PIP-4048MS and PIP-5048MS inverters

non-EV Solar, Wind and other renewable power sources
non-EV batteries and other energy storage stuff
Forum rules
Important!
This forum is for discussion of Non-EV matters.
hennejg
Noobie
Posts: 5
Joined: Wed, 30 Aug 2017, 05:36
Real Name: Jörg Henne

Re: PIP-4048MS inverter

Post by hennejg »

*sigh*
I just realized that THE BUG is only part of the problem. The insufficient range of parameter 29 (low battery cut off) is the other.

Which raises another question: is there an elegant way of disabling a) the charger and b) the inverter in response to demands from the BMS (a just as another level of safety, b as a matter of necessity) without using a contactor or a beefy SSR?
For b) a relay can be connected in serial to the inverter power switch, but for a) my best idea for now is to interrupt the PV inputs.
User avatar
weber
Site Admin
Posts: 2936
Joined: Fri, 23 Jan 2009, 17:27
Real Name: Dave Keenan
Location: Brisbane
Contact:

Re: PIP-4048MS inverter

Post by weber »

hennejg wrote: Wed, 30 Aug 2017, 06:15 * Will it even work with LiFePO4 cells which I naively just assumed to be the case?
Don't know.
* Is the bug present on the 24V models as well?
Don't know.
* Do you know of any resources (similar to this forum) dedicated to the 24V models?
No.
* Right now, I could probably make the switch to 48V and sell my 24V box on eBay, but after I've got the battery that makes no sense anymore, so the question is: is it worth it?
We can help you with the 48 V model. We can't help you with the 24 V model. I understand the 24 V model is very different. I think it even uses a different microprocessor architecture.
just realized that THE BUG is only part of the problem. The insufficient range of parameter 29 (low battery cut off) is the other.
Right. Also the fact that the unpatched code requires the voltage to go 4 volts below the float setting (in the 48 V case) before it will start another bulk/absorb charge cycle.
Which raises another question: is there an elegant way of disabling a) the charger and b) the inverter in response to demands from the BMS (a just as another level of safety, b as a matter of necessity) without using a contactor or a beefy SSR? For b) a relay can be connected in serial to the inverter power switch, but for a) my best idea for now is to interrupt the PV inputs.
It depends what you consider elegant. But surely patching the code gives the most elegant solution.
One of the fathers of MeXy the electric MX-5, along with Coulomb and Newton (Jeff Owen).
User avatar
coulomb
Site Admin
Posts: 6357
Joined: Thu, 22 Jan 2009, 20:32
Real Name: Mike Van Emmerik
Location: Brisbane
Contact:

Re: PIP-4048MS inverter

Post by coulomb »

hennejg wrote: Wed, 30 Aug 2017, 06:15 * Is the bug present on the 24V models as well?
It's hard to guess. Most likely, the two models share some common code, so the likelihood is that the bug exists in the 24 V models.

But surely you could tell us that. I would think that Weber's test would apply to the 24 V model, or at least could be adapted. [ Edit: perhaps I misread your post, and you don't have the battery yet. It will still be good to know. ]
MG ZS EV 2021 April 2021. Nissan Leaf 2012 with new battery May 2019.
5650 W solar, 2xPIP-4048MS inverters, 16 kWh battery.
Patching PIP-4048/5048 inverter-chargers.
If you appreciate my work, you can buy me a coffee.
hennejg
Noobie
Posts: 5
Joined: Wed, 30 Aug 2017, 05:36
Real Name: Jörg Henne

Re: PIP-4048MS inverter

Post by hennejg »

Thanks a lot @weber, @coulomb for your responses!

