Working on some stuff recently I got very annoyed that the Yihua 948 solder sucker doesnt suck hard enough. I've seen posts related to putting holes in the side of the thing to exhaust the air more directly. I've seen posts related to changing out the vacuum pump. I took a slightly different approach to it.
When studying the circuit layout, I arrived at the attached schematic. Its not quite accurate but more about that later.
Gripe #1 about the circuit: The 500 ohm resistor and the small sot-23 NPN transistor are fighting each other whenever the pump isn't running. The 500 ohm resistor is mean to provide the base current for the Darlington when the pump is running, but all other times the 30-something milliamps through that resistor are going straight through the tiny transistor. Wasteful and making that little transistor hot hot hot.
Gripe #2 about the circuit: Lots of voltage drop from the supply through the Darlington, yielding only ~10.2V at the pump. Not great when we've got 16V available on the board. I tested the pump from a bench supply from 10 up to 17V and it seemed happy enough and had much more chooch at the higher voltages
Schematic problem #1: I depicted the small-ish electroytic cap on the board across the supply rail. Thats not where it lives. The one on the board goes across the motor terminals. Probably for handling some of the noise from the brushed motor?
Schematic problem #2: The black diode body shown still on the board in the "parts removed" photo is a Zener that was apparently limiting the base voltage in the original. I ended up removing that too as it will interfere with the Gate voltage of our P-channel replacement going all the way high and shutting off the transistor.
So the approach:
- remove the 500 ohm resistor
- remove the small NPN transisor and bridge the two terminals to pass the signal from the opto directly to the gate of our P-channel
- remove the zener diode
- remove the NPN darlington transistor
- remove the heat sink to give room to work and put the P channel FET on caddywumpus
- LEAVE the 100K pullup on the output of the opto / gate net. This is the shutoff for the P channel
- Place a P channel FET of appropriate ratings in place of the darlington - but you have to reverse the source and drain terminals. I crossed them over but I suppose just as easy would be install those backwards and bridge the low-current gate signal around to the through hole that originally held the Base of the Darlington.
- I added a 470 and 10uf cap on the supply side of the 500 ohm resistor that was removed and tied the ground side to the low terminal on the opto just next door - This is probably optional but I liked the idea that we wouldnt be completely losing the voltage on the output with every cycle of the 120hz diode bridge output. 500-ish uF may or may not be enough to accomplish that goal. Circuit sims suggested 1mF or more would be better.
Two photos attached of the after condition. One from off the edge of the board looking in at the transistor. The other looking outward from the center of the board to the output side of the opto where the caps were added.
Testing on a live project still pending. Also, long term durability of this is TBD since I just finished the mod, but wanted to share with the group.
Happy tinkering -
Jason
Home
Help
Search
Login
Register
