I have seen this problem and do not know a work around. Have you received an answer from N.I. yet? N.I. is currently working on a another Modbus problem we saw and is the main reason why we have not switched from 3.8 to 4.X. The pole rate can be dramatically effected by consecutive reads i.e. reading 0x01,0x02,0x03...0x0100 will be fast but if there is any break in the numbers the scan rate is slower. i.e 0x01,0x03,0x05...0x099. You can actually pole 100 points consecutive reads faster than you can poll 20 points non consecutive. There should be a fix this month I hope.