I've decided to switch to 48V and ordered a PIP-4048MS and the battery accordingly. Now I can only hope that your patches still work with the recent models.
User avatar
coulomb
Site Admin
Posts: 6357
Joined: Thu, 22 Jan 2009, 20:32
Real Name: Mike Van Emmerik
Location: Brisbane
Contact:

Re: PIP-4048MS inverter

Post by coulomb »

hennejg wrote: Fri, 01 Sep 2017, 02:43 Now I can only hope that your patches still work with the recent models.
Yes, that's something we'd like to know as well. If your machine comes with firmware version 73.00, then you're definitely safe. We have successfully downgraded PIPs from 73.00 to 72.70b patched firmware. [Edit: We no longer recommend putting version 72.xx (not even our patched versions) into inverters that are supplied with 73.00. We recently discovered that this causes a problem with a relay that has been added inside these new machines, that can connect Neutral to Earth on the AC output. We now have a 73.00 firmware file.] If it's 73.XX, you are probably still safe. If its 74.XX or later, then it's unclear whether the downgrade will work, and what new features you might be throwing away. There is presently no way of going back to factory firmware other than asking nicely for it. I think in a situation where you tried to correct a defect that they created, and in the process bricked or severely limited your machine, then presumably they would provide the firmware file. But it is a risk.

Of course, there is the chance that the charging bug will be fixed in 73.XX where XX > 00, or in 74.YY. But it seems unlikely to me that they will also have fixed the other issues affecting lithium batteries, such as needing to see a 4.0 V drop in battery voltage before re-starting a charge cycle (we patched that to 1.0 V [Edit: Now 1.2 V]).
MG ZS EV 2021 April 2021. Nissan Leaf 2012 with new battery May 2019.
5650 W solar, 2xPIP-4048MS inverters, 16 kWh battery.
Patching PIP-4048/5048 inverter-chargers.
If you appreciate my work, you can buy me a coffee.
Brandt66
Noobie
Posts: 6
Joined: Sat, 02 Sep 2017, 19:58
Real Name: Joerg Brandt
Location: Thailand

Re: PIP-4048MS inverter

Post by Brandt66 »

Hello

I am running a solar system from 12KW with 3 PIP-4048MS Inverters in parallel the battery is an 48V 1300AH Batterie. After I found this website I update the firmware to the “dsp_Pb1_72.70b” and install the monitoring software ICC. Now I get aware that during the system run on Batterie the Inverters take for a sort time High currents peaks up to 500A without that there is an equivalent load.

https://1drv.ms/i/s!Ai20LeuVVwNVltg6C__5BnayuoAwjg

Is here in this forum somebody that could give me the advice this is normal for this type of Inverter or what can I do to stop this behavior ?
User avatar
coulomb
Site Admin
Posts: 6357
Joined: Thu, 22 Jan 2009, 20:32
Real Name: Mike Van Emmerik
Location: Brisbane
Contact:

Re: PIP-4048MS inverter

Post by coulomb »

Brandt66 wrote: Mon, 04 Sep 2017, 20:06Hello
Hi, welcome to the forum.
Is here in this forum somebody that could give me the advice this is normal for this type of Inverter or what can I do to stop this behavior ?
That sure doesn't look right to me. My guess is that somehow, the monitoring software is sometimes getting garbled responses. There are probably not all that many three-inverter systems running ICC, and maybe the timing needs to be changed a bit to cope with the delays caused by three inverters (if that's the case).

But first, it would be really good to rule out that those current spikes are really happening.

You can probably find this out from the instantaneous battery voltage. Do you see the battery voltage sag in time with the purported surges in battery load? A 250+ amp drain will cause a noticeable voltage sag on pretty much any battery.

The other things is you could watch the battery current on one of the LCD displays. Some of those apparent surges are about 10 second in duration, so you should be able to see them on the display if they are real. Almost certainly, they are not, so this is a result of the monitoring software not working with your setup.

