(Draft) [Model 64112] (Outback FM Charge Controller) Disagrees with last Outback Spec.

[ke6jjj]

Somewhat unwisely, OutBack seems to have inserted a new field, CCconfig_Temp_Comp_Slope into their model map between the time they first published it with the alliance and the first version available to me in their ModBus development package (AXS Port Development Package 97.zip, described in issue #273).

This field ought to be inserted after CC_Config_temp_comp_hlimt and before CC_Config_auto_restart in the current code.

DID	Start	End	Size	R/W	Name	Type	Units	Scale Factor	Contents	Description
64112	28	28	1	R/W	CCconfig_Temp_Comp_Slope	uint16	Mili Volts (sic)	N/A	Programmable	RTS temp compensation Slope 2-6 mV per Degree C

Additionally, the latest spec (AXS Port Development Package 105.zip) adds several fields not currently documented:

TBD

[jsbosch]

Additionally, the latest spec (AXS Port Development Package 105.zip) adds several fields not currently documented:

That's not the latest spec. Unfortunately, some people made changes that should not have been made. SunSpec models should never have insertions, any new models should be appended, but only before a model is published. One downside of SunSpec is the lack of a version point in the model header. But that's a different battle.

Have you asked OBP USA for a copy of the last (current?) AXS Port sdk? They may release it, possibly without NDA.

[ke6jjj]

That's not the latest spec.

You seem to have a lot of knowledge. Can you tell me what is the latest spec?

Again, I'm not arguing that I have the most complete information at the moment. I'm arguing that even package 105 is way more modern than the definitions in this repository at the moment. It doesn't make sense to roadblock updates just because there might be even more modern definitions.

[jsbosch]

I'm not trying to block updates unless they would be incorrect.

Please see my profile.

[ke6jjj]

I'm not trying to block updates unless they would be incorrect.

I have three versions of the OutBack ModBus spec that support the update I'm asking for in this issue, yet you keep insinuating that the update I am proposing is incorrect, but when I ask for clarifications you haven't been forthcoming with any other evidence. If you have additional info, please post it here, so that everyone can see it and evaluate it.

[jsbosch]

I insinuate nothing, nor am I trying to be mysterious. I don't have this information at my fingertips. If I have access to it at all, it would likely be in email archives, which will take up to a couple of days to check.

[jsbosch]

The latest release version was 4.0.0.2052 on May 8, 2020. (The latest development version was 4.0.0.2064, but it was not released.) The model change log does not indicate any updates to 64112, so those changes occurred before my time. I conjecture that OBP submitted 64112 to SunSpec, made changes, and didn't resubmit, which should have been done. (In our documentation, we advised customers to use our specs, not the custom models published on SunSpec's website. I think SunSpec was relatively new at that time and the developers were still figuring it out. Again, that's conjecture.)

Any changes to the published model should have been by appending to the model and updating the length point. At least that would not break existing implementations based on the published model.

These are the supported models in the last version:

• SUNSPEC_COMMON_MODEL_BLOCK_DID (0x0001)
• SUNSPEC_AGGREGATOR_BLOCK_DID (0x0002)
• SUNSPEC_END_BLOCK_DID (0xffff)
• SUNSPEC_INVERTER_SINGLE_DID (101)
• SUNSPEC_INVERTER_SPLIT_DID (102)
• SUNSPEC_INVERTER_3PHASE_DID (103)
• SUNSPEC_112_INV_FLOAT_DID (112)
• SUNSPEC_120_NAMEPLATE_DID (120)
• SUNSPEC_121_INV_CONTROLS_DID (121)
• SUNSPEC_122_INV_STATUS_DID (122)
• SUNSPEC_123_IMMED_INV_CONTROLS_DID (123)
• SUNSPEC_124_BASIC_STORAGE_CTRLS_DID (124)
• SUNSPEC_126_VOLT_VAR_DID (126)
• SUNSPEC_129_LV_RIDE_THRU_DID (129)
• SUNSPEC_130_HV_RIDE_THRU_DID (130)
• SUNSPEC_132_VOLT_WATT_DID (132)
• SUNSPEC_134_FREQ_WATT_DID (134)
• SUNSPEC_135_LO_FREQ_RIDE_THRU_DID (135)
• SUNSPEC_136_HI_FREQ_RIDE_THRU_DID (136)
• SUNSPEC_137_LV_RIDE_RM_CONN_THRU_DID (137)
• SUNSPEC_138_HV_RIDE_RM_CONN_THRU_DID (138)
• SUNSPEC_139_LV_RIDE_THRU_MOM_CESS_DID (139)
• SUNSPEC_140_HV_RIDE_THRU_MOM_CESS_DID (140)
• SUNSPEC_145_EXT_INV_CONTROLS_DID (145)
• SUNSPEC_BASIC_CC_DID (64111)
• SUNSPEC_OUTBACK_DID (64110)
• SUNSPEC_OUTBACK_FM_CC_DID (64112)
• SUNSPEC_OUTBACK_FX_DID (64113)
• SUNSPEC_OUTBACK_FX_CONFIG_DID (64114)
• SUNSPEC_OUTBACK_GS_SPLIT_DID (64115)
• SUNSPEC_OUTBACK_GS_CONFIG_DID (64116)
• SUNSPEC_OUTBACK_GS_SINGLE_DID (64117)
• SUNSPEC_OUTBACK_FNDC_DID (64118)
• SUNSPEC_OUTBACK_FNDC_CONFIG_DID (64119)
• SUNSPEC_OUTBACK_SYS_CONTROL_DID (64120)
• SUNSPEC_OUTBACK_OEM8048A_STATUS_DID (64123)
• SUNSPEC_OUTBACK_OEM8048A_CONFIG_DID (64124)
• SUNSPEC_OUTBACK_STATISTICS_DID (64255)
• SUNSPEC_OUTBACK_OEM8048A_MFG_DID (64900)

Here's the changelog. As you can see, very few changes were made to the models, most involved appending points or adjusting names or sizes for models under development. Note that the SunSpec models as published pertain to the AXS Port, which means that newer devices such as Mojave and FM100-v2 would not be present as the AXS Port (and Mate3s) did not support them.

[jsbosch]

I suggest asking OutBack Power USA for the SDK or permission for me to share it. (I thought I had deleted it, but found a backup source.)

Looking at that list, there must be a later dev version, because I don't see the 700-level models listed. I doubt I still have that version, but it probably wouldn't be all that useful anyway.

[ke6jjj]

This is great, @jsbosch ! I'll try contacting them. It's super helpful to have clues like this when making the request. We'll see if they ever answer. I had a warranty claim with an FM80 and they straight up disappeared last year -- just failed to answer anything.

[jsbosch]

The OBP website has two docs that contain SunSpec models. The two disagree with each other. One has two points inserted at offset 28, the other shows the model as published with SunSpec.