Rigol DS1052E CH2 microphony ?

[Fraser]

I was doing some tests on my new DS1052E when I discovered that Channel 2 appears to be sensitive to the DSO's case being tapped or bumped. The effect only occurs on the 5mV or 2mV ranges in AC or DC, but not GND, and appears as a very brief burst of noise on the flat trace (no signal being input to DSO). My first thoughts are "dry joint" and in the past I have found some SMD components that only have a pressure contact with the solder pad and a poor or non existant solder joint, so that is a distinct possibilty here.

I would like to check that this isn't just a 'feature' of the DS1052E present on other units before I dive inside it's case.
Has anyone else got this occurance on their DS1052E please ? If not, the next step will be to grab the screwdriver and do some hunting inside the CH2 input module.

 

[sonicj]

might want to swap leads before you go tearing into it. 
-sj

[PeterG]

Does this happen when you remove the test leads?

If it does, have a look at this video Dave did a while ago showing something very similar to what your seeing.
http://www.eevblog.com/2009/06/21/eevblog-14-a-rather-unusual-oscilloscope-phenomenon/
 Just my 2cents.

Regards

[Fraser]

Thanks for the suggestions guys.

I have seen Dave's interesting video on the proble pulse but this differs in scenario and is a complex noise band across the screen.  

My quick and dirty test was carried out with the DSO running with open inputs, i.e. no test leads attached and with BNC inputs unterminated.

I considered static discharge and loose input connectors etc but the odd thing is that it is only found on CH2 and at a very low level. I could just ignore it but I have a niggling suspicion that there is a dry joint in there scratching away when the DSO's case is bumped. I will do further tests with a decent input signal and see how it behaves.

The idea of hunting an intermittant contact inside a DSO stuffed with SMD parts does not fill me with glee  

If/when, I discover what is going on, I will report here.

[sonicj]

its actually not all that scary inside. the inputs have a shield over them though and im not sure how much effort is involved with removing it. the other thing is that you'll be breaking the warranty seal (if it still applies).
-sj

[Fraser]

Hi Sonicj

Thanks to Dave’s excellent teardown and the content of this great forum I understand the process of disassembling the 1052E. This information is invaluable as it saved me doing any damage whilst attempting entry  Sadly the access that I will require requires that he main PCB be removed and this in turn requires that I disassemble the front panel and remove the SMPSU first 

I have studied the high-resolution images of the input stages and considered possible tests to identify the fault if it is not easily visible. I dread the prospect of a cracked SMD capacitor or resistor ! The PCB layout looks pretty friendly though. I am hopeful that a high magnification visual inspection will reveal the intermittent connection. I have even considered a faulty input relay contact.

I am only holding back on the repair because I have been considering my options with regard to the 100 MHz upgrade. I am running hardware V58, FW 2.04 SP1  , so the firmware upgrade is highly risky at this time and I don't fancy 'bricking' my new scope. As the unit will already be disassembled I am thinking that the most sensible way forward would be to repair the fault and disable or modify the 50MHz  low pass filter circuit at the same time.

Regarding the warranty... a good point , but I purchased from Aidetek in the USA and I am in the UK . The cost of shipping and risk of UK Customs complications basically makes the warranty pretty worthless to me on a £220 DSO.

It could be a busy but enjoyable weekend 

[Simon]

bear in mind that the scope does pickup noise from the SMPS, I have demonstrated this by shielding it with aluminum foil and then when turning the scope back on the trace with no input was under "0" instead of on the zero line requiring a recalibration. I know now that the aluminum foil was not an ideal shield but it proved a small point, the scopes calibration seems to deal with the biasing and slight noise induced but maybe hitting the scope causes something to move (the SMPS inductor ?) causing an induced noise on the input ? if you look at the main board you will notice that only the inputs have shielding, there is a whole load of analogue circuitry up to the 5 ADC's (3 on top and 2 underneath) that is not shielded.

when I get around to it I'll put a steel screen parallel to the mainboard

[Fraser]

Hi Simon,

Thanks for that information.

The problem I am experiencing is only present on channel 2 and at the highest sensitivity ranges with no cables connected or signals input. I titled this thread 'microphony' because I wondered if the CH 2 is vulnerable to the influence of other components when the case is tapped. Your comment on the SMPSU is interesting and I will investigate when looking for the cause of the problem. I was in two minds whether to just ignore the sensitivity to case impact as it isn't a great problem for me at this time but my curious side is getting the better of me and I don't like unknowns in my test gear  
Take a look at the photo below. This was captured with the scope in single shot trigger on a 'pulse' setting. The input attenuator is at 5.0mV and the impact to the case was me knocking the area under CH2 input with my knuckle so it is a softish impact rather than the harsh type a hammer would produce  

In fast timebase sweeps you only see the CH2 trace 'jump' several times but at the slow sweep you can see the waveform that the impact produced. CH1 has some small response but CH2 definitely doesn't take kindly to case impacts. The greatest response is found when knocking the bottom of the case between the input channels. Other 1052E's may in fact behave the same at these settings... I would be really grateful if someone could repeat my simple test with no input cables attached to see if this is just a design 'feature' caused by a component placement or just a component that is sensitive to shock.

