I have been trying to optimize a calibration routine for the ADE 7758. The results are not good.

Let me start with a simple question regarding how to derive the Watt Gain register setting using Line Accumulation and an Accurate Source.

I do not need the pulse output for anything. Can I ignore the CF registers (leave as default)? If not, what should I set them to?

Also, what is the (4 * Meter Constant) in the equation required for? I have no meter to provide me with a constant, so I simply chose 3200. Is this okay?

My setup is power factor = 1, providing about 120V, at about 10 Amps, and the line frequency is 60Hz. I am calibrating one phase at a time. My accumulation time is 15 seconds. I have no WDIV factor applied right now, so the register has been left at the default setting.

If I calculate the **expected value** based on equations 60 and 61, I get either 64d or 4032d depending on whether I include the 0x3f default in the APCFDEN register as a factor in the equation. ** However, when I read the raw value from the WATTHR register after 15 seconds**, I get 0x4060, or 16,480d. If I then try to compute a register value using expected and the accumulated WATTHR number, the result is obviously wrong.

For example,

((4032 / 16,480) - 1) * 2^12 = -3094. That is a register overflow given a signed 12 bit register.

I know from experiments that the proper adjustment is typically not < -150 or > +150.

I have made sure that the WATTHR register is cleared before i begin the accumulation step.

Can someone advise me as to what I am doing wrong?

Thanks so much!

I should add that if I use equations 45-47 to generate settings for APCFNUM and APCFDEN I get 507/1.

Adding that to the to the registers and equations does not help things. I still wind up with register overflow.

Interestingly, using 277/1 pulls things more into line. However, this system is 120V nominal, not 220V.