You probably need to work with the author to get this resolved. I'm pretty sure that you will get good results asking questions or sending private messages on the following forum: http://powerforum.co.za/forum/94-axpert ... -software/ .
MG ZS EV 2021 April 2021. Nissan Leaf 2012 with new battery May 2019.
5650 W solar, 2xPIP-4048MS inverters, 16 kWh battery.
Patching PIP-4048/5048 inverter-chargers.
If you appreciate my work, you can buy me a coffee.
paulvk
Senior Member
Posts: 457
Joined: Fri, 23 Oct 2015, 23:45
Real Name: Paul
Location: Sydney

Re: PIP-4048MS inverter

Post by paulvk »

I have been running the Pb version of the patched software on my two parallel PIPs for a couple of months.
They still tend to go back to float a bit early.
I have just installed the 130v version of I-Panda 40a MPPT charger on the battery so now I have the two PIP 60A chargers and the I-Panda 40A
The 40A unit is doing a much better job of bulk charging, its working very well with the PIPs so those with a 60A PIP could use it to get to 100A MPPT and fix the bug in the PIPs .
The 40A unit is a esmart series 3 $160 AU on e-bay.
I have 9 x 250w panels on it , one PIP has 12 x 240w the other 15 x 240w.
User avatar
coulomb
Site Admin
Posts: 6357
Joined: Thu, 22 Jan 2009, 20:32
Real Name: Mike Van Emmerik
Location: Brisbane
Contact:

Re: PIP-4048MS inverter

Post by coulomb »

Revlac wrote: Fri, 25 Aug 2017, 20:56 Now I fitted a ferrite Round cable suppression sleeve around the cord of the appliance, tried a few sizes best results form the small one that fits snug around the cord. less noticeable flicker on the house lights and no bounce on the battery charger volt/amp meter. All Good
Just thought someone may find that helpful.
I remembered your post about the ferrites when I was called to a charger that was making a terrible audible racket some of the time. It was a very harsh sound, most unpleasant. So I tried adding ferrite cores around the thick cables from the charger to the battery. (There were some already installed in the bowels of the boat where the battery and charger were located.) This was a 6 kW 24 V charger, so these were 35 mm² cables (quite thick). The sound changed; it was lower in frequency and not so harsh. The customer had more ferrites, so we clamped a few more on. Each time the frequency changed, going lower and lower with more and more ferrites. Finally, we removed all the ferrites from the battery cables, and the acoustic noise stopped. I think that the added inductance, instead of smoothing the electrical switching noise as I expected, seemed to induce some sort of instability. I'm not saying that what you did is bad, far from it. But under some circumstances, ferrites might not be the best idea. Fortunately, they are usually designed to be easy to add and remove with clips, so you can just try them and see what happens.
MG ZS EV 2021 April 2021. Nissan Leaf 2012 with new battery May 2019.
5650 W solar, 2xPIP-4048MS inverters, 16 kWh battery.
Patching PIP-4048/5048 inverter-chargers.
If you appreciate my work, you can buy me a coffee.
ghatikar
Noobie
Posts: 2
Joined: Thu, 28 Sep 2017, 23:31
Real Name: Anil Ghatikar

Re: PIP-4048MS inverter

Post by ghatikar »

andys wrote: Thu, 17 Aug 2017, 14:49 Just tried a new generator - its a Honda 13HP petrol with 8kW alternator. Reading a strong 240VAC while maintaining 50 amps PIP charge rate, for quite some time before dropping out. Then it gets worse and worse, shorter periods until the next drop out.

I tried removing all other AC loads and it just seemed to make it worse - now only making it up to 40A charge rate before dropping out.

So, still not sure what is going on.
Greetings from Abuja Nigeria - Found the forum through Power forum South africa when I was looking for tips on integrating LIfepo4 with my MPP _Solar PIP MS4048
I have used the PIP-HS/.MS and different variations since 2013 when they first came out. They have a peculiar habit when working with most of the generators. The reasons is the High-frequency switching of the Generator results into waveform distortion. In some cases, it is so bad that you can see the flickering if you connect a light bulb.

