A friend offered me this stove board, after he had to replace it. This board is expensive over $300 – I guess they can almost charge what they want, profit margins on appliance repair components must be very high.
I was not expecting to find that much here to look at, and on initial inspection it doesn’t look much just 8 relays that switch power to the elements an MCU and some power related components. However the display looked interesting and digging deeper there is some neat technology to look at more closely.
Vacuum Fluorescent Display
On the back of the board fitted in the plastic frame is this display board.
This is display type I have never looked at, after a bit of research I figured out it is a Vacuum Fluorescent Display. A type of display that has mostly been replaced by the ubiquitous LCD, but apparently are still being put in some appliances.
Getting it off the board turned out to be difficult it is not only soldered (14 pins along one etch) but is attached to the board with a 1cm wide strip of the strong adhesive. Physically it is a glass box, 9mm thick with 2.5mm thick glass plate top and bottom connected with a 4mm glass spacer. Underneath the rear glass plate is a PCB (Which due to the reflective glass was difficult to photograph). The round evacuation hole is visible where the unit is evacuated and sealed with gases – presumably Neon.
At first I thought the board was glued to the underside, but now I am pretty convinced the glass plate is actually the substrate, the metal pattern (Chrome?) is patterned on the glass – using the same or similar process to how photo-lithographic masks are made. this is then coated with a black insulating material with openings on the inside for connections (To anodes).
This is a zoom in on the traces through the glass with a low power microscope objective
The glass pieces are connected with a glass frit bond and I couldn’t find a way to separate them, I tried all sorts of heat. Nothing worked so in the end I resorted to smashing it apart!
These displays work by having a tungsten filament wire(s) cathode that generate electrons. The positive anodes are coated with phosphors that glow when hit by electrons, in between are thin grids that are switched on/off (-ve) to control electron flow to the anode. You can see these elements here
There are ~20 mesh control grids of which a couple can be seen here in this macro image.
I took a peak with the microscope. This was taken with a 5x objective around the degree symbol above. You can see the carbon anode material, the phosphor, and you can see the mesh grid, which appears to be gold plated.
At this point I thought I had done with the display and through the pieces away. Only going through the board I couldn’t figure out how the display was driven, I traced the cathode connections which I believe are driven with 12V AC. But I didn’t think the MCU would drive the display, so I dug the broken bits of glass out of the bin and took another look. Hidden under the right metal plate that connects and supports the cathode wires is this (4.5mm x 2mm) driver chip
This was imaged with the chip still attached to the glass, on one side the bond wires are cover with a white undersell material (That appears black in this bright file image – white and shiny surfaces like the bond wires here appear dark in BF not sure why). Here is a full size image after I removed it from the glass
There are 64 channels, I think the structures at the side is a shift register and looking at a slice through one section from top to bottom what I think you have is a latch circuit, then level shifter in the middle, and the bottom is a long (Wide in transistor terminology) driver transistor.
Moving on to the PCB
I don’t know if there is some sort of standard for symbols printed on PCD. I did a bit of googling but nothing turned up. I have never seen the hand before, and it points to a track connecting to the large cap and the display connection.
and beside that below the piezo beeper is a 8 pin dual comparator LM358
The MCU is a Renesas R5F series
And this is it imaged with dark field. Not sure you can see anything more (Possibly less) but the image for me is better to stare at.
Power Integrations LNK306GN
The 4 pins on the bottom of chip traced back directly to the mains live pin (230V in this case.) It is not often you see a silicon chip connected directly to the mains input.
A datasheet exists for this part. It is an ‘Offline Switcher IC” and is described as a lowest cost and component count buck converter solution. It has a 700V breakdown voltage with a transistor that also has less than 10Ω on-resistance (RDS) which is impressive.
About two-thirds of this 2.5mm x 2mm die are taken up by the humungous power transistor
I spent quite a bit of time trying to understand the power device, but was unable to figure it out with just the plan view. This is a device that needs a cross-section to properly understand. You can see the wide channels, which are likely a low-doped drift regions that allow the device to withstand 700V, and I believe the polysilion gates, and the source/drain connections are shunted by Aluminum down there whole length to minimize the RDS.
I thought that was all the silicon on this device, but then when I was trying to understand the display driver circuit I discovered this device is not what I thought (A transistor). It is in fact a ST L78M05 a positive voltage regulator, capable of an output current unto 1.5A.
So I dug the board back out of the bin (Again) and decided to depot this chip.
Like many of these small analog devices, it is easy to image (Large features) and quite interesting. This is one where you can relatively easy work out (reverse engineer) the circuit.
The Bipolar device at the top of the device, has a really unusual layout. You can see the emitter is connected to the VOUT pin on the right. And the collector is connected to the VIN pin on the left. The base connection that goes off to the other control circuitry has a strange sawtooth like doping regions. It really is quite remarkable how different the device layouts are that are used for analog functions.