Proposed WhizBang Senior
Again, you all are digitally smarter than me. If this concept could be integrated with Midnite Charge controllers it would yield the following benefits.
1. Ability to compare multiple bank charge/discharge/float performance.
2. A better MV resolution of float currents in each bank . (@250ma float .19mv improved to 6.19mv)
3. Maintains the ability to calculate total bank currents.
I feel confident this would be a better mouse trap if the guys smarter than me can digitally integrate the hardware and software.
Humbly yours, Barry
That IC would make an awesome WB Senior !
However, you only get 2 differential input channels which is what would be needed.
Not sure if this IC allows it to be biased up in the middle of the range at 0V input to a differential channel?
Quote from: boB on March 08, 2024, 06:52:34 PMNot sure if this IC allows it to be biased up in the middle of the range at 0V input to a differential channel?
Not so sure differential inputs would be required.
Specs seem to indicate + and - 2.5 volt supply.
Hope this helps
Barry
ps more expensive differential input ad7177-2 data sheet attached
Quote from: Barry Fields on March 08, 2024, 02:43:36 PMProposed WhizBang Senior
Again, you all are digitally smarter than me. If this concept could be integrated with Midnite Charge controllers it would yield the following benefits.
1. Ability to compare multiple bank charge/discharge/float performance.
2. A better MV resolution of float currents in each bank . (@250ma float .19mv improved to 6.19mv)
3. Maintains the ability to calculate total bank currents.
I feel confident this would be a better mouse trap if the guys smarter than me can digitally integrate the hardware and software.
Humbly yours, Barry
Well I know it's not really a new product idea . This was on some threads sometime ago
[/quote]
Well I know it's not really a new product idea . This was on some threads sometime ago
[/quote]
I know I mentioned this in December in my post on parallel bank monitoring. If you are aware of previous posts concerning this concept, I would be interested in reading their rational and suggestions.
Thanks Barry
I did find some old references to WBsr asking for remote battery sensing.
The ad7177-2 could give you that flexibility.
Ya might also add Bat temp and ambient temp.
The WB Junior has ambient temperature on it now and the Classic and other controllers already have battery temperature.
What would be nice to have on a battery shunt monitor would be the remote battery voltage sense itself since it is already close to the battery terminals. You might actually be able to not connect the battery negative or that RB V sense in that case.
The new products have remote battery sense connections of course.
boB
Upon reflection, my prior suggestion of tying the low side signal of the shunts together to avoid the need for differential inputs, may be not be viable.
It would be dependent on REALLY spiffy and consistent connections between the shunts and BUSS negative.
There are two bits of good news here.
1. I stand corrected.
2. The ADFS7124-8 ($20) has multiplexing capabilities that would allow for occasional comparison of the low side shunt signal and BUSS negative. This would verify the integrity of the Shunt to Buss connections.
Just found this. Informative read.
Quote from: Barry Fields on March 29, 2024, 12:44:41 PMJust found this. Informative read.
Some times we use op-amps with shunts.
TI has many great informative papers.
I feel I should back off this topic. I do not want to be a nag (first wife,bless her heart).
I believe I have effectively put forward the advantages of a WBsenior.
1. Better resolution of Bank currents (including float currents).
2. Ability to monitor individual bank health.
3. Maintains ability to calculate coulombs in/out.
4. The flexibility to control Float current modification of Absorb time.
5. Ability to verify shunt/buss connections.
It would be nice to know if MidNite is seriously considering this option.
Quote from: Barry Fields on March 30, 2024, 12:31:29 PMI feel I should back off this topic. I do not want to be a nag (first wife,bless her heart).
I believe I have effectively put forward the advantages of a WBsenior.
1. Better resolution of Bank currents (including float currents).
2. Ability to monitor individual bank health.
3. Maintains ability to calculate coulombs in/out.
4. The flexibility to control Float current modification of Absorb time.
5. Ability to verify shunt/buss connections.
It would be nice to know if MidNite is seriously considering this option.
No, not working on or considering this right now. It would be nice but you need to realize that it costs resources, time and money to add more products. Not to mention all of the software and coordination between engineers and support and documentation.
Multiple battery banks are probably < 1.0 % of the market too. So the ROI is questionable.
You and I are part of that 1% maybe, but we are not the marketing department either.
boB
Well, I would like your post, but let's just say I appreciate your honesty.
Is there any reason I should not approach Multitel with this concept?
Quote from: Barry Fields on March 30, 2024, 03:31:35 PMWell, I would like your post, but let's just say I appreciate your honesty.
Is there any reason I should not approach Multitel with this concept?
No reason... It would be a nice product all by itself. But it really should be able to communicate with a system if it is going to help with things like Ending Amps, etc. too.