See the attached waveform where a PIP MS 4048 is connected to KIPOR ( UK) 7.5 KVA 3 phase generator with charging current set to 30 Amps ( 1.5 KW). The issues are not with the Generator per say but the excitation system /AVR used in the generators which do not cope well with the rapid changes resulting from the unpredictable behavior.
You cannot see this behavior in normal meters designed to capture RMS currents and voltages ( as rms values by nature are average over a period of 20 ms)

The simplest solution at a fraction of cost is to add steady resistive load ( at least 25 % to 35 %) of the charger load - Bulbs or room heaters
e.g. PIP set as 50 Amps then 50 amps x 54 V /0.9 = 3KW so the resistive load will be at least 1 KW and the generator should have 70 % load means generator will be 3 KW +1 KW/0.7/0.8 (PF)= 7.14 KVA . Note that MPP solar recommends 1.5 times the rated capacity of inverter at least because of the same reason

In borderline cases, You will notice that as the terminal voltage of the battery increases over time the Inverter may stop charging and resume charging starting from zero - Easily distinguishable by a sudden increase in Generator sound and again slowing down as PIP increases the current. Can be solved by reducing the charging current

In one case of Hyundai 5 KVA generator, I have a peculiar issue where the AVR was blowing off till I put meters on it. Turns out during switching on of charging stage due to capacitors being added the terminal voltage raises as much as 15 V for few seconds.

In short for those of you having problem


1. Start at max 30 % load of your generator and see if it runs stable if not step 2
2. Add resistive load _ I normally use 100-watt bulbs as needed - Bulbs should come on earlier or together with inverter
3. If runs stable increase the charge current ( max load limit 70% of Gneretaor nameplate rating in KVA) till you come across the hunting phenomena. Set value to next lowest and you are good to go.

4. If the above steps do not solve your issue then you have no option but to buy a low-frequency charger as used to charge traction batteries.

In any case, note that running PIP MS on small generators is not recommended a long-term solution as it overheats the rotor on long-term use and will cause the premature death of the generator you are better off bypassing the load directly to the generator ( BYPASS mode). This can be easily achieved using OSO ( Solar only) as charge setting.
Use generator only in emergencies when batteries are totally flat or a very cloudy day

p.s. By profession, I am a consulting electrical Engineer for High voltage electrical networks but I have self-installed a PIP- MS in 3 phase configuration in my home and have helped may friends install one for their houses over the years. For their price, there is nothing out there which can match the options of settings and convenience. Just keep one spare around or connect the bypass switch for emergencies :D
Attachments
waveform001.jpg
waveform001.jpg (77.27 KiB) Viewed 7499 times
waveform003.jpg
waveform003.jpg (99.18 KiB) Viewed 7499 times
waveform005.jpg
waveform005.jpg (87.05 KiB) Viewed 7499 times
andys
Groupie
Posts: 78
Joined: Wed, 20 Jul 2016, 19:44
Real Name: Andrew Snow
Location: Sydney

Re: PIP-4048MS inverter

Post by andys »

Thank you ghatikar, most interesting. This situation is quite frustrating as I'm not sure what generator to buy for my new system (4 x parallel PIP) because I don't know what it will be able to charge at.

Also I am not using generators with AVR, just the constant RPM direct AC type.

I have been looking into 48VDC generators instead - there's an Australian company selling them but they are quite expensive (see Eniquest Husky, its about $10k AUD). Promising work is being done by someone on whirlpool forums using an 8 phase 48V AC generator that has very low ripple for directly charging DC batteries with no voltage regulator (just DC rectification)
hennejg
Noobie
Posts: 5
Joined: Wed, 30 Aug 2017, 05:36
Real Name: Jörg Henne

Re: PIP-4048MS inverter

Post by hennejg »

As "promised" in my last post I switched to a 48V PIP and a LiFePO4 battery with an EMUS BS at the helm. The patched firmware runs fine - thanks again for the great work!
coulomb wrote: Mon, 04 Sep 2017, 13:14
hennejg wrote: Fri, 01 Sep 2017, 02:43 Now I can only hope that your patches still work with the recent models.
Yes, that's something we'd like to know as well. If your machine comes with firmware version 73.00, then you're definitely safe.
The device is from May 2017 but (still?) came with a 73.00. As I sad, the patch seems to run just fine.

