Commodore 128 Power Supply Testing and Repair

The Commodore 128 takes regulated +5 V DC and 9 V AC from an external power supply through a 5-pin square DIN plug.[1] Ray Carlsen describes it as a switching supply that runs cooler and lasts longer than the Commodore 64's linear supply; its weaknesses are the factory glue, which becomes conductive and corrosive with age, and solder joints that break loose.[2] Tynemouth Software gives its rating as 2.5 A at 5 V and 1 A at 9 V AC.[3] The procedure below tests the supply on a load, then opens it to clean off the glue, resolder, replace fuses and the two main filter capacitors, check the over-voltage crowbar and set the 5 V output.
Several makers built the supply. Commodore's service manual prints two circuits "for reference only", a Mitsumi supply, part 252449-01, and a Dee-Van supply, part 252449-03.[4] World of Jani drew the schematic of the PAL supplies made by Ismet, 310416-01 to -03, from six units.[5] Labels do not always match the contents: Muldrf's Hobbytronic found a Mitsumi 252449-01 board inside a supply labelled 310416-05.[6]
Symptoms
[edit | edit source]Carlsen lists a dead computer as the commonest power supply fault, and also blank screens with the power light on or off, freezes, garbage on the screen, hum bars moving up the picture and intermittent operation.[7]
- No output at all: Tynemouth found a blown fuse on the 9 V AC side, and both joints of the main 5 V output inductor cracked so that it was loose, giving no 5 V.[3]
- 5 V low: the same supply gave about 4.5 V after the inductor was resoldered, and was trimmed back to 5 V.[3]
- 5 V far too high: World of Jani reports that on the Ismet supplies the 5 V output rises to between 12 and 15 V if the 4700 µF capacitor C9 across the 5 V rail goes open circuit; the crowbar then holds it down until the crowbar fails. On his supply the crowbar failed and he measured 12 V on the 5 V rail.[5]
- No 5 V with nothing connected: some switching supplies need a minimum load before they start. A contributor to an EEVblog repair thread suggested testing a C128 supply with a 10-20 Ω resistor across 5 V and ground instead of the computer.[8]
If the supply tests good on a load and the computer is still dead, the fault may be on the main board: a dirty power switch, the line filter or the +12 V regulator. See Commodore 128 Power Switch and Regulator Repair.
Safety
[edit | edit source]Warning: the supply contains mains wiring, a mains fuse and the transformer primary. Unplug it from the mains before opening it.
- The supply plug has the same shape as the Amiga 500, 600 and 1200 supply plugs, but the supplies are not compatible; using one on the other can damage the hardware.[9] On the Amiga plug two of the pins carry ±12 V where the C128 has 9 V AC.[3]
- Do not connect a supply of unknown history to the computer. Measure its 5 V output on a load first, because a failed output capacitor and crowbar can put 12 V or more on the 5 V line.[5]
Tools and parts
[edit | edit source]- A multimeter, and for the load test a 10-20 Ω resistor.[8] At 5 V a 10 Ω resistor dissipates 2.5 W, so use a resistor rated for at least that.
- A Phillips screwdriver; a dental pick or similar for scraping off old glue.[2]
- A soldering iron and fresh solder. Carlsen applies new solder to every joint he reworks; reflowing the old solder is not enough.[2]
- Fuses of the original ratings. The Mitsumi schematic gives F1 600 mA 250 V in the mains input and F2 1.5 A 125 V in the 9 V AC winding.[4] Tynemouth's supply had a 1.4 A fuse on the AC side.[3]
- Capacitors. In the Mitsumi supply the two electrolytics are C4, 2200 µF 35 V, after the bridge rectifier, and C9, 4700 µF 16 V, on the 5 V output.[4] Hobbytronic found only these two in his Mitsumi supply and fitted 105 °C low-ESR replacements.[6] World of Jani also replaces C4 and C9 in the Ismet supplies with low-ESR 105 °C parts rated for 10,000 hours.[5] The Dee-Van circuit is different; its electrolytics are listed in the capacitor guide.
