Hey everyone, I'm diving into the VRM phases of my Legion Tower 5 26ARA8 AMD B650 motherboard, but I'm running into a bit of a wall. This motherboard is a proprietary OEM model, so information online is pretty scarce. I've tried to analyze it myself using ChatGPT, but I need some extra eyes on my findings. I'm particularly unsure about the configuration because I didn't notice any doublers as suggested at the end of the ChatGPT response.
Here's the info I have on my board:
- **Top array:** There are 5 pairs of chokes and chips labeled "MPS2403 MP87691 A3N024A TH" located around the top side of the CPU socket. Notably, there's a wider gap between the 3rd and 4th pairs, which makes me wonder if there are two groupings of 3 and 2 pairs.
- **Left array:** This also features 5 pairs of chokes and chips with the same labeling around the left side of the CPU socket.
- Additionally, there's another choke and chip marked "MXL7630S KBQ5446.01 2231 Y" at the bottom end of the left array.
- Lastly, I spotted a chip labeled "MPS2342 MP2857 R7655288X" further north of the top array.
You can check out the analysis I got from ChatGPT [here](https://chatgpt.com/s/t_68fc13c6c04c8191a9736cfdaba5abb3). I'd love to hear your thoughts!
3 Answers
Regarding your upgrade to either the 7800X3D or 9800X3D, I’d say it largely depends on how well your current motherboard handles the thermal demands of those CPUs. If the VRMs aren’t overheating and you ensure adequate cooling, it should hold up fine! Just keep monitoring the temps during your usage scenarios.
From what I can see, it looks like there aren’t any separate doublers on the board. Usually, if a controller has fewer phases, it connects two power stages per phase, which is likely the case here. Your assessment of a 10+2 board seems accurate based on your description. And honestly, asking ChatGPT about this might have been a bit of a gamble considering the complexity of VRM phases. But hey, it’s a great starting point for understanding things better!
It sounds like you’re onto something there! If the MXL7630S isn’t acting as the PWM controller and you’ve identified the MX2857 as the main one driving the phases, that changes things. An 8+2+1 configuration does make sense, especially if you’re not seeing any doublers. If you can post some photos, even if they're not perfect, it might help others provide better insights. Plus, making sure those VRMs are cool with tiny heatsinks is a smart move!
I get what you're saying. It's always tricky relying on AI for this kind of technical analysis, but it can help get the ball rolling. Just make sure to validate anything it says with real-world evidence, like checking your board hardware against the data!