Thank you so much for your work on the diyBMS and keeping it open source! I can't wait for you to receive the current monitoring board and hopefully greenlighting it for the rest of us to order.
Thanks a lot for your work... I didn't need one shunt for now, but when will be ready I'll buy/make one just to see and learn something new :-) After that, the next can be a DIYBMS-MPPT project :-) Congrats again !!!
If the shunt gets hot it would be best to isolate it from the PCB, as to not heat up the board. That is, produce temperature gradients across the PCB. Temperature gradients can influence sensitive voltage measurement circuitry through the Seebeck effect. On the PCB there are many different types of metals connected together (like a thermocouple), so the only way to avoid these thermal EMFs is to make sure all areas of the board are at the some temperature (it is best to look at the final design with a thermal camera to make sure any temperature gradients are kept to a minimum around the most sensitive nodes). In some very sensitive instrumentation you will see sensing circuit PCBs with larger and many internal layers. This is done to spread out heat evenly on the board. You may even see the PCB contain an Aluminum metal core, or the entire board (circuitry only on top), bonded to a 0.250" aluminum plate. You can connect the shunt to the PCB using a wire or a flat bus bar. If the shunt and PCB are rigid within the same enclosure a rigid bus bar will not vibrated as easily as a wire. A low value shunt wire connection must be held rigid as to not vibrate. If the installation is not subject to vibration than a wire will work. I like making custom bus bars because once soldered to the board they stay in position relative to the shunt sense terminals, when you have to separate the two. Why? for high gain circuits you can calibrate the circuit with the first stage gain or just have a huge selection or shunts available from the vendor and pretested as to measured value. Then you can select the shunt that best matches the amp, and not have to calibrated (change a resistor) the gain. So, on a production line you would have a selection of shunts and amps, and then have a calibration procedure to just match them up. With the worse case being having to change a gain setting resistor for the first stage instrumentation amp.
Hopefully it works first time. I am also waiting on JLCPCB for 5 of your controller boards. Looks like Digikey for the attinys and other parts. They emailed me that they now ran out of the 1.25V reference chip.
Hello First of all, thank you for the DIY BMS project i also have BATRIUM and i have more problems with it than with DIY BMS. I am asking (due to problems with parts) to answer if it is possible to use it for the DIY BMS shant BATRIUM project? I have an extra one, hence the possibility of using it Thanks in advance for your answer Regards, Jerzy
Doesn't matter to the shunt. The rest of the system might care for various reasons. On my system I cannot put a shunt in the negative because on my system the negative is tied to earth ground as are various other places in the system.
Mike, technically the INA228 supports both options. Reading through this guide, it seems that high side has more benefits - www.microchip.com/en-us/about/blog/developer-insights/high-side-versus-low-side-current-sensing
@@StuartPittaway Yes, it supports both and there are definitely positives and negatives to either option. I just wanted to confirm that it is intentional to use the POS terminal of the battery. I'd suggest having a very clear assembly diagram as part of the documentation for this device. When you have this ready to produce, in mass I'd be happy to help with a USA order (same as was done for the v4.2 controller PCB). I have the tools (and parts!) to assemble these in bulk as well. One Q, since you have SMT components on both sides of the PCB, are you having JLCPCB assemble the back side only?
Stuart, any progress on this? I don’t see anything posted on your GH. I’m similarly have started to work on a power monitoring PCB using the INA228AIDGST and an ESP32. My initial plan is to physically separate the shunt away from the PCB as this adds a lot of flexibility to use any shunt.
@@StuartPittaway I can think of MANY ways to calibrate. More important is long-term stable results! For example a "short burst" pulse of a load. Once you have a few points of the curve the rest is interpolated. The "load shunt" resister should have a stable curve over any style of the fuse. Or a dummy load for the full long-term load numbers. I use a 12v heater!
how about an open loop current sensor instead of a shunt? TI has a 100amp busbar demo unit using a pair of their DRV421. PDFs to google for: TIPD205 describes the demo; SLOA237 for the Theory.
Hi Stuart great info loved the way you explained very easy to understand keep it up. I need your help to design on BMS for my own business could you please help me with?
