Hantek DSO5102P repair

A while ago, my Hantek DSO5102P oscilloscope simply stopped working. I discovered that the power supply was goosed. Specifically R14 (3x 1R5) and R11 (?) had popped. Unable to procure a replacement power supply board, I attempted to repair it but when this failed, I decided to reproduce the various supplies using off-the-shelf adjustable supply modules. At the time, I recorded the process I went through to achieve this. I thought it might help others in a similar situation if I shared this and so here it is. I started with investigating the power supply board in order to replace any components which I deemed faulty. Helpful links here, here and here.

Chips on power supply board are:

• CR6850T: Current mode PWM controller
• PS817C: Photocoupler
• XL1509-3.3E1: PWM buck (step-down) DC/DC converter, capable of driving a 2A load

My plan is to replace the faulty power supply with multiple buck DCDC converters all running from a single 12V mains adapter:

• 3V3: 100mA, LM2596S Buck Converter
• 5V:   2A, LM2596S Buck Converter
• 14V: 400mA, XL6009 Boost Converter
• -8V:  200mA, XL6019 Buck/Boost/Pos/Neg Converter
• 12V: ?mA, fan motor, use 12V direct from adapter
• TR-AC signal: 50Hz square wave at 5V

Power supply cabling

TR-AC signal generation circuit

CR6850 reference circuit

Looks like Hantek have used the above reference circuit ... Q1 = 4N60C, R11 = 100R, R14 = 0.5R (3x 1R5 in ||), N2 = CR6850T ... also F1 fuse = T3.15AL 250V ... have ordered all these from amazon and will attempt to fix the PSU board in 1st instance and if that fails, build own PSU with discrete buck and boost regulators.

Have replaced components with those I ordered and fuse popped immediately after switching on! Replaced fuse with wire to find where the possible short is and switched on again and this time the common mode choke burnt out!

This scope is specified at 30W max so with 240V rms, I = 30/240A = 125mA. So have ordered replacement choke (CR6850 datasheet shows T1 = 20mH) Farnell 1636281 (500mA) and Farnell 2860718 (800mA). They are not a perfect fit however. Also ordered 0.1uF/275VAC caps (Farnell 2495780) - might as well as have to pay a £10 handling charge whatever!

Replaced common mode choke and tried again and this time RT1 (in line with the fuse) popped! RT1 = NTC 5D-9 Inrush Current Limiter, Power Thermistor 5 ohm 3Amp. Also the track from mains input to RT1 and track to D4 both burnt out! DMM across D3, D4 shows a short in both directions! Also common mode choke coils are now o/c! Giving up on repairing this!

Have now made up my own DIY power supply board. Below is a photo of the new board. Need to (1) set all the voltages using the pots (2) program the Raspberry Pico to output a square wave on GPIO15 for TR-AC (3) wire up the 14V adapter to scope on/off switch.

New power supply board with schematic

The 3V3 50Hz square wave is being generated with a Raspberry Pico and then level-shifted to 5V and passed to the 'TR-AC' input. The Pico program is toggle.py (attached) which has been renamed to main.py (so it will autorun) and saved to the Pico.

Raspberry Pico program

Note that the setting of the -8V supply requires moving the pot and then waiting for the huge capacitance to charge or discharge! The scope is now working again with 14V input (from mains adaptor) and drawing 1.2A.

New power supply board in position

Scope now working again!

OK, that was short-lived. Tried scope yesterday and the display comes on OK but the trace remains at 0V with noise but is not responding to the scope probe. Checked voltages (under load) and -8V is only -5V and so increased this to -8V (which increases the +8V to +11V!). 5V is OK. 3V3 is actually only 2.6V and unable to increase it (also LM317 getting very hot) and so replaced with LM2596 buck DCDC and now getting 3V3 ... and scope trace is now working! DMM shows 14.4V/1.4A.

Unfortunately autoset keeps coming on and the trace has low frequency (32Hz) modulation on it (even when probe connected to its own 5V/1KHZ test point) and so the scope remains unusable! If I short the scope input coax connector, I still get the 32Hz modulation. Tried factory reset (power on then power off before boot complete and then power on again and select F2) and no change. Tried FW update with *.up file from Hantek website but it says no update file found on USB stick.

The autoset is perhaps not a problem after all. Maybe the light is meant to always be on and when pressing the autoset button, it takes a few secs to do its job (during which its light goes off).

The low frequency noise is certainly an issue but I've noticed that it isn't present if I turn scope on when it is cold! Have noticed that the -8V (actually now adjusted to -10V for better results?) drops gradually after switch on and it also starts to waver up and down by 1V or so. I think it's the -8V which is causing the LF noise. Need to get a better negative supply for this. Also, the -8V seems to switch off completely (down to ~0V) if left on for a few hours.

After watching the video here, I've now configured the LM2596 as an inverting converter to generate the -8V and it seems to work fine! The scope now works as it should! The only problem is that sometimes on power on, some of the button lights will flash on/off continuously and the scope never actually powers up. I guess this is due to the inverting converter because that's the only thing that's now different. Perhaps the -8V should only be switched on after the other supplies are on. Have now wired in a manual switch on the 14V to the inverting converter so it can be switched on *after* the main power comes on. Actually it is probably the inrush current on the inverting supply which is causing the 14V mains adapter to switch off and then restart over and over again. The switch on the -8V actually doesn't really help i.e. sometimes it works and sometimes it doesn't! Updated power supply board is as shown below. Tried a 15V/5A mains adapter to see if extra current capability (albeit 1V extra) might be the solution but ... LM2596 inverting converter now no longer works ... arrgghh! Fortunately, after replacing it with another LM2596 board, it's all back to working again, phew!

Updated power supply board