This Time It's Serious - Hung Chang HC-5504 Oscilloscope Repair

Parts have started to arrive for the planned repair of my oscilloscope. I will document what I do here, and try and do it properly so anyone who has the same problems or needs to replace the HV tripler in this type of oscilloscope might stand a chance, based on my success (or failure) to do the same. Some information here is duplicated from previous posts.

The oscilloscope in question is badged as an IsoTech ISR441 40Mhz scope. I believe that these were branded by RS Components in the late 80's and early 90's. Following the research and lucky finds I had on the internet I can identify that the model I have is a re-badged 'Hung Chang' 5504 40Mhz oscilloscope. The link to the manual can be found here from the EEVBlog forums. (If this stops working leave a comment and I'll see if I can fix it.)

Hung Chang 5504 40Mhz scope

The problem I had with the scope is basically it now has no trace. Prior to the trace disappearing I had noticed that it had started to become dim, but only sometimes. Once it was dim it would sometimes go back to being bright again, sometimes it would not. Changing the controls had no real affect that I can speak of. One day, the trace was bright as normal, and the scope was on. After a few minutes I turned to use it and the trace had disappeared.

Switched on but no trace

Checks revealed that the test signal on the front still works and power supply rails are at the correct voltages.

Inspection of the board with the HV circuit (there are two main boards in the scope - the one in question is the board mounted on the side of the scope) revealed that there were apparent burn marks on the HV tripler unit at the point where the Anode cable leaves the unit. There were also signs that the plastic case had started to melt and had expanded and 'bubbled up' in some places. 

Burn marks close to Anode cable

Deformed plastic case

Removal of the HV tripler from the circuit board also revealed a large hole in the plastic with evidence of the expulsion of burnt potting compound onto the heatsink of a nearby power transistor (Q2001).

Hole in faulty tripler

Q2001 heatsink shows the material expelled
from the faulty tripler

This scope makes use of a separate Flyback Transformer and HV Tripler unit rather than the more conventional combined units as seen in CRT TVs and CRT based computers such as the iMac G3. 

Page 37 of the manual shows the circuit diagram for the voltage tripler unit. These units are not manufactured and so the only option to repair this scope is to build a new tripler from discrete components using the parts list and circuit diagram as a guide.

Tripler circuit diagram (page 37 of manual)

The components required are:

• 6 x 1000pf 6kV ceramic capacitors
• 6 x Y-16GA or equivalent 10kV high voltage diode (typical forward current 5mA)
• 1 x 1/2W 10M ohm solid resistor

The components I have purchased are:

• 6 x 1000pf 15kV ceramic capacitors
• 10 x 10kV diodes
• 2 x 1W 10M ohm resistors

Note that the ratings I have may differ from the service manual. This is primarily due to availability of components, especially the capacitors. Note that in most cases it is acceptable to pick a voltage rating higher than the original component. 

Capacitors and 1W resistors

I will also be making use of the original Anode cap and cable. To contain the unit, once constructed, I have 3D printed a small box that matches as closely as possible the dimensions of the original ABS box. This initial 3D print is made with PLA which, despite being an excellent insulator at room temperature, begins to lose its insulating properties as its temperature rises towards 65 degrees celsius. As such, I would not recommend PLA for the final box since it sits very close to the heatsink for the large power transistor Q2001 - a D880. Using ABS instead of PLA will provide a more suitable box since ABS has a much higher 'glass transition' temperature i.e. the temperature at which the plastic starts to become 'rubbery'.

In my case the PLA box will be used to evaluate whether the constructed tripler fixes the oscilloscope and, if so, a more appropriate container will be procured/constructed/printed.

Two sizes of replacement box

Q2001 and Heatsink
(Note proximity of tripler box outline)

Space is very limited within the confines of the board and the fly-back arrangement is also covered by a silicone (or rubber) lined aluminium cover. I intend to use the space as the factory unit did with the 3D printed box, if possible but it will be tight due to the number of components required to build the unit. If the aluminium cover is to be retained, which I very much hope I can do, the potential for going beyond the original unit size is very limited due to the mountings of the two main boards.

Aluminium cover - limited access

Tripler Construction

To start, I laid out the components that I have in a rough pattern of where they would sit. Note that in this picture I was still waiting for the final delivery of diodes to complete construction. As such, the final diode is missing from this picture which would connect from the top right most diode to the resistor and capacitor bottom right.