On the ither hand, the BMS's behaviour towards the end of the charge cycle isn't optimal, unfortunately. Which is due to the two devices stepping on each other's toes.

If we were to ask the EMUS BMS, it would like to control the charger's current/voltage through CAN. In absence of CAN support it can control a charger by slowly toggling it on/off. Now, the only way to turn off the PIP's charger is by cutting the solar power supply (my installation doesn't use the grid input). Which, unfortunately happens at the end of the charge cycle, preventing "direct" use of solar power through the PIP's internal bus. A work-around is to set the BMS's end-of-charge cell voltage somewhat higher than the PIP's cut-off voltage. The BMS might still turn off the charger when it needs to balance the cells, but only for relatively short periods of time. A downside of this approach, though, is that the BMS's coulomb counter gets out of what due to the charge cycle never coming to an end.

How did you out there solve this problem (which may not be present with other BMSs)? In theory, I could build a CAN-to-serial translator that takes the BMS's input and translates it into configuration updates for the PIP. However, I suspect that this would quickly wear out the EEPROM cells holding those values.
weber wrote: Thu, 31 Aug 2017, 08:29
hennejg wrote: Wed, 30 Aug 2017, 06:15 Which raises another question: is there an elegant way of disabling a) the charger [...]
It depends what you consider elegant. But surely patching the code gives the most elegant solution.
In an ideal world, the PIP would have an extra command which takes U/I values for the charger which don't get stored in EEPROM. For robustness reasons the would remain in effect only for a few seconds unless refreshed continuously which would let the charger fall back to safe defaults of the charger/BMS communication was lost.
Did your hacking efforts ever add code to the PIP's firmware or did you "only" (yes, no small feat by itself) tweak certain parameters and constants?
andys
Groupie
Posts: 78
Joined: Wed, 20 Jul 2016, 19:44
Real Name: Andrew Snow
Location: Sydney

Re: PIP-4048MS inverter

Post by andys »

I am updating the PIP twice day over serial via PBFT command to begin and end rapid charging. Other than wearing out the EEPROM (not sure how many cycles it would support?) it is working well for me.
User avatar
weber
Site Admin
Posts: 2936
Joined: Fri, 23 Jan 2009, 17:27
Real Name: Dave Keenan
Location: Brisbane
Contact:

Re: PIP-4048MS inverter

Post by weber »

hennejg wrote: Wed, 04 Oct 2017, 00:31 If we were to ask the EMUS BMS, it would like to control the charger's current/voltage through CAN. In absence of CAN support it can control a charger by slowly toggling it on/off. Now, the only way to turn off the PIP's charger is by cutting the solar power supply (my installation doesn't use the grid input). Which, unfortunately happens at the end of the charge cycle, preventing "direct" use of solar power through the PIP's internal bus. A work-around is to set the BMS's end-of-charge cell voltage somewhat higher than the PIP's cut-off voltage. The BMS might still turn off the charger when it needs to balance the cells, but only for relatively short periods of time. A downside of this approach, though, is that the BMS's coulomb counter gets out of what due to the charge cycle never coming to an end.

How did you out there solve this problem (which may not be present with other BMSs)? In theory, I could build a CAN-to-serial translator that takes the BMS's input and translates it into configuration updates for the PIP. However, I suspect that this would quickly wear out the EEPROM cells holding those values.
Since we're using the LyteFyba BMS, which we designed, we arranged for each cell monitoring unit (CMU) to bypass, not at a specific cell voltage, but at a specific cell state of charge (call it 98%), based on its estimated open-circuit voltage. The BMS master unit (BMU) tells all the CMUs what the current is, and they estimate their internal resistance based on their temperature, and use that to compensate for the voltage rise or drop due to current.

For 16 LFP cells, we set the PIP's float voltage to 53.7 V (3.356 v per cell) which, if all the cells are balanced, puts them just below the voltage at which their CMUs will bypass (3.368 V), given near-zero amps. The bypass current of our CMUs is about 0.65 A.