@@StuartPittawayThank you. So for an 800V lifepo4 battery i will need to do 3 batteries connected in series. I guess there is no reason why it should not work.
Stuart, I'm curious ... did you consider a hall effect current sensor vs shunt? I'm using shunts now and have been thinking hall effect would be the way to go given 2 fewer physical connections in the main power path ... plus no extra heat / power drain.
![Current Monitor Progress [DIYBMS]](http://i.ytimg.com/vi/gtKEEv8j9DU/mqdefault.jpg)
![Current Monitor Progress [DIYBMS]](/img/tr.png)
![[UNCUT] เก็บ "สจ.โต้ง" เอี่ยวมาเฟีย รู้ตัวคนบงการ ก่อนสั่งลั่นไก I คนดังนั่งเคลียร์ I 13 ธ.ค.67](http://i.ytimg.com/vi/o-J_p62D40w/mqdefault.jpg)
Looks realy good...all the parts we need...they talk to each other....reasonably priced...we can build a sensible ess and monitor everything! LOVE IT!
INA228 is a wonderful ADC, nice to see that you are going to use it.
Looks good to me, excellent specifications for the money
@@StuartPittaway Yes I have used it myself for a WiFi Voltmeter
Thank you Stuart! So been waiting for this. Since the controller, I was searching how to make or design one!
Me too! This is a fantastic development by Stuart.
@@MiniLuv-1984 Glad you like it - lets hope it works!
@@StuartPittaway I'm pretty confident Stuart. The guy that designed and is building it has a pretty amazing track record...you may know him? :)
Looking good. All the pieces are coming together :-)
Thanks Adam.
Thank you for that video! I didn't know the INA228. Interesting device.
Cheers, Renee
You're welcome!
cant wait for it to be finished
Thank you so much for your work on the diyBMS and keeping it open source! I can't wait for you to receive the current monitoring board and hopefully greenlighting it for the rest of us to order.
I'll be logging a complaint with JLC I think!
truely nice work stewart...
Thank you kindly
thanks Stuart
Thanks Stuart
Thanks a lot for your work... I didn't need one shunt for now,
but when will be ready I'll buy/make one just to see and learn something new :-)
After that, the next can be a DIYBMS-MPPT project :-)
Congrats again !!!
Great.
Great, thanks Ttuart.
Awesome job Stuart
Thank you kindly
Great, and looking good :)
Thanks! 😊
If the shunt gets hot it would be best to isolate it from the PCB, as to not heat up the board. That is, produce temperature gradients across the PCB. Temperature gradients can influence sensitive voltage measurement circuitry through the Seebeck effect. On the PCB there are many different types of metals connected together (like a thermocouple), so the only way to avoid these thermal EMFs is to make sure all areas of the board are at the some temperature (it is best to look at the final design with a thermal camera to make sure any temperature gradients are kept to a minimum around the most sensitive nodes).
In some very sensitive instrumentation you will see sensing circuit PCBs with larger and many internal layers. This is done to spread out heat evenly on the board. You may even see the PCB contain an Aluminum metal core, or the entire board (circuitry only on top), bonded to a 0.250" aluminum plate.
You can connect the shunt to the PCB using a wire or a flat bus bar. If the shunt and PCB are rigid within the same enclosure a rigid bus bar will not vibrated as easily as a wire. A low value shunt wire connection must be held rigid as to not vibrate. If the installation is not subject to vibration than a wire will work. I like making custom bus bars because once soldered to the board they stay in position relative to the shunt sense terminals, when you have to separate the two. Why? for high gain circuits you can calibrate the circuit with the first stage gain or just have a huge selection or shunts available from the vendor and pretested as to measured value. Then you can select the shunt that best matches the amp, and not have to calibrated (change a resistor) the gain. So, on a production line you would have a selection of shunts and amps, and then have a calibration procedure to just match them up. With the worse case being having to change a gain setting resistor for the first stage instrumentation amp.
They are all very good points you raise. The shunt doesn't really get hot - and its also insulated by the plastic base.
Great work
Thanks
Thanks man!
No problem!
So great - keep it up
Thank you! Will do!
Could you make a video for the shunt assemble, thank you. Keep up the great work you doing!!
Thanks for the idea!