Rough Layout

First, I soldered the two groups of three capacitors together, making sure to leave a small amount of leads between the caps so that I had something to connect the diodes to. Then I twisted alternate ends of the diodes together to make an extended 'W' shape, being sure to put the cathode (the end with the markings) pointing in the correct direction. Following the circuit diagram, with voltage coming in from the left, the cathode marking was on the 'right' of every diode following through the ladder.

Once this was done I soldered the legs of the diodes to the gaps between the caps. This resulted in the following:

Partially constructed

Due to the limited space in the box, I put the caps so that they will be at the bottom of the box with the diodes connected on top. There is also a 10M ohm resistor that is the final component before the cable that runs to the anode cap at the CRT.

Resistor added to end of Anode cable

The leads of the first two caps are long enough to extend down through the bottom of the box and into the correct places on the circuit board i.e. the two holes at the 'In' end of the tripler outline. The two holes at the 'Out' end of the tripled are only used for anchor points (an assumption) as they both appear on the ground plane underneath and the actual 'out' voltage is the anode wire and cap. 

Tripler location next to transformer
Note 'In' and 'Out' markings

I intend to to use a 'u' shaped piece of silver wire to perform the same anchor function in the 3D printed box I will use. In the picture above the space for the tripler has two holes on the left which are the ground 'anchor' points, and two holes on the right. The hole nearest to the transformer (top) is also ground. The hole at the bottom is the 'in' voltage that arrives from one side of the flyback transformer.

Completed construction
(note second diode has become disconnected)

Following completion of the tripler, I installed a small piece of silver wire to support the box at the end where both connections connection to ground. Next, I placed small pieces of heat-shrink on the joints of the capacitors to try and reduce the potential for arcing between the components due to the high voltage. Then I slid a large piece of heatshrink over the end of the anode cable before soldering the end of it to the final diode/cap joint in the ladder.

With this done I slid the whole thing into the box, lining up the two legs of the first section of the ladder with two holes I made in the bottom of the box. These holes line up with the 'in' and 'ground' connections on the 'in' side of the tripler. Finally, I put a piece of insulating tape over the top to hold the cable in.

I have some concerns with this very temporary design. First, I don't know if my efforts to insulate the components have gone far enough. Ideally, the whole thing should be potted into the box. BUT, as mentioned above the temporary box here is made of PLA and is not ideal for scenarios with temperatures over 65 degrees celsius and so it will be replaced. Potting it would make box replacement almost impossible without re-building it all over again. Second, the final 10M ohm resistor presses up against the PLA. I don't know how hot it may get and so I am wary of leaving it in there since, if it gets too warm it may melt the PLA exposing the resistor to potential arcing targets.

So, this arrangement (as shown below) will at least show if this is going to work.

Assembly complete
Ready for installation

Re-installed on the board.

Awaiting main unit re-assembly

Re-assembly of the oscilloscope was a reverse of disassembly. There were three wires to re-solder into board, one brown wire (connects to a BNC connector on the front panel) and the two wires that drive the CRT 'x' deflection. I had to be careful soldering these back in as the pads on the board started to lift when I removed the cables originally.

All the screws were re-installed and the plastic cover re-attached with the black pegs holding it in place. The cover is important as it offers some protection against accidentally touching the high voltage section of the board - so not to be left off! The numerous connectors were re-attached with their positions being fairly obvious due to the orientation and natural position of the various plugs, so there were no issues there. A couple of them are deep inside the centre of the board and I found a long pair of tweezers were helpful in getting these re-connected.

The main power cables were re-connected to the board and these were the only connections I hesitated on as I realised I did not have a good photograph of them pre-disassembly. I verified that I had them correct by checking the connector orientation and voltages in the manual. 

Re-attaching the CRT neck board was straightforward as the connector is keyed and will only sit flush if attached correctly. Finally, the anode cap was re-attached to the CRT.

This is now in a suitable condition to try a switch on. 

Ready for test run

To ensure maximum safety I plugged the unit into an extension cable and then took it and placed outside of the garage door. The thinking being that in the event of a catastrophic failure the damage would be limited to my garden and not the garage and its contents. In addition, I took the added precaution of using a piece of wood to turn on the main power switch. 

And the result was... a working oscilloscope. This is a fantastic result but not the end of this project. As mentioned above I intend to 3D print an ABS box which will be far more heat resistant that the PLA one currently installed. Once that is done, which will require another disassembly and re-assembly, then I will move on to the calibration process which, thankfully, is well documented in the manual.

Success!