Round crossover duct between supply and return - leave it open or close it off?

I dealt with a similar setup in my rental last winter, and it was definitely an inefficient loop. My heating system used these large rectangular supply branches, and the smaller crossover you're describing was right where the main air handler sat. The damper on that thing was a flimsy piece of galvanized metal, and it seized up completely after a particularly harsh season. It felt like we were paying to heat the basement only to have the warm air dumped right back into the return intake before it could hit the upstairs zones. If the damper is truly stuck, trying to force it shut could snap the linkage and create a much bigger problem. What I'd do is inspect the actuator mechanism itself. Sometimes these cheap levers just lose their spring tension, causing the stuck position. I remember on my last job with a Lennox unit, we found the internal linkage had rusted, so we swapped the whole duct damper for a solid, custom piece of sheet metal secured with screws, rather than relying on the factory lever. That stopped the heat recycling immediately. Before you start taping anything off, just double check the surrounding metal for any debris or old insulation that might be wedged around the pivot point of that damper. A small piece of fiberglass can completely bind it up. Fixing the mechanical failure of the damper is the right first step before worrying about system pressure or flow balance.

When I was working on a zoned system a few years back, I saw a similar crossover and it wasn't just a sloppy loop. It looked like it was part of a manifold design intended to manage backpressure. If the blower is pushing too much air out of the supply side, having a controlled bypass or return path helps regulate the static pressure in the entire plenum. It’s often used to stabilize the system when the fan rapidly shifts speed upon startup.

That warmed-up duct you're seeing might not be wasted air, it could be air that's momentarily recirculating to keep the airflow balanced while the system ramps up. If the damper is stiff, it suggests the internal mechanism might be designed to partially close under certain pressure conditions. If you force it fully shut, you could actually be causing a pressure spike that stresses the motor or the entire housing.

My instinct is that you should investigate the blower's performance curve first, before altering the ductwork. Check your static pressure readings to see if the restriction is coming from the filter, the vents, or the blower itself. Sometimes those ducts are simply built that way to handle the system's 'breathing' cycles.

It seems like the fan might be working too hard trying to push through that stuck spot. I saw a house once where the heater was running really loudly because the blower was struggling to force air through that restricted path. If the unit has to push through a tight blockage like that all the time, could the blower motor itself start to get damaged? I just worry about that kind of strain.

I worry that the static pressure differential across the main heat exchanger registers could be significantly elevated due to the compromised flow paths. When the forced airflow is forced through a partially constricted damper, the system is inherently generating unnecessary resistance which directly impacts the overall pressure readings. It woud be highly advisable to measure the static pressure in the supply plenum and compare that value against the published specifications for your specific blower unit. If that pressure delta is inconsistent with the design intent, you are almost certainly dealing with a genuien efficiency loss beyond merely the localized air recirculation you currently observe.

I think relying solely on swapping out that seized damper for a piece of custom sheet metal might not be the whole answer. While a broken lever is certainly a failure point, you're treating the symptom before you've mapped out the bigger picture. If that crossover loop is truly an inefficient recycling path, the physical damper failure is just a secondary issue. Before you modify the airflow physically, you really need to measure the static pressure across the main supply registers. That tells you exactly how much resistance that crossover is putting on the system, regardless of wheather the lever is stuck or not.