It would involve adding 2 variable resistors (potentiometers) - one in series with the ground wire, and one in series with the +5 volt wire going to the TPS. Stay with the TPS you have in there now if you want to.
But before you add the variable resistors, you would make some measurements. They would be:
1. With the TPS installed but the connector disconnected, measure the following on the TPS itself:
**a. Resistance from ground pin to +5V pin.
**b. With throttle at idle position, resistance from center (wiper) pin to ground pin.
**c. With throttle held to WOT position, resistance from center pin to ground pin.
2. With the TPS installed and the connector connected and the ignition on, measure the following on the wires going to the TPS:
**a. Voltage from ground wire to +5V wire.
**b. With throttle at idle position, voltage from ground wire to center wire.
**c. With throttle at WOT position, voltage from ground wire to center wire.
You would provide me with those numbers and the WOT voltage that causes the CEL/limp mode. Then we would agree on a value for the WOT voltage - one that would be just below that CEL/limp mode value - based on the 4.64 volts you mentioned in your OP, maybe go with 4.55 volts as a safe value - i.e., close to the max., but with some reasonable margin so the CEl/limp mode doesn't get tripped by small system variations (due to temperature, vibration, etc.).
From all of that information, I would develop the formulas to calculate the values of resistance to dial into the two variable resistors which would at the same time give any desired idle position voltage and the safe high WOT voltage.
IOW - you would end up with a table that would have 3 columns. First column would be a list of idle position voltages, starting with the unaltered idle position voltage that you measured before putting in the variable resistors, and then values increasing in, say, 0.05 volt increments up to, say, 1.2 volts. Then columns 2 and 3 would be the values of the ground wire variable resistor and the +5V wire variable resistor, respectively, to achieve the desired idle position voltage and the high safe (i.e., non-CEL) WOT voltage.
From there, it would be up to you to adjust the two resistors according to the table for different idle position voltages for experimenting. As you moved it further up, you would eventually hit an area of idle position voltage in which the performance/fuel mileage would start decreasing. Then you could dial it back down to the sweet spot.
Beyond that, you could either leave the variable resistors in for further experimenting, or you could replace them with optimum value fixed resistors.
If you're up for that, so am I. But I don't want to start it if we aren't going to agree take it to completion.
'98 LXi - Later Concorde gages (black w/ chrome rings)/'99 LX - LHS gages (white) - HIR bulbs