Reply 20 of 23, by Doornkaat
Sphere478 wrote on 2021-11-15, 05:26:Are you saying that the lower voltage makes the regulators hotter despite the k6 + processors themselves using less watts than other processor types?
🤔 makes sense, higher the voltage difference between input and output the more conversion heat is produced. I see this on my solar controllers. But that’s a difference of nearly 100v didn’t think a 1v difference would be that bad.
Yes, the lower voltage will heat up the LDO more at the same current. Your solar controllers probably use a switching design though to which this doesn't necessarily apply. (Assuming you are talking about the kind you use to power your house.)
Another thing to consider is that lower heat dissipation doesn't necessarily mean less Amps drawn. If we view the CPU with the capacitor in between LDO and CPU as a inear load at the same total power consumption a 2.0V CPU will draw roughly 40% more current compared to a 2.8V CPU.
If we now run the same 2.0V CPU at 2.2V (and still assume the CPU and capacitor together act as a linear load) we get roughly 60% more current draw compared to a 2.8V CPU with the same nominal power consumption.
Of course this is a simplification but I would be cautious nonetheless. 👍
Oh, and I wrote 'draws more power' instead of 'draws more current' in my previous post. My mistake. I'll fix that.
Repo Man11 wrote on 2021-11-15, 04:17:The TX chipset MB8500TTD has the same issue, so Feipoa added a small fan to cool the VRM when running a K6-2+. Re: mb-8500ttd manual needed
From the pictures in that thread it appears the MB-8500TTD is using a switching mode VRM to power the CPU. It can probably handle a lot more current than the simple LDO on OP's MB-8500TVX.