Hopefully it works first time. I am also waiting on JLCPCB for 5 of your controller boards. Looks like Digikey for the attinys and other parts. They emailed me that they now ran out of the 1.25V reference chip.
They do seem to be having issues with parts, looks like a global issue of semiconductors.
Hello
First of all, thank you for the DIY BMS project
i also have BATRIUM and i have more problems with it than with DIY BMS. I am asking (due to problems with parts) to answer if it is possible to use it for the DIY BMS shant BATRIUM project? I have an extra one, hence the possibility of using it
Thanks in advance for your answer
Regards, Jerzy
No I don't think that's possible
Aren't most shunts using the battery negative lead rather than the positive?
Doesn't matter to the shunt. The rest of the system might care for various reasons. On my system I cannot put a shunt in the negative because on my system the negative is tied to earth ground as are various other places in the system.
Mike, technically the INA228 supports both options. Reading through this guide, it seems that high side has more benefits - www.microchip.com/en-us/about/blog/developer-insights/high-side-versus-low-side-current-sensing
@@StuartPittaway Yes, it supports both and there are definitely positives and negatives to either option. I just wanted to confirm that it is intentional to use the POS terminal of the battery. I'd suggest having a very clear assembly diagram as part of the documentation for this device.
When you have this ready to produce, in mass I'd be happy to help with a USA order (same as was done for the v4.2 controller PCB). I have the tools (and parts!) to assemble these in bulk as well.
One Q, since you have SMT components on both sides of the PCB, are you having JLCPCB assemble the back side only?
Awesome job Stuart, board is very nice but can I use a 300A 50mv shunt with this board ?
I already have one sitting around.
Thanks.
No reason why not.
Stuart, any progress on this? I don’t see anything posted on your GH.
I’m similarly have started to work on a power monitoring PCB using the INA228AIDGST and an ESP32. My initial plan is to physically separate the shunt away from the PCB as this adds a lot of flexibility to use any shunt.
Yes, its fully complete and finished - github.com/stuartpittaway/diyBMS-CurrentShunt
@@StuartPittaway Hello. Is it possible to connect two or more current monitors?
@@kaldeis its not currently supported, although there isn't a hardware restriction, its just not written into the controller code.
@@StuartPittaway Do you plan to add this functionality?
@@kaldeis If there is enough interest in it, I can. I think you are the first to mention it!
Question: Why not use a Fuse? You need one in the path. No extra loss. Cheap. Calibrate for each Fuse swap.
A fuse instead of the shunt do you mean? Fuses are not that accurate, how would you calibrate it?
@@StuartPittaway I can think of MANY ways to calibrate. More important is long-term stable results! For example a "short burst" pulse of a load. Once you have a few points of the curve the rest is interpolated. The "load shunt" resister should have a stable curve over any style of the fuse. Or a dummy load for the full long-term load numbers. I use a 12v heater!
@@myCloudWatcher How linear do you think a fuse is?
Could you post a link to the schematic?
Did you need to separate analog and digital grounds?
Hi, I've not published the circuit yet, effectively it's just the reference design from the datasheet. Doesn't need separate grounds
how about an open loop current sensor instead of a shunt?
TI has a 100amp busbar demo unit using a pair of their DRV421. PDFs to google for: TIPD205 describes the demo; SLOA237 for the Theory.
It could work, however this only gives you the current - we also need an accurate voltage to calculate the power and energy used
Hi Stuart great info loved the way you explained very easy to understand keep it up. I need your help to design on BMS for my own business could you please help me with?
What is the max number of series connections you can do with the diybms
Technically over 100 are supported, but for performance I'd recommend less than 40
@@StuartPittawayThank you. So for an 800V lifepo4 battery i will need to do 3 batteries connected in series. I guess there is no reason why it should not work.
@@arhitektche should work, by I really really really recommend you don't stick to lower safe voltage less than 50V DC
@@StuartPittaway Thank you for the advise!
Stuart, I'm curious ... did you consider a hall effect current sensor vs shunt? I'm using shunts now and have been thinking hall effect would be the way to go given 2 fewer physical connections in the main power path ... plus no extra heat / power drain.
You still need to measure the voltage, shunts give you both options and I think they are more accurate