I recently acquired a 1980s Acorn Archimedes A3000 32-bit computer for my collection (which I successfully restored in another article), which came with a matching Acorn AKF-17 CRT monitor. Both were in good condition and came with several accessories, but were sold-as-seen with partial testing.
The Acorn AKF-17 is a 14″ 15KHz CRT colour monitor with an analogue RGB video input, compatible with A3000, A3010, A3020, A4000, and A5000 computers. As with most similar monitors of the era, it is essentially a rebadged Philips chassis (in this case, a CM8833 Mk.II) with some of the optional components not fitted.
These were often bundled together with Acorn Archimedes computers in electronics stores, so sold quite well in the UK. However, CRTs were often binned and they are notoriously difficult to ship, so it can be a bit tricky to get hold of an AKF-17.
The monitor arrived safely and seemed to work well, with the exception of the power switch – initially it was stuck in the “on” position, then it worked OK for a few power cycles, then it failed completely and wouldn’t stay latched on any more, so I had to hold the switch in to keep the monitor powered on. Broken power switches are a common problem in CRT monitors of this era.
I previously swapped out the power switch on my Commodore 1084S-D2 CRT and it was a relatively simple procedure, so I decided to just bite the bullet and do the same here. The first step was to dismantle the monitor, and check over everything inside.
Before we start, a word of warning: like all other mains-voltage electronics, CRTs can be dangerous to work on, so it is important to take care and use the proper precautions – if you don’t know what you’re doing, it’s probably best to leave it to someone who does.
The high voltage side of CRTs can contain extremely high voltages (up to 30 KV), and the “low” voltage side has mains voltage and large line filter capacitors, all of which can give you a nasty shock – most CRTs contain safety features to automatically discharge any stored charge (i.e. bleeder resistors), but even if these exist they can fail, so it’s always best to err on the side of caution.
If you work on a CRT, you do so at your own risk – keep one arm behind your back so as to minimise shock across you torso; if possible, work on the CRT when it is unplugged and powered down, and safely discharge the CRT anode cap before starting work (preferably using a high-impedance lead so as not to damage the CRT coating).
Due to a phenomenon known as dielectric absorption, capacitors (including the CRT itself) can self-charge again even after being discharged – if you’re worried about this, keep any potential hazards shorted out while you’re working.
If you do need to work on the CRT while it’s powered up, use an isolation transformer.
The rear panel is held in place with four screws, and simply pulls off; it is connected to the mainboard via a speaker cable, which can be gently pulled from its header.
The CRT anode cap can then be discharged and removed – I use my homemade CRT discharge tool, which includes a 10 MOhm power resistor to prevent arcing.
The neck-board can then be pulled (carefully) from the CRT neck; the mainboard is connected to the front panel by several multi-cables and the CRT dag cable, which need to be removed – I’d recommend taking lots of pictures at this point, to make sure that all of these go back in the correct place and orientation later on; the mainboard carrier is held in place with four screws, and simply pulls out – take care not to damage the front controls.
The mainboard simply slides out (carefully) from the plastic carrier.
Whilst the mainboard was so easily accessible, I gave it a good clean and checked all of the electrolytic capacitors, and all seemed OK.
Then, I turned my attention to the power switch – the rear metal chassis plate is held on with two screws, and the power switch cover simply pulls out from the switch itself.
I compared the original part with a similar new-old-stock part which I had available to ensure that they were compatible, and it appeared that the only functional difference was its metal collar, which would therefore need to be swapped between the two parts.
The next step was to remove the original power switch, which is a relatively simple process – the switch itself has four pins with plated through-holes (despite the CRT having a single-layer PCB), and the metal collar has three split-pins which are splayed out for extra mechanical support. I first flowed all of these areas with fresh solder (which takes time on the collar, as this draws a lot of heat away from the soldering iron), removed all the solder using my desoldering station (a Duratool D00672), bent the split-pins inwards, then gently rocked the power switch until it fell out of the board.
As mentioned previously, the only functional difference between the original switch and its replacement was the metal collar, so this would have to be changed – it is held in place by two locking tabs on the top, so when these are bent out, the collar slides out.
The installation of the new switch is simply the reverse of its removal – all the contacts are soldered and the PCB cleaned, the plastic switch cover pushes onto the end of the switch, and the metal chassis cover screws into place with two screws.
I took this opportunity to test the power switch before reassembling the CRT, and the plastic switch cover seemed to be a bit gummed up – after applying a small amount of silicone grease, the switch worked exactly as it should.
The reassembly of the CRT is the opposite procedure to its disassembly.
After this work, the CRT seemed to work well – no more dodgy power switch.
- Display works correctly on all video inputs.
- Green-screen mode works OK.
- Audio output (stereo) works OK.
- All switches, knobs, and dials work OK.
- Power LED works OK.
EDIT: a big thank you to Hackaday, who recently featured this post in a short article!