe - that's why you do the compensation of each probe on each channel.  Each probe has a unique serial number which is read by the scope when it is plugged in, and it uses that s/n to recall the compensation parameters for that probe/channel combination.

[rx8pilot]

Clever features cause chaos if not known to the end user. Maybe not a bug, but a learning curve?

[cody198879]

The probe TPP0250 is a new type probe and it is have no a general probe compensation. It is compensation by the soft. Most time, it seems good. But if you try to test some PWM signal, like your power IC's LX pin PWM, you can't get a true result. The PWM's waveforms top and bottom will distort. You can see it  when you zooming in 100 or 200 mV.  This bug no on TDO3000 series.

Note that you need to compensate EACH probe one EACH channel.  You only have to do this once.  When you've done this, the scope remembers the compensation factors for each probe when you connect it.  The compensation coefficients are stored in the scope, not in the probe - that's why you do the compensation of each probe on each channel.  Each probe has a unique serial number which is read by the scope when it is plugged in, and it uses that s/n to recall the compensation parameters for that probe/channel combination.

I know that. I compensate every channel when i first use it. This bug is true because TEk's engineer go to our company to try to fix it. He spend one hour to  compensate new probe but no useful. And he just say  he must report this to USA develop team but can't do no thing.

[cody198879]

Clever features cause chaos if not known to the end user. Maybe not a bug, but a learning curve?

I want this is no any bug, MDO3000 is a new series. But this bug I write before is test and verify with DPO3000.  MDO3000 is more cheaper and TEK reduce some analog hardware.

[tekman11]

By the way, our supplier says that this bug is on all new MDO 2000/3000/4000. They are the same.

can you post a few pictures of the problem

[cody198879]

By the way, our supplier says that this bug is on all new MDO 2000/3000/4000. They are the same.

can you post a few pictures of the problem

OK,I post some pictures.

Test 1

This is a buck IC's LX pin waveform. Vin=5V, Vout=1V, Iout=0A. CH1 is TPP0250, this probe is compensated , CH2 is P9139B   You could easy to see when fine scale is 1V, They are all the same. But when fine scale is 100mV, TPP0250 waveform is distort.

Test 2

This is a buck IC's LX pin wavefrom. At first Vin=14V, CH1 is TPP0250, The result of 5V,1V and 500mV fine scale is the some.  And then, Vin=20V, CH1 is TPP0250. You could easy to see that in 500mV fine scale, the bottom of the  wavefrom is no on zero. This is a false result. Because in other fine scale, they are on zero.

When you use DPO3000 and  set the fine scale form 1V to 500mV?use 10x probe) , you can hear a tick of a relay, I think the on this time, scope change the amplifier. But in MDO2000/3000/4000, no this hardware. So it make Test 2 bug

If you are TEK's engineer, could you answer this problem?

[rx8pilot]

Sad to see this thread die. I was very curious about the resolution.

Could not tell what was going on in the screen grabs. Are you able to explain what you are seeing in the traces and how it relates the problem you are having?

[Lukas]

By the way, our supplier says that this bug is on all new MDO 2000/3000/4000. They are the same.

can you post a few pictures of the problem

OK,I post some pictures.

In the pictures with high vertical sensistivity, the waveform is clipped. You're measuring the overdrive perfomance of the scope's amplifiers and ADC rather than the signal. When an amplifier is overdriven, it saturates and needs some (often unspecified time) to recover.

[free_electron]

this is a frequent problem with oscilloscopes. they use a differential stage inside and once you drive that out of its common mode it is adios signal.
there is nothing you can do . you are simply fishing for a small signal on a large offset.

Agilent (now Keysight) has solved that with their new current probe : http://literature.cdn.keysight.com/litweb/pdf/5991-1711EN.pdf

that probe has an internal analog adder in the probe. the scope establishes where you want to look and prgrams a DAC in the probe top to add or subtract a common mode range BEFORE the signal is sent to the scope . this overcomes this problem.

These probes need two scope channels. one is used to show the signal peak to peak , the other one gives the summing output of the DAc and the signal. a clever bit of software in the scope looks at the area you selected and programs the DAC.

i have used these probes to sucessfully look at things in the microampere range happening superimposed on multiple ampere events. the probe has not problem doing this.

drawback : (advanatage ) only works with agilent scopes liek the 4000 , 6000 7000 and 9000 series.

Sod Trektopnix   ( dutch joke. Trekt op niks = trektopnix  means literallyay pile of garbage ). Switch to Agilent (Keysight) and don't look back

if you want to stick with tek you will need a probe where you can dial in an offset but that needs ot be done by hand and you will have to verify how much offset you set.... using a reference current. very annoying and cumbersome.

[tekman11]

this is a frequent problem with oscilloscopes. they use a differential stage inside and once you drive that out of its common mode it is adios signal. there is nothing you can do . you are simply fishing for a small signal on a large offset.


Sod Trektopnix   ( dutch joke. Trekt op niks = trektopnix  means literallyay pile of garbage ). Switch to Agilent (Keysight) and don't look back

pic' xxxx104, two probes different probes are used with different results. Is this the result of probe compensation not being correct?

what other limitation have you noticed with tek scopes?

[cody198879]

this is a frequent problem with oscilloscopes. they use a differential stage inside and once you drive that out of its common mode it is adios signal. there is nothing you can do . you are simply fishing for a small signal on a large offset.

PIC 104 is a picture that I want to show different probe have different results. All probes are compensation. But TPP0250  get a bad wavaform

And alos try the issue 2 on TDS1000 series. But in Old scopes,  they don't have this issue. Any time , if you can hear a tick sounds when you set the fine scale form 1V to 0.5V, they will no have this issue. But new serise,  TDO 4000B, All MDO series all have this bug.

Sod Trektopnix   ( dutch joke. Trekt op niks = trektopnix  means literallyay pile of garbage ). Switch to Agilent (Keysight) and don't look back

pic' xxxx104, two probes different probes are used with different results. Is this the result of probe compensation not being correct?

what other limitation have you noticed with tek scopes?