We set the PIP's absorb voltage to whatever it takes to get the cells to that 98% SoC while putting in the maximum current. This is typically 55.2 V, but it depends on the relative sizes of the PV array and the battery. The BMU resets its coulomb counter to 100% when all the CMUs report an estimated open circuit voltage at least 40 mV higher than the bypass voltage (3.408 V).

As backup protection, if any CMU reports an estimated open circuit voltage of 3.608 V or more, the BMU turns off contactors to disconnect the PV array and the utility power (or generator) from the PIP. It doesn't turn them back on again until all CMUs have stopped bypassing (i.e. all cells have fallen below an estimated open circuit voltage of 3.368 V).
In an ideal world, the PIP would have an extra command which takes U/I values for the charger which don't get stored in EEPROM. For robustness reasons the would remain in effect only for a few seconds unless refreshed continuously which would let the charger fall back to safe defaults of the charger/BMS communication was lost.
Did your hacking efforts ever add code to the PIP's firmware or did you "only" (yes, no small feat by itself) tweak certain parameters and constants?
I was having thoughts in that direction too, and intended to raise the question with coulomb, but you beat me to it. :-) We have added small pieces of code.

One of the difficulties is that the solar charge controller (SCC) has its own processor, of a different type from the main board, and firmware which we have not analysed, and possibly its own EEPROM to store settings. The main board sends serial commands to the SCC. What we found in the past is that when we send a PBFT (set float voltage) or PCVV (set absorb voltage) command to the PIP, the SCC will stop charging and will begin a new maximum power point tracking (MPPT) sweep from scratch, taking maybe 30 seconds to get back to charging at the new voltage.

So in addition to avoiding EEPROM wear, we would want to stop that MPPT reset from happening. That may be relatively easy if it is directly commanded by the main board. But if the SCC just decides to do it, all by itself, whenever the main board tells it to change its settings, or if we have to worry about SCC EEPROM wear, then it will be very difficult. But it certainly deserves investigation.
One of the fathers of MeXy the electric MX-5, along with Coulomb and Newton (Jeff Owen).
andys
Groupie
Posts: 78
Joined: Wed, 20 Jul 2016, 19:44
Real Name: Andrew Snow
Location: Sydney

Re: PIP-4048MS inverter

Post by andys »

weber how do you measure open circuit voltage while the battery is still a member of a possibly charging or discharging pack?
User avatar
weber
Site Admin
Posts: 2936
Joined: Fri, 23 Jan 2009, 17:27
Real Name: Dave Keenan
Location: Brisbane
Contact:

Re: PIP-4048MS inverter

Post by weber »

andys wrote: Wed, 04 Oct 2017, 10:25 weber how do you measure open circuit voltage while the battery is still a member of a possibly charging or discharging pack?
Hi andys. I never claimed to measure it. You will see that I refer to it as "estimated" every time I mention it. In short, we take the measured cell voltage and we subtract the measured current times the estimated internal resistance, where charge current is positive and discharge current is negative.
Here's a longer explanation of the method, including how we estimate the internal resistance:
viewtopic.php?f=31&t=4332&p=59994#p59994
One of the fathers of MeXy the electric MX-5, along with Coulomb and Newton (Jeff Owen).
dakoal
Noobie
Posts: 5
Joined: Sat, 08 Jul 2017, 21:56
Real Name: Bernhard STrunz
Location: Europe/Austria

Re: PIP-4048MS inverter

Post by dakoal »

Hello.

I am one of the DIY Powerwall guys.
There is one question coming here less or more often:
How to control the AMP's the PIP is charging to the batteries from grid (utility charge) without changing the value on the panel or with serial port?
Serial Port has 2 downsides:
1. EPROM wearout on heavily changing values all the day
2. only 10Amp steps (aprox. 500Watts with 50Volt Battery)

That function can be useful on grid tied PV installations that have other Inverters connected beside with the PIP as utility charger for minimizing grid meter movements.

