FIXED: Hameg HM 203-4 Scope faulty second channel

[Str1ker]

Hi there!

I recently scored myself a nice looking Hameg HM230-4 oscilloscope which seemed to work perfectly on the first glance. After getting familiar with it I noticed, that the second channel is faulty. When I connect the scopes channel 1 to the scopes test CAL signal everything is displayed correctly, but when I connect the signal to the scopes channel 2 it's not showing the square wave (like channel 1), but the signal shown in the pictures. When I switch the channel 2 to GND, it shows a steady line as it should (That's why I didn't notice the fault when buying the scope).
My first uneducated guess when I saw the signal was, that the fault must be somewhere on the Y-input and attenuator board of the scope and be a faulty or dry capacitor (the Signal looks to me like the one of a capacitor discharging). Sadly I cannot find any faulty caps in the accessible areas of the scope to this point...

Has somebody seen such a signal before, or know, which caps (in which areas) or which other components might come in question, so I don't have to disassemble the whole scope for no reason?

(Channel 2 is the upper line in the pictures where both inputs are plotted)

[shakalnokturn]

Start with switch contacts, if you have the same symptoms on all V/Div. settings check the coupling switch. (The AC/DC/GND one.)
If you don't want to disassemble the scope, stand it upright before spraying contact cleaner in it, then exercise it a lot.
What you're seeing is typical of a square wave through a RC with a very small time constant.

[Str1ker]

Ok, thanks for the advice! I'll try cleaning the contacts. While switching them back and forth I noticed, that the signal is correctly displayed by the second channel when it is switched to DC... I'll check the input coupling capacitor, maybe it is the one that's faulty...

[Str1ker]

UPDATE: It indeed was the input coupling capacitor. I measured it before unsoldering it at ~15nF (should be 100nF) so I thought it was damaged. Then (just for curiosity) I measured it again after desoldering it and it measured 100nF like in the first channel. I then soldered it back to the PCB and measured it again... 100nF, spot on. So, in the end, it must have been a dodgy solder joint connecting the capacitor to the PCB.

[shakalnokturn]

A few days ago after digging out one of my older scopes, a HM203-6, preparing to sell it I gave it a quick check.
Plenty of crust in switch contacts right round, the coupling switches got their share of cleaning spray before being manoeuvred energetically.
After that CH1 had lost AC coupling similar to the 203-4 above. I assumed I'd over-done it a little on the switch.

Disassembling revealed that the ends had cracked off the orange 100nF Philips MKT capacitors, switches tested OK.

[niconiconi]

Has somebody seen such a signal before

UPDATE: It indeed was the input coupling capacitor. I measured it before unsoldering it at ~15nF (should be 100nF) so I thought it was damaged. Then (just for curiosity) I measured it again after desoldering it and it measured 100nF like in the first channel. I then soldered it back to the PCB and measured it again... 100nF, spot on. So, in the end, it must have been a dodgy solder joint connecting the capacitor to the PCB.

Just a tip for you, so you'll know if you encounter the same waveform in the future... The strange waveform you saw on the broken oscilloscope was the textbook example of an RC high-pass filter, also known as a differentiator. It won't allow any low-frequency signals to pass, only high-frequency signals can survive. If C is getting lower, its impedance increases and it filters more low frequencies, which is the problem in your case. For a scope-calibration square wave around 10 kHz, it will kill the fundamental frequency and almost all low-order harmonics, leaving only the highest frequencies in a square wave, which is its edge (since the edge is fast-changing, high-frequency), so you get a strange waveform with sharp edges.