What is Multitel ? Sounds like a phone company.
https://www.multitel.com/
They already do Float current monitoring. (FCCP)
They also do modbus to snmp conversions. https://www.multitel.com/solutions/modbus-snmp-protocol-conversion/
I know little of communications protocols.
1. If a stand alone "WBsenior" had a modbus connection, could it communicate with MidNite charge controllers?
2. For my own info, could multiple op amps be connected to the same SHUNT. (my best guess is Yes)
Thanks for your input. Barry
Quote from: Barry Fields on March 31, 2024, 01:45:55 PMhttps://www.multitel.com/
They already do Float current monitoring. (FCCP)
They also do modbus to snmp conversions. https://www.multitel.com/solutions/modbus-snmp-protocol-conversion/
I know little of communications protocols.
1. If a stand alone "WBsenior" had a modbus connection, could it communicate with MidNite charge controllers?
2. For my own info, could multiple op amps be connected to the same SHUNT. (my best guess is Yes)
Thanks for your input. Barry
Yes and yes
We were thinking about doing snmp at one time
two questions if I may.
1. the WBjr circuit board is designed to work only with a 500a 50mv shunt. I do not see adjustments for other shunts.
2. Am I correct in saying that MODBUS only describes the interconnection of units. SPI and uart can be used on that buss? If so what serial protocol does Midnite use?
Please edumacate me.
Thanks Barry
The WbJr has programmable gain and is turned up all the way for 50mV/500A. That means it will work for smaller shunts but adjustments would have to be made on the Classic.
It works the way it is now with any scaling of that 50mV/500A. We also use 100mV/1000A and 200mV/2000A as is. Good to 2500A DC.
The serial protocol is ours but basically 8 bits 1200 baud command and then the WbJr responds with the values and CRC. Simple. In between commands, the WbJr is powered by the Classic.
I hear you saying that any ADC wishing to communicate would need to be programed with MidNite's serial protocol.
Is the same protocol used with all MidNite's charge controllers?
Quote from: Barry Fields on April 01, 2024, 06:10:45 PMI hear you saying that any ADC wishing to communicate would need to be programed with MidNite's serial protocol.
Is the same protocol used with all MidNite's charge controllers?
Yes and Yep.
If the ADC has an internal processor, then it would have to be programmed to and interfaced like the WB Jr.
Any reason why the attached would not work?
I have attached the LTspice file.
Quote from: Barry Fields on April 02, 2024, 08:10:07 PMAny reason why the attached would not work?
I have attached the LTspice file.
What is the electrical relationship between the separate shunts and batteries ?
As per your request.
Ahhh yes. Thank you.
Yeah, it would probably work fine for that. But you can't tell which battery you are looking at since they are all combined into one amplifier.
750 uO is kind of high for the typical home or industrial power battery system but will work of course.
boB
Quote from: boB on April 04, 2024, 11:28:05 PMAhhh yes. Thank you.
Yeah, it would probably work fine for that. But you can't tell which battery you are looking at since they are all combined into one amplifier.
I will get to individual monitoring in a minute.
750 uO is kind of high for the typical home or industrial power battery system but will work of course.
I use that because I plan to use 100A/75mv shunts. The circuit should work with with any 4 equal shunts. The divider R14/R15 would be select-able for 50mv vs 75mv shunts.
Before I continue, is there any combination of charge current/SOC/Rconnection that would put U1 at risk? Would different resistor values for U1 preclude that? I would not want a poor connection between the shunt and negative buss to smoke U1.
Thanks for your input.
Only thing I can think of that would be bad would be HIGH current that would place the input to U1 too far out of its common mode voltage range. Any battery current that is normal should be OK. A lower shunt resistance might help but at the expense of lower resolution on the amps being measured.
You could also raise those 1K Ohms to something higher, maybe 5K Ohms
I have added the 50mv/75mv shunt selector switch and changed the R's to 5k.
Just for conversation I added a couple of 3.1v zenner diode with indicator leds.
Would this further protect the excess common mode excursions?
As far as the voltage drop across the Shunt to neg buss connection, I would expect it to be it's worse on a high discharge current (100A-200A). Would this not cause a high negative voltage on the low side shunt signal.
What voltage might one expect or want to limit during a high discharge?
I apologize for thinking out loud as I post.
I have rethought the LED thing, never mind.
I have included the shunt to buss resistance in the schematic. I reckon all the Rconn's have to be the same or very close.
I have attached a schematic and LTspice for a 555 latching ckt which would latch an overly negative voltage across the shunt to buss connection. It would usually activate on a large discharge current.
The negative voltage at which it would trigger is up to debate.
