Strange OmniStat2 issue, 64F heat setpoint

[dschoppe]

Hi,

I just noticed a very strange issue with my Omnistat2s (firmware 1.02c), connected to HAI T-stat plug-in (vers 1.0.16.0) under HS 2.3.0.19.

If I change the heat point using the pull-down temperature list in the HAI RC Thermostat Status page (i.e., under Plug-Ins -> HAI RC Thermostat) I am not able to set the heat point to 64F. The HS log indicates that the heat point has been changed to 64F, but instead the actual heat point on the stat is changing to 63F.

However, if I instead control the heat set point using the up/down controls on the heat point device from the HS status page, I have no problem stepping through all temperatures, including 64F.

Would someone with an Omnistat2 please give this a try and let me know if I'm going crazy?

Thanks,
Don

[dschoppe]

Bump...

Before I submit a help desk ticket about this issue, would someone with a Omnistat2 please try to change the heat point to 64F (from the pull-down temperature list in the HAI RC Thermostat Status page) and check if the thermostat's actual heat point becomes 63F instead?

If someone with an older stat such as the RC-80B could try this as well, I'd appreciate it.

Thanks,
Don

[HacksawHA]

Works with my RC-80B

I just tried it with my RC-80B and it works fine. I tried 63 and then 64 from the droplist and the 'stat went to 63 and 64 as expected.....

[dschoppe]

I just tried it with my RC-80B and it works fine. I tried 63 and then 64 from the droplist and the 'stat went to 63 and 64 as expected.....

Thanks for checking. Maybe this is just a problem then with the Omnistat2s or the specific firmware in my units. All 3 are running 1.02c and behave the same.

Just need to know if any other Omnistat2s are working the same and then I'll ask HS about this or check with HAI.

Thanks,
Don

[dschoppe]

After using a serial spy program and referencing the HAI thermostat protocol document, the issue is clearer. HAI's protocol uses a lookup table to map "omnitemp" values to temperatures (which can be fractional). See attached table copied from the HAI documentation.

Here are four packets that the HS plug-in sends when changing the heat point to 63F, 64F, 65F, and 66F respectively via the drop-list in the plug-in status page. The 4th byte in each packet is the Omnitemp value in hex (note: HAI table uses decimal).

01 21 3c 71 cf

01 21 3c 72 d0

01 21 3c 73 d1

01 21 3c 75 d3

You can see that the plug-in sends 0x73 (63.5 F) for 64F rather than 0x74 (64.4 F).

Apparently the older series stats display 63.5 F as 64 F, whereas the Omnistat2s display it as 63 F. The new stats must instead be sent 0x74 (64.4 F) in order to show 64 F.

This can be verified by "stepping" consecutive heat points via the "up" control on the heat point device rather than selecting heat point directly using the drop-list on the thermostat status page.

It would great if the HAI plug-in could add an option to configure the type of each connected stat (legacy vs series2) and then adjust the omnitemp / temperature mapping accordingly.

Thanks,
Don

Here are a few more temperatures at which this issue surfaces:

To set 55F the plug-in sends 0x69, but Omnistat2 requires 0x6a
To set 73F the plug-in sends 0x7d, but Omnistat2 requires 0x7e
To set 82F the plug-in sends 0x87, but Omnistat2 requires 0x88
To set 91F the plug-in sends 0x91, but Omnistat2 requires 0x92

The pattern suggests that old HAI stats round up XX.5 F for display, but the Omnistat2s round down. Thus, omnitemp values for the Omnistat2 must be one value higher in these cases.

[mterry63]

Did this ever get updated in the plugin?

[dschoppe]

Did this ever get updated in the plugin?

As far as I'm aware, HST has not updated the plugin since I reported this issue. I think it is a minor fix and probably on the back burner compared to other items HST has on their plate currently.

Don