It is likely the operating temperature is not too high under nominal factory test conditions. I don't plan on further troubleshooting the default software packages for now as more increased cooling appears to be the easiest fix for instabilities from enabling included software.
Specifically, when enabling AdGuard, the AXT1800 appear to be stuck at some of its hardware limits, with the device stuttering and losing connections. Disabling AdGuard did not stabilize the unit. Fan does not turn on at factory settings. Forcing air into the unit did temporarily stabilize, but factory resets were required to stabilize.
AdGuard setting problems elsewhere:
Before seeing the reply, I already disassembled and started summarizing.
I can't tell if 5021-8600 is an old or local PN. Molex Picoblade 51021-0400 appears to be the globally available equivalent wire housing to whatever the fan is using.
3mm wide posts limit about 14.3mm down to about 11.5mm available height
M2x0.4 x 6mm, PWH P0
1.25~1.3mm pitch
Molex Picoblade 51021-0400 housing equivalent, crimps potted with clear material for vibration
1, RED: +5VDC
2, YEL: TACH
3, BLU: PWM
4, BLK: GND
Original fan:
ELEPEAK B4006ASHNF2000TT
5V 0.5A
40x40x6
Equivalent or old laptop fan model which ELEPEAK appears to replace:
SUNON EG40060S1-1C010-SAA
E Series
G motor
40mm frame size
06mm frame thickness
0: 5VDC
S: Sleeve
1: High Speed
1C010: Customer Code
SAA: Function Code
5.5~6.3 CFM
AlElec cap:
- 8mm Dia
- 12mm H
- ~19.3mm max under chassis
Markings:
1500
10V
MB
Likely part from GJT:
-
http://www.gjt.net.cn/en/aluminum-electrolytic-capacitors/559830 -
http://www.gjt.net.cn/aluminum-electro-lytic-capacitors/559830 - 105C
- 20%
- 2khr
- low ESR
But no exact matching PN, and labeling format is not 100% consistent with GJT descriptions.
Comparable 1500uF, 10V, 8mm D, low ESR or not, parts
are 20mm tall, and radial.
Bottom chassis:
- 2x fasteners, M2x0.4 x 6mm, PWH P0
- 7x latching clips