I chose R5 & R6 for a trigger around -2 volts. Obviously a lower voltage would be necessary for a 12v system.
V1 duplicates a reset signal for evaluation only. I would suggest a manual reset button.
The circuit would be duplicated for each battery bank. I am not familiar with the quad 558 IC which would probably be usable.
C1 could be labeled LLJ1 (little leyden jar) ;D
As I have not received any "OMG you can't do that" responses, let me proceed.
The prior schematics of the bank combiner, the connection monitor and this configuration would all share the same bank shunts.
Below is attached a method for a monthly evaluation of Bank Health. A total KWH charge data for each bank could be compared and reset monthly.
The performance of the WBjr should not be affected (save not needing its large shunt). It would still deliver total Battery array data and still allow for float current controlled Absorb time (nudge,nudge)
I don't like to use zeners, and especially low voltage zener diodes for things like protection because they can be very leaky. 5K Ohms is not real high so it might be acceptable at that low impedance.
You know... Summing all 4 shunts together only tells you what one bigger shunt does. It doesn't tell you if all 4 banks are sharing.
boB
Quote from: boB on April 10, 2024, 01:24:00 AMYou know... Summing all 4 shunts together only tells you what one bigger shunt does. It doesn't tell you if all 4 banks are sharing.
boB
Below is attached a method for a monthly evaluation of Bank Health. A total KWH charge data for each bank could be compared and reset monthly. (my previous post)
After reflection, the acceptable limit of voltage drop across the shunt to buss connection should likely be in the area of -100mv. We already accept 50mv to 75mv across the shunt itself. (open to discussion)
U2 enables better trigger resolution of U1 (555 flip flop). R5 & R8 set the trigger point. As is it is approx -100mv. R5 & R8 should be replaced with a potentiometer for adjustment.
Just so I understand.
If I use the Wbjr, I lose the ability to PWM. TRUE?
If I have 2 classic 150s in follow me, can I use the aux2 port on the second 150 to PWM?
The communication between 150s is Modbus. TRUE?
The first 150 is the master and the second is a unique addressable slave. TRUE?
Can the slave query the master for data? (ie SOC, set point, wbjr current)
Could another non-150 unique addressable slave do the same?
Just trying to get smarter. May the 4th be with you.
Barry
Yes, Aux 2, the only PWM port is also used for WB Jr so pick one.
Yes, you can hook up the WB Jr. to either one.
When you say MASTER, the only thing that is master about it is the battery temperature sensor reading comes from that Classic.
The default modbus address for each Classic is 10. You can change it up or down. Usually the main Classic (whatever that really means?) stays at ten (10) and the next one is usually 11. Then it is easier to look at the second Classic with the first Classic's MNGP and vice-versa.
There is a special button push combination to tell the Classic's processor that it's modbus address is now what the MNGPs present address is. I can't remember if this particular action also changes the default MNGP address (both saved in their own EEprom) I think that is the case.
Classic and MNGPs communicate back and forth of course but there are only certain things that are done. All the information and settings for a Classic could be transferred over the modbus but it is limited.
What you could also do is take some small single board computer, add modbus or just use TCP/IP Ethernet (much easier I think) and you could write a program to transfer all sorts of data with your own protocols. All the features that we put in and the ones we didn't put in :) ;)
boB
Just a couple more.
MNGP allows a forced BULK. I presume that WBjr can force FLOAT. Can ABSORB be forced in the same way?
If true can that be done on the MODBUS?
Are PWM setpoint & width, or PWM% and SOC available over MODBUS?
Can WBjr data be injected over MODBUS?
;D ;D Sure Barry just get a computer and program it. ;D ;D ;D
As we progress here, I will need some help.
Best regards Barry
The WB Jr. information, which is basically just amps is available over modbus.
You can force a bulk/absorb or float using modbus.
This is what Follow-Me does. It sends the charge stage that Classic is in.
So the other units change to Float if one of the other Classics goes to Float.
The other units will also go to Bulk/Absorb when another Classic on the modbus goes to Bulk/Absorb if, say, you force a Bulk in the MNGP.
The MNGP is just a "dumb terminal" basically where it reads the requested number or mode or whatever from the Classic and the Classic tells the MNGP the new number.
So, it is all done over the RS-232 modbus and most of the registers are available on our MODBUS document. It should be updated but the WB Jr. stuff is in there too at least as well as force Bulk or Float or EQ.
https://www.midnitesolar.com/pdfs/classic_register_map_Rev-C5-December-8-2013.pdf
boB
When dealing with 2 classics in "follow me" is a firmware update reflected through to the second classic over the modbus or is a second usb connection required for an additional update for the second classic?