Insights from full recap and restore of Trayloader G3 iMac
- Thread starter 68krazy
- Start date
Great work, 68krazy! Reminds me, I should probably work on a list of caps and a servicing guide for the eMac's analog board. That's something Apple did not intend for the user to service... If you try finding a new CRT, let alone a new analog board, Apple would just give you a whole new top half that goes to the front bezel.
Caps don't have to leak to be bad, but perhaps that's what you're getting at.
iMac G3s are known to have issues, especially on the down converter board, so bad caps in these wouldn't surprise me. And they were very much budget/consumer machines - and also subject to tons of heat because cooling in them is almost non existent, so again bad caps wouldn't surprise me.
Anyone want an iMac? I have a rev D 333 MHz for parts, needs work. If you want it it's yours for the cost of shipping or free for local pickup.
yo! I’m having a strange issue with my 1998 iMac G3. It works without any problems initially, but when it’s left on for a long time, a weird bug occurs that’s difficult to pinpoint. After being on for an extended period, my iMac G3 simply goes into a black screen mode if there’s any change in video resolution. Additionally, when I turn it off and try to turn it back on, the power LED stays green, never turning off, and the machine doesn’t power back on. The same happens during a reboot; instead of the LED turning yellow and then green, it doesn’t return from the reboot. Even with the startup chime playing, the screen remains black. The only solution is to unplug it and plug it back in. After being unplugged for a while, it returns to normal, but the bug reappears if it’s left on for several hours, typically two or more. I’m not sure if this is an issue with the mainboard, power supply, or analog board. I’ve already recapped 100% of the capacitors, and while the bug used to be permanent, it’s now intermittent. At first, I thought the recap had solved the problem since I only used the iMac for a short time, but when I use it for longer periods, the bug always comes back. It only stabilizes if it’s unplugged for at least 10 hours. Any attempt to turn it back on after the bug appears is futile. Do you recommend checking any specific chip or shutdown signal? Thanks for the help. Oh, I forgot to mention, only the electrolytic capacitors were replaced during the recap.
This is my imac G3: https://area31.net.br/wiki/Restoring_a_1998_iMac_G3_rev_A
This is my imac G3: https://area31.net.br/wiki/Restoring_a_1998_iMac_G3_rev_A
After troubleshooting, the iMac G3 Rev. A (M4984) developed a very unusual intermittent fault that initially appeared to be related to the analog board, power supply, flyback transformer, or synchronization circuitry. The machine would boot and operate normally when cold, but after remaining powered on for a few hours it would begin exhibiting severe display-related problems. Any attempt to switch video modes, such as changing from 640x480 to 800x600, would immediately cause the CRT to shut down and display a black screen. Once the failure occurred, the flyback SCREEN control could no longer generate a raster, making it appear as though the monitor section had completely powered off. The problem was even more severe at 1024x768, which consistently failed to operate correctly.
The fault also affected the power management behavior of the machine. After the display failure occurred, the power LED would remain green even when the system was shut down, instead of returning to the normal orange standby state. Warm reboots would fail, producing the startup chime but never restoring video output. The only way to recover the machine was to disconnect AC power completely and leave it unplugged for several hours. Because the symptoms involved display modes, synchronization, power sequencing, and CRT shutdown behavior simultaneously, several components became suspects during the investigation, including the flyback transformer, TDA4853 H/V sync processor, B+ regulation circuitry, high-voltage protection circuitry, and various capacitors in the horizontal section.
The root cause was ultimately traced to transistor Q708, a Toshiba 2SC5404 Horizontal Output Transistor (HOT). Although the original transistor was not shorted and the machine remained partially functional, it had apparently degraded with age and thermal stress. Replacing only the 2SC5404 completely resolved all symptoms. The CRT now operates normally at 640x480, 800x600, and 1024x768, warm reboots function correctly, the power LED returns to the proper standby state, and the machine no longer requires AC power removal to recover. This confirms that a marginal or thermally unstable HOT can produce symptoms that closely resemble failures in the flyback transformer, synchronization processor, power supply, or protection circuitry, despite the actual fault being limited to the horizontal output transistor itself.
The fault also affected the power management behavior of the machine. After the display failure occurred, the power LED would remain green even when the system was shut down, instead of returning to the normal orange standby state. Warm reboots would fail, producing the startup chime but never restoring video output. The only way to recover the machine was to disconnect AC power completely and leave it unplugged for several hours. Because the symptoms involved display modes, synchronization, power sequencing, and CRT shutdown behavior simultaneously, several components became suspects during the investigation, including the flyback transformer, TDA4853 H/V sync processor, B+ regulation circuitry, high-voltage protection circuitry, and various capacitors in the horizontal section.
The root cause was ultimately traced to transistor Q708, a Toshiba 2SC5404 Horizontal Output Transistor (HOT). Although the original transistor was not shorted and the machine remained partially functional, it had apparently degraded with age and thermal stress. Replacing only the 2SC5404 completely resolved all symptoms. The CRT now operates normally at 640x480, 800x600, and 1024x768, warm reboots function correctly, the power LED returns to the proper standby state, and the machine no longer requires AC power removal to recover. This confirms that a marginal or thermally unstable HOT can produce symptoms that closely resemble failures in the flyback transformer, synchronization processor, power supply, or protection circuitry, despite the actual fault being limited to the horizontal output transistor itself.
