I had one that threw me for a loop this week. I have been rebuilding a 1949 AC Model G. The cleanup was going fine until I got around to connecting up the hydraulics. The Model G has front and rear cylinders for lifting equipment. The rear cylinder has a spring return and a needle valve to adjust the timing between the two. The operator can adjust the needle valve to delay the lifting of the rear cultivator at the row end so the front would be out of the ground when the rear finishes the row.

          After I cleaned and reassembled the pump, I remounted it and connected the rear cylinder to test the system. I started the tractor and the hydraulic cylinder wouldn’t move. It had a problem: the system would only reach about 100 PSIG. The pump bypass relief valve consists of a spring that pushes on a ball bearing that seals the pump output from the return sump. When the system pressure exceeds about 900 PSIG, the valve opens and bypasses the fluid to the sump to keep the system from over pressurizing. I made new valve parts because it seemed like the relief valve was opening too soon and not allowing the pump pressure to reach the required 900 PSIG. After the 5^th or 6^th pump removal, I decided to replace the spring and ball with a machined rod that would seal the bypass to determine what the ultimate pump pressure would be. I was astonished to find the pump only produced the same 100 PSIG. That drove me nuts until I happened to spy the problem: the G has a latch on the clutch pedal to keep the clutch plates from sticking together during periods of inactivity. I haven’t adjusted the clutch yet because the tractor is on stands. I hooked the clutch rod up and set the adjustment midpoint and assumed I would adjust it when it came time to mount the wheels. As luck would have it, the clutch was just barely engaged when the latch was used. The amount of engagement was just enough to reach 100 PSIG before slipping. Had I turned the clutch nut 1 turn more or less, the problem would have been obvious from the start.