[armandas]

I tried it, but couldn't reproduce the result. The noise levels seem to be a bit high on your scope. With probes disconnected, I get something around 500 uV.

Did you try running self-calibration?

[alm]

Check if your probe attenuation settings on the scope are the same, since it will multiply/divide all displayed values by ten. If Aurora's scope is set to 10x, and armandas's scope is set to 1x, that could explain the difference. Or a difference in knock amplitude/frequency, gotta calibrate everything .

If not, it might be a piezoelectric effect (high K capacitor vibrating?) in Aurora's scope.

[Simon]

yea look at the scope attennuation, it can be set from 1-1000X so you may have an issue there, you picture did not come out in your post, I'll have a go on mine in a minute see what happens (I don't have the shield installed any more)

[Simon]

ok, just given mine a "bashing", it did not really pick up anything, some really hard hits produced some bursts but really nothing unusual.

[Fraser]

Thanks for your comments everyone.

My scope is set for a X1 probe and when sat on a bench produces a good, low noise flat line. It is only when I give the case a sharp tap, or bump it, that the event occurs. I am sure the wise would say "if it only does it when you thump it..... don't thump it"  
I found it interesting to see that CH1 shows some deviation during the bump event, but nothing like the amplitude of CH2 which is what initially caught my eye. Before anyone reprimands me ... I know that it is not good practice to go around thumping your test kit  

The comment regarding piezo effect has peaked my interest... the effect does have that characteristic to it. I had assumed that I had a dry joint as I have often seen such a joint produce electrical noise when agitated. I will hold off ripping my scope apart for the moment and carry out further tests to establish whether I am chasing my tail here  

If anyone else wishes to set their DS1052E to the 100.0uS TB,  5mV range (X1 probe setting) with a single event pulse trigger and then give it a good bump on the underside, I would welcome the results.

Thanks everyone.

[Simon]

maybe it is the wires from the PSU to the main board bouncing around

[d0ss]

Had to give it a good beating to get anything to trigger, poor scope never had so much abuse.

[Fraser]

Well that confirms it.... my Rigol has a little 'issue' with being bumped 

I shall have to do some internal investigating. If truth be told, I am putting off the job because it's such a faf to remove the motherboard. Whatever happened to easy access.... these days you seen to have to strip an item to virtualy a bare chassis to get to the motherboard fixings 

Ah well it will have to be done as I hate unknowns. I shall report back here when the cause has been identified. Watch this space.

[Fraser]

Ok it's fixed 

I stripped the 1052E down to a bare chassis and visually inspected the motherboard around the CH1 and CH2 input circuits. As it turned out I needn't have bothered ! Read on.....

1. Whilst disassembling the scope it was noted that the motherboard retaining screws and BNC connector nuts were loose. Whilst it crossed my mind that this could be the cause of the problem, I decided to investigate further and as the scope was already disassembled it gave me the opportunity to look at the build quality of the PCB.
2. Once the motherboard was extricated from the chassis, it was visually inspected in the region of the CH1 & CH2 input circuits with a USB microscope (damned useful bit of kit). No suspect joints were found. The general quality of component placement and soldering was very good though PCB ‘hygiene’ was not great with fingerprints and minor detritus visible.
3. The scope was partially re-assembled to the point that it could operate but the CH1 & CH2 screening cover was not refitted. A 'knock' test was carried out on every component in the CH2 input circuit and adjacent PCB. On the 2mV range It was possible to produce a ringing waveform if the PCB was struck hard (but not too hard  ). It was noted that a similar ringing waveform could be induced into CH1. This tended to disprove this as a fault as it is more likely to be a normal microphonic response within a component in the input circuit.
4. Specific capacitors in the input circuits of CH1 & CH2 were checked for shock sensitivity. Two large pink capacitors did produce impulse type responses and this is likely to be piezo effect as already suggested by another member of this forum.
5. It was not possible to reproduce the original fault symptom as detailed at the start of this topic

The scope was carefully re-assembled and all screws and nuts were tightened to a decent torque. The scope was then placed in Self Calibration mode after which it was tested for impact response.

My DS1052E now behaves normally and it takes a significant thump on the case to cause any sort of response on either CH1 or CH2. Both channels provide similar ‘bump response’ on the 2mV range. This is of negligible amplitude and may be attributed to normal component impact sensitivity and related effects.

So it would appear that my scopes sensitivity to impact stemmed from loose screws and nuts !
It was in interesting little exercise with which to occupy my evening