- For crowbar repairs on the Mitsumi circuit: V9, a 6.2 V zener, and V10, a TIC106D thyristor.[4] World of Jani lists equivalent parts for the Ismet supply's V8 and V10.[5]
Procedure
[edit | edit source]Test before opening
[edit | edit source]| Pin | Signal |
|---|---|
| 1 | +5 V |
| 2 | Shield |
| 3 | 9 V AC |
| 4 | Signal ground |
| 5 | 9 V AC |
Individual Computers notes that the pin layout seen at the solder side of the computer's socket matches the outside of the cable's plug.[9]
- Measure the AC voltage between pins 3 and 5 of the plug. Expect about 9 V AC.[1][3]
- Connect the load resistor between pin 1 and pin 4 and measure the DC voltage across it. Expect 5 V; Commodore's own checks on the computer allow ±10 %.[8][10]
- If 5 V is present only with a load, the supply may be working normally; if there is no 5 V even with a load, or it is high, open the supply.[8][5]
Open the supply
[edit | edit source]Carlsen's procedure:[2]
- Unplug the supply from the mains and from the computer.
- Remove the four screws from the bottom of the case. Hobbytronic found plastic plugs in the screw holes of his supply and had to drill a small hole in one to pull it out.[6]
- Lift off the top half. It may be stuck to the transformer by double-sided tape; pull the chassis out of the bottom half-shell to free it, without stressing the circuit board, which can crack.
- Clean the plastic residue off the screw ends before refitting them later.
Remove the glue and resolder
[edit | edit source]
- Look for factory glue on both sides of the board. New glue is tan or light brown; near hot parts such as the power transistors it turns brown and then black. Black glue absorbs water from the air and causes leakage between circuits and corrosion of tracks and component leads.[2]
- Scrape the brown and black glue away from the wiring and components. Aged glue comes off easily; Carlsen uses a dental pick.[2]
- Replace any component the corrosion or the scraping has destroyed, and bridge any broken track with wire soldered across the break.[2]
- Resolder the terminals of the power transistors and the large parts such as the filter capacitors with fresh solder, and any other joint that looks to be breaking loose.[2] World of Jani found cold joints in several of his supplies.[5]
- Check the output inductor. Tynemouth found it loose despite a cable tie, with both solder joints cracked, and resoldered it.[3]
- Check the circuit board for cracks from rough handling.[2]
Fuses
[edit | edit source]- Check F1 in the mains input and F2 in the 9 V AC winding with the meter.[4]
- Replace a blown fuse with one of the same rating. Tynemouth's supply had no output until a blown fuse on the AC side was replaced.[3]
- A fuse that blows again points to a fault in the supply or the computer. Carlsen traced a blown fuse on the 9 V AC line to a shorted SID in the computer.[2]
Replace the filter capacitors
[edit | edit source]- Note the polarity of C4 and C9 and remove them.[4][6]
- Fit replacements of the same capacitance and at least the same voltage. Hobbytronic's replacements were shorter and slightly smaller in diameter than the originals.[6]
- On later Ismet boards World of Jani found an extra capacitor, C12, and R13 fitted at 100 Ω; C12 can be retrofitted to earlier boards.[5]
Check the crowbar
[edit | edit source]On the Ismet supplies the crowbar pulls the output down if it exceeds 6.2 V.[5] In the Mitsumi circuit it is zener V9 (6.2 V) driving thyristor V10 (TIC106D) across the 5 V output, with R13 (1 kΩ).[4]
- If the supply has ever produced a high output, or blows its fuse, check V10 for a short and V9.[5][4] World of Jani's failed supply had a leaking capacitor and a blown V10.[5]
- Replace C9 as well, since its failure is what lets the output rise.[5]
Set the 5 V output
[edit | edit source]- With the load connected, measure the 5 V output.[8]
- Adjust the trimmer to 5.0 V. On the Mitsumi supply it is R11, 200 Ω.[4] Tynemouth set a supply that had drifted to about 4.5 V back to 5 V this way.[3]
Reassemble
[edit | edit source]Carlsen's reassembly:[2]
- Set the assembly back in the bottom half. The transformer fits into a recess, and the board should sit flat on the bottom of the case. If it tilts up, a wire is pinched underneath; reroute it.
- Press the mains and DC cords into their places.
- Fit the top. Check that it seats fully and evenly all round before refitting the screws; if not, wiring is holding the board up.
