PD or Iota for my upgrade?
I'm thinking about a converter and battery upgrade. The batteries will be wet cell lead acid 6 volt Possible T-105s or the 230AH, EGC2 from Sam's club. I'll probably go for four. (Does nayone have...
Forum Discussion
I'm closing in on a more complete understanding of the PD design.
Here's how I now think it works:
The FOD 2741 optically isolated error amp is connected to the UC3846 current mode controller. The input to the FOD monitors the output of the converter which must be at the "target voltage". If it's not, the FOD will produce an error voltage and tell the 3846 to send more current corresponding to that error voltage to try to get the output up to the target voltage or the current limit is reached. The higher the error voltage, the more current the 3846 sends. (I'm going to pretend that the error voltage is the difference between the battery voltage and the target voltage, but in reality, it's scaled down so that the target the FOD tries to reach is always 2.5 volts)
There's one resistor that sets the "target" voltage for each mode (boost, normal and float). The microcontroller selects the mode and its matching target voltage based on various inputs or the manual override.
A second resistor sets the relationship between the error voltage and the current output. That resistor (actually it's multiple resistors that PD can adjust or change) sets the slope of the voltage decay that's being discussed and that I'm trying to control. IOW, the difference between the target voltage and the battery voltage sets the current level from the PD. As the battery voltage climbs,this difference decreases and the current drops. By changing this resistor the PD can be convinced to increase the current output for a smaller voltage differential.
A third resistor sets the current limit by limiting the maximum permitted error voltage. Even if the battery voltage is down to 5.8 volts and the target voltage is 14.8 volts, which would produce a 9 volt error voltage, this third resistor will limit the error voltage to a value that corresponds to 80 amps.
Last night I modified resistors 2 and 3. The actual work was trivial - I just had to solder two resistors - each one in parallel with another existing resistor to lower the combined resistance - no cutting or desoldering needed. The problem was the extensive calculations of resistor values required. I was shooting for a 33% increase in the ratio between output current and error voltage (resistor 2) so that 60A became 80A and a corresponding decrease in the current limit voltage (resistor 3) so that the current limit remained at about 80A.
Overnight, I drew down about 40 amp hours from a fully charge 460AH bank. I didn't want to do a big discharge, turn on the converter, see that I made a mistake and discover that my current limit was 200 amps, with the converter trying to make lots of smoke!
I saw it go up to 73 amps at 14.3 volts and then rose fairly quickly to 14.7 volts at reduced current. I think that's an indication I didn't do anything too horribly wrong, so I'm currently going for a 50% discharge and I'll test the charging tonight to see if it now holds at a more steady current limit for longer.
It's fun :)
Here's how I now think it works:
The FOD 2741 optically isolated error amp is connected to the UC3846 current mode controller. The input to the FOD monitors the output of the converter which must be at the "target voltage". If it's not, the FOD will produce an error voltage and tell the 3846 to send more current corresponding to that error voltage to try to get the output up to the target voltage or the current limit is reached. The higher the error voltage, the more current the 3846 sends. (I'm going to pretend that the error voltage is the difference between the battery voltage and the target voltage, but in reality, it's scaled down so that the target the FOD tries to reach is always 2.5 volts)
There's one resistor that sets the "target" voltage for each mode (boost, normal and float). The microcontroller selects the mode and its matching target voltage based on various inputs or the manual override.
A second resistor sets the relationship between the error voltage and the current output. That resistor (actually it's multiple resistors that PD can adjust or change) sets the slope of the voltage decay that's being discussed and that I'm trying to control. IOW, the difference between the target voltage and the battery voltage sets the current level from the PD. As the battery voltage climbs,this difference decreases and the current drops. By changing this resistor the PD can be convinced to increase the current output for a smaller voltage differential.
A third resistor sets the current limit by limiting the maximum permitted error voltage. Even if the battery voltage is down to 5.8 volts and the target voltage is 14.8 volts, which would produce a 9 volt error voltage, this third resistor will limit the error voltage to a value that corresponds to 80 amps.
Last night I modified resistors 2 and 3. The actual work was trivial - I just had to solder two resistors - each one in parallel with another existing resistor to lower the combined resistance - no cutting or desoldering needed. The problem was the extensive calculations of resistor values required. I was shooting for a 33% increase in the ratio between output current and error voltage (resistor 2) so that 60A became 80A and a corresponding decrease in the current limit voltage (resistor 3) so that the current limit remained at about 80A.
Overnight, I drew down about 40 amp hours from a fully charge 460AH bank. I didn't want to do a big discharge, turn on the converter, see that I made a mistake and discover that my current limit was 200 amps, with the converter trying to make lots of smoke!
I saw it go up to 73 amps at 14.3 volts and then rose fairly quickly to 14.7 volts at reduced current. I think that's an indication I didn't do anything too horribly wrong, so I'm currently going for a 50% discharge and I'll test the charging tonight to see if it now holds at a more steady current limit for longer.
It's fun :)
