KeepItSimpleStupid
- Joined Mar 4, 2014
- 5,088
Let's use the space between our ears a bit. Vbe is Voltage, Base to Emitter. Maybe better expressed as V(be)
First a little research: https://aetechron.com/Diagnosing-overtemperature-7548-7796.shtml say heat sink temperature below 100C, Using the setpoint number in the 500 amplifier, room temp of 25C and the slope, you get 96.8 degrees as a setpoint. Makes sense.
You need a transistor (a p-n junction) with a Vbe of around 420 mV at room temperature.
You need a transistor (a p-n junction) with a Vbe of around 463+600mV = 1.063 V at room temperature.
Anyway, it has to be bigger than 463 mV.
Usually transistors made in the same batch will have the same characteristics. You might have to buy 20 and then pick 2 for Q16 and Q42. You might get lucky and have two already. Remember close, not the same. Every 10 mV is a degree C.
The lower the Vbe at room temperature, the circuit will "read" a hotter temperature.
"selected" components are not unusual. manufacturer's might say "Q42 selected" and that's it.
We are doing a bit of reverse engineering here.
R121/R126 and R122/R127 are voltage dividers. I won't do the math. They are dependent on the supply voltage. This kinda shows you that ripple and the value of the +15V supply matter. Other critical designs like a voltmeter will use a voltage reference. here we don't care if it's between 90C and 110C that the fan comes on or one comes on at 95 and the other 110C.
"Too hot" doesn't mean > 98.6F like it does in a human.
The entire point of using that diode is to NOT get 15V and to prove an open. It was NOT designed to turn the overtemperature LED off.
We want to as few measuremnts as possible to identify the fault(s).
First a little research: https://aetechron.com/Diagnosing-overtemperature-7548-7796.shtml say heat sink temperature below 100C, Using the setpoint number in the 500 amplifier, room temp of 25C and the slope, you get 96.8 degrees as a setpoint. Makes sense.
You need a transistor (a p-n junction) with a Vbe of around 463+600mV = 1.063 V at room temperature.
Anyway, it has to be bigger than 463 mV.
Usually transistors made in the same batch will have the same characteristics. You might have to buy 20 and then pick 2 for Q16 and Q42. You might get lucky and have two already. Remember close, not the same. Every 10 mV is a degree C.
The lower the Vbe at room temperature, the circuit will "read" a hotter temperature.
"selected" components are not unusual. manufacturer's might say "Q42 selected" and that's it.
We are doing a bit of reverse engineering here.
R121/R126 and R122/R127 are voltage dividers. I won't do the math. They are dependent on the supply voltage. This kinda shows you that ripple and the value of the +15V supply matter. Other critical designs like a voltmeter will use a voltage reference. here we don't care if it's between 90C and 110C that the fan comes on or one comes on at 95 and the other 110C.
"Too hot" doesn't mean > 98.6F like it does in a human.
The entire point of using that diode is to NOT get 15V and to prove an open. It was NOT designed to turn the overtemperature LED off.
We want to as few measuremnts as possible to identify the fault(s).
Last edited: