Yes, the 4.8 has a belt-driven pump.

LOL. The cooling system on the N62 is basically water soluble. I've replaced everything over the past few years.

Not that I have found so far. Not sure where else to look.

Kind of the same symptoms, but, IIRC, there was no heat and not off and on heat.

Cooling system has been pressure tested, and I use a vacuum to fill and bleed the system so the coolant is fresh and I don't suspect any air pockets in the heater core.

One thing has got me thinking. When the system started to misbehave the heater core temperatures dropped really fast and really far (80C to 5C in 5s? How is that possible?). Also, sometimes one side would drop and the other would climb, etc. Only when the heater core temperatures rose above 40C would the system start to work properly. Now I'm wondering if I have a bad ground connection somewhere.

Yes, that's kind of chicken or egg if the temp reading is not accurate or the coolant is not circulating properly. Can you get anywhere near the heater hoses to measure temp or even get your hands on them? Gasses from a bad head gasket could definitely do what's going on. Antifreeze that's too strong can cause poor heat but not so much the on and off again.

Any history of overheating?

If it is head gasket issue there is a good chance the heat/circulation will improve if you leave the rad cap off. Be careful with that though.

You would think a bad ground or electrical issue would generate a code??

Any updates on this?

I am beginning to suspect the secondary water pump which I replaced only 1 year ago, as the symptoms seem to match. If I rev the engine so that the primary water pump pushes harder then the heater core temperature rises and the IHKA can do its job of pushing warm air into the cabin.

The secondary water pump is buzzing like it is working, but perhaps the impeller is broken.

I hope to be able to take it out and bench test it sometime this week.

That's a very sound theory especially if it was a cheapy pump. Both hoses should be really hot if the pump is working.

It was a Pierburg pump, which is OE, so not a cheap pump.

Yeah, not so likely for a Pierburg pump to give out. Does it seem noisier than it should be? Heater valves are always open by default so disconnecting should provide for heat and help to eliminate the IHKA.

Any chance that hoses/connections got messed up with a rad replacement?
Otherwise you could be back to a possible head gasket, I'm thinking when you did a pressure test it wasn't to find a head gasket issue?

Removed and bench tested the auxiliary water pump - it's OK. Will move on to the water valve next.

FYI - I pressure tested the cooling system after replacing a failed hose to ensure there wasn't any additional leaks.

Removed the heater control valve. My understanding is that the valve is supposed to be fully open when it is disconnected, but mine is fully shut. Can anyone confirm that the heater control valve is normally open?

Also, is there any way to test the valve? I get consistent resistance readings between pins 1-2 and 1-3. My understanding is that the IHKA sends a PWM signal to pins 2 and 3 to close the valve to a particular setting.

After disassembling and reassembling the heater control valve, both valves are now open. One of the coils had coolant on it, so perhaps some coolant slipped by a seal and either gummed things up or caused the electronics to fail.

I'm going to replace the valve.