Page 7 of 11

### Re: new program for monitoring charge readings

Posted: Wed Mar 27, 2019 9:56 am
while in progress

### Re: new program for monitoring charge readings

Posted: Fri Mar 29, 2019 2:00 am
DBMandrake wrote:In these conditions Canion reports 4.075 - 4.085 volts (correct) and Hobdrive beta's 88 cells page reports values from 3.95 - 3.98 volts. (incorrect) It's simple knowledge that 3.95 volts is not a fully charged lithium Ion battery as these cars charge the cells to exactly 4.1 volts, so 3.95 - 3.98 cannot be correct.

I don't know how you're calculating the voltages but for my 2011 Ion they are simply wrong and do not agree with the official Peugeot diagnostic tool. The error seems to be greater at low state of charge. The calculation in previous versions of Hobdrive was correct.

Voltage can be read separately from each block of 1-12 CMU and displayed on a separate screen, i.e. there will be 12 requests for 12 different blocks. Measurement step in these block is 0.005
And you can read a single request from the block BMU. Measurement step 0.01.
Replace files from this topic
http://hobdrive.com/forum/viewtopic.php ... 100#p30361
and compare the Diagram (CMU) and Diagram (BMU) screens

### Re: new program for monitoring charge readings

Posted: Thu Apr 11, 2019 6:12 pm

### Re: new program for monitoring charge readings

Posted: Thu Apr 11, 2019 6:27 pm
i think what is puzzling to everyone is why do the BMU and CMU voltages not match up exactly--there is about 0.100 volts difference between the two for the same cells.

Which value is correct, and why is there an indicated difference in the readings?

This is a problem that needs to be answered before anyone will have confidence in the app or the data.

### Re: new program for monitoring charge readings

Posted: Fri Apr 12, 2019 12:57 am
kiev wrote:i think what is puzzling to everyone is why do the BMU and CMU voltages not match up exactly--there is about 0.100 volts difference between the two for the same cells.

Which value is correct, and why is there an indicated difference in the readings?

The BMU values (min voltage and max voltage) are correct and match Canion and Diagbox. The individual cell values in the 88 cell screen are incorrect. The error is actually much greater than 0.1v at a lower SoC, as I showed in my earlier screenshots.

The formula to calculate the cell voltages on the 88 cell screen was correct in the last stable release of Hobdrive but is wrong in the current beta releases. I hope Serhge will revert to the old calculation which has been proven to be correct.

I think Serhge is skeptical of the values reported by Canion however I have a Diagbox/Lexia 3 official dealer diagnostic tool of my own which says (at least on my 2011 Ion) that Canion reports exactly the right individual cell voltages (within 5mV) while current Hobdrive beta builds are way out and read too low.

Perhaps early and later model cars use a different calculation - this could be a source of confusion and error among those trying to reverse engineer the formula to calculate the voltages from the PID values.
This is a problem that needs to be answered before anyone will have confidence in the app or the data.

This is my concern as well. Most of the data reported in Hobdrive is correct (apart from the 88 cell voltages in current beta builds) however I have noticed a few individual parameters that are incorrect or have very implausible readings. When I get time I'll compile a list of the parameters with questionable accuracy by comparing them to my Diagbox unit.

### Re: new program for monitoring charge readings

Posted: Fri Apr 12, 2019 1:13 am
I initially said that the CMU voltage readings were incorrect but I've just discovered that isn't the case. (So I've amended my original post)

In fact the individual CMU reported voltages are correct, it's only the 88 cells page (which shows all cells on one page instead of per CMU) that is incorrect. See below:

Having different results from these two pages is obviously a concern...

### Re: new program for monitoring charge readings

Posted: Sat Apr 13, 2019 2:28 am
DBMandrake wrote: ... it's only the 88 cells page (which shows all cells on one page instead of per CMU) that is incorrect.

I believe that he needs to multiply those values (ignoring the decimal point) by 5 (for 5 mV per sample), and add 2100, to get the result in mV.

For example, 397 x 5 + 2100 = 4,085 mV = 4.085 V.
396 x 5 + 2100 = 4,080 mV = 4.080 V.

These then agree with the other page.

The problem is that some of the values come from the CMU, and others from the BMU, and I believe that they are encoded differently. I have no idea why the encoding differs.

[ Edit: corrected typos, 4,095 mV → 4,085 mV; 4.095 V → 4.085 V. ]

### Re: new program for monitoring charge readings

Posted: Sat Apr 13, 2019 5:44 pm
values ​​in the test hobd. ** 45 values ​​from the converter formula.
The values ​​of all volts of cells from cmu blocks, as well as mut3.
In the original Mut3, the voltage values ​​of all cells are present only for titanium, and there the formula for calculations is not the same as for lithium-ion. therefore, both readings are presented in this test version. I also have a professional scanner, mut3, X431, so I have something to compare with.
and moreover, if you put three numbers after a comma and compare them with numbers after a comma with two, an error occurs against the general background, and this is normal, but still, bolts with three values ​​after multiplying a comma: 3.000 , 3.005 , 3.010 .
so did the manufacturer and do nothing with it.

### Re: new program for monitoring charge readings

Posted: Sun Apr 14, 2019 12:11 pm
coulomb wrote:
DBMandrake wrote: ... it's only the 88 cells page (which shows all cells on one page instead of per CMU) that is incorrect.

I believe that he needs to multiply those values (ignoring the decimal point) by 5 (for 5 mV per sample), and add 2100, to get the result in mV.

For example, 397 x 5 + 2100 = 4,085 mV = 4.085 V.
396 x 5 + 2100 = 4,080 mV = 4.080 V.

These then agree with the other page.

The problem is that some of the values come from the CMU, and others from the BMU, and I believe that they are encoded differently. I have no idea why the encoding differs.

[ Edit: corrected typos, 4,095 mV → 4,085 mV; 4.095 V → 4.085 V. ]

Very good idea.
How did I immediately did not pay attention)
The same formula is used to calculate the min and max voltage

Replace the file from this post:
http://hobdrive.com/forum/viewtopic.php ... 770#p30770

### Re: new program for monitoring charge readings

Posted: Mon Apr 15, 2019 6:17 am
arbeitet
works