I thought on some pot/resistor in the PIP that can be modified to change the actual charge AMPs the PIP does to the batteries.
Or to "fake" the actual AMP's measured to let the Logic do the work.

Or how does the controller "tell" the power part of the PIP the AMP's to be charged? Maybe there is a way of injecting some logic.
andys
Groupie
Posts: 78
Joined: Wed, 20 Jul 2016, 19:44
Real Name: Andrew Snow
Location: Sydney

Re: PIP-4048MS inverter

Post by andys »

weber wrote: Wed, 04 Oct 2017, 10:46 Hi andys. I never claimed to measure it. You will see that I refer to it as "estimated" every time I mention it.
Ah yes! Very clever indeed, especially the way you determine charge completion.

(Although in practice I've noticed the voltage can be "sticky" during periods of transition from charge to discharge and back, but I guess that doesn't matter much for these purposes)
User avatar
coulomb
Site Admin
Posts: 6357
Joined: Thu, 22 Jan 2009, 20:32
Real Name: Mike Van Emmerik
Location: Brisbane
Contact:

Re: PIP-4048MS inverter

Post by coulomb »

dakoal wrote: Wed, 04 Oct 2017, 11:05 Serial Port has 2 downsides:
1. EPROM wearout on heavily changing values all the day
I'm not convinced that EEPROM wear is a problem; see my post on this a year ago. (I've just added it to the index.)
2. only 10Amp steps (aprox. 500Watts with 50Volt Battery)
That function can be useful on grid tied PV installations that have other Inverters connected beside with the PIP as utility charger for minimizing grid meter movements.
Yes, it seems strange to enforce the 10 A steps with the MUCHGC command. It makes sense on the LCD, but why reject commands that ask for say 37 A? It's less code to just check the minimum and maximum currents. We could add it to the existing patch, but I'm hoping the patch eventually won't be needed. Plus it's moderate work patching every new firmware that comes out, and the official update files are getting very hard to find.
I thought on some pot/resistor in the PIP that can be modified to change the actual charge AMPs the PIP does to the batteries.
I'm not aware of any resistor that could be adjusted to modify the charge current.
Or to "fake" the actual AMP's measured to let the Logic do the work.
That might be possible, but there is always the risk of modifying other behaviour, for example preventing inverter overload.
Or how does the controller "tell" the power part of the PIP the AMP's to be charged?
I believe that the current set by a PWM value in a register (for mains charging). For solar charging, there are commands sent to the SCC over a dedicated serial port between the TMS320 DSP processor and the SCC's processor.
Maybe there is a way of injecting some logic.
That's possible for solar charging, but I really think it's better to let the PIP take control of charging current. Have the BMS intervene only in the case of cell stress. Yes, it's drastic dropping contactors, but if the system is set up properly, that should be a rare event.

[ Edit: "last December" -> "a year ago". ]
[ Edit: -> "that could do that" -> "could be adjusted to modify the charge current". ]
MG ZS EV 2021 April 2021. Nissan Leaf 2012 with new battery May 2019.
5650 W solar, 2xPIP-4048MS inverters, 16 kWh battery.
Patching PIP-4048/5048 inverter-chargers.
If you appreciate my work, you can buy me a coffee.
hennejg
Noobie
Posts: 5
Joined: Wed, 30 Aug 2017, 05:36
Real Name: Jörg Henne

Re: PIP-4048MS inverter

Post by hennejg »

coulomb wrote: Wed, 04 Oct 2017, 17:34 I'm not convinced that EEPROM wear is a problem; see my post on this a year ago. (I've just added it to the index.)
It would indeed be interesting to check how many updates the BMS issues during a typical charge cycle and then how long the 1M cycles would last.
I believe that the current set by a PWM value in a register (for mains charging). For solar charging, there are commands sent to the SCC over a dedicated serial port between the TMS320 DSP processor and the SCC's processor.