brunnhilde's air conditioner

It seems these days my things are breaking faster than I can fix them. This remind me of the MTBF calculation, the more parts in a system the lower the MTBF. The number goes to hell in a hand basket in a hurry if they have strong Independence.

The ticking noise of Brunnhilde's cabin AC is traced to a fray on the drive belt. The roof top air conditioner is made by Kerstner of Germany. The most unusual thing about it is it is a 12Vdc unit rated at a whopping 100A. Why so high current? It just take that much energy to refrigerate even a small room. It equals roughly (equivalent AC power is measured in RMS) a household window mount unit drawing 10A at 120Vac so it is not big at all BTU wise.

blowing out the pine noodles

ticking noise is caused by this frayed edge

It turned out finding the replacement belt is more difficult than first seems. I have to do considerable amount of research to find out the cryptic part marking of this Pix brand translates to generic metric part number of 5K630. It means it is a micro-v serpentine belt with 5 v-ribs and with a inner loop length 630mm. Here in US these belts typically has a generic part number of xyzK5. Instead of mm xyz is the inner loop length in inches measured to one decimal place. Hence the equivalent belt would read 248K5. I nearly fell victim to this transposition convention tribal knowledge and bought the wrong belt.

Problem is for a belt manufacturer the permutation of this micro-v serpentine belt is huge. The number of ribs ranges from 3 and up. The loop length ranges from roughly 20" to over 100". Very few applications implement belt as short as one I need because typically there is a spring-loaded tensioner wheel in the design. In the case of the AC in Brunnhilde there isn't one due to space constraints.

Short of ordering the exact replacement belt from Europe, the closest belt I can locate in the US is a 237K5 so has a nominal inner loop length of 602mm. Will it fit?

Checking it is way too much trouble right now so I resorted to the photos I took last year. Assuming my existing belt has not stretch much (a bad assumption) there is about 1.2 - 1.5 mm of adjustment range for a shorter belt. Note that the displacement on the adjustment is roughly half of the belt loop length so judging from the photo it would be a tight fix if at all possible.
very cryptic part markings on the Pix belt

Paying more attention to the photo one can see the DC motor is mounted to the steel plate with 4 bolts. The motor mounting flange has slotted mounting holes to afford some lateral movement (for alignment purpose). However you cannot see the space on the side where the motor need to move closer to the compressor in the photo.

There is one more place one may be able to gain a bit more adjustment range for a shorter belt. It is to lengthen the slot of the adjustment arm towards to the compressor side. It is unclear if the compressor can rotate this extra amount without inspecting it.

Anyway, I think it is a very good chance I can make the 237K5 belt fit.