Testing
[edit | edit source]- Repeat the load test: 5 V between pins 1 and 4 and 9 V AC between pins 3 and 5.[1][8]
- Connect the computer and measure +5 V at the positive lead of C107 or pin 25 of the SID (U5), and +12 V at pin 28 of the SID, within ±10 %.[11][10]
- Run the computer for a while and recheck. Carlsen lists intermittent operation among the supply's symptoms.[7]
Revisions
[edit | edit source]| Label or board | Maker | Notes |
|---|---|---|
| 252449-01 | Mitsumi | Circuit in Commodore's service manual, printed p. 78[4] |
| 252449-03 | Dee-Van | Circuit in Commodore's service manual, the page after the Mitsumi circuit[4] |
| 310416-01, -02, -03 (PAL) | Ismet | Schematic drawn by World of Jani from six units[5] |
| 310416-03 | Photographed above; made in West Germany[12] | |
| 310416-05 | Mitsumi board 252449-01 inside | Found by Hobbytronic[6] |
Related pages
[edit | edit source]- Commodore 128
- Commodore 128 Power Switch and Regulator Repair
- Commodore 128 Capacitor Replacement Guide
- Commodore 128 General Maintenance
- Commodore 128 Troubleshooting Guide
- Commodore 64 Power Supply Testing and Repair
- Capacitor Failure Symptoms
References
[edit | edit source]- ↑ 1.0 1.1 1.2 1.3 Service Manual C-128/C128D (1987), schematic 310378 rev. 6, sheet 2 of 4, printed p. 67 (PDF p. 72): power connector CN11 pin assignment.
- ↑ 2.00 2.01 2.02 2.03 2.04 2.05 2.06 2.07 2.08 2.09 2.10 Ray Carlsen, "C128 power supply repairs", updated 7 April 2012: glue, solder joints, the fuses, disassembly and reassembly.
- ↑ 3.0 3.1 3.2 3.3 3.4 3.5 3.6 3.7 3.8 Tynemouth Software (Dave Curran), "Commodore 128 Repair - Part 1 PSU", 9 October 2014: ratings, the Amiga connector, the blown AC-side fuse, the cracked inductor joints and the 4.5 V output.
- ↑ 4.00 4.01 4.02 4.03 4.04 4.05 4.06 4.07 4.08 4.09 4.10 Commodore Business Machines, Service Manual C-128/C128D Computer, PN-314001-08, November 1987, power supply schematics "for reference only": Mitsumi PN-252449-01, printed p. 78 (PDF p. 89), and Dee-Van PN-252449-03 (PDF p. 90). Hosted as File:Commodore 128 128D Service Manual 314001-08 (1987 Nov).pdf.
- ↑ 5.00 5.01 5.02 5.03 5.04 5.05 5.06 5.07 5.08 5.09 5.10 5.11 5.12 World of Jani, "C128 PSU Schematic (PAL) P/N 310416-01,-02,-03", 12 April 2024: cold joints, the 6.2 V crowbar, C9 failure, replacement parts and the C12/R13 revision.
- ↑ 6.0 6.1 6.2 6.3 6.4 6.5 Muldrf's Hobbytronic, "Commodore 128 Power Supply Recap", 23 February 2019.
- ↑ 7.0 7.1 Ray Carlsen, "C128 chips and common symptoms", updated 8 August 2013, "Power supply" entry.
- ↑ 8.0 8.1 8.2 8.3 8.4 8.5 "Commodore C128 PS repair HELP PLEASE!", EEVblog forum thread: minimum load and the 10-20 Ω test resistor.
- ↑ 9.0 9.1 Individual Computers, "Powersupply FAQ", section "Commodore 128": the connector and its incompatibility with Amiga supplies.
- ↑ 10.0 10.1 C128 Diagnostic Instruction and Troubleshooting Manual (1986), section 2.2.1, p. 2-2: +5 V at the positive lead of C107.
- ↑ Commodore Business Machines, C128 Diagnostic Instruction and Troubleshooting Manual, Assembly #314060-01/02 (1986), section 2.1, p. 2-1: +5 V at SID pin 25, +12 V at SID pin 28, ±10 %. Hosted at Commodore 128 Diagnostic Instruction and Troubleshooting Manual (314060).
- ↑ Bultro, File:Alimentatore C128.jpg, Wikimedia Commons, CC BY-SA 4.0: part number 310416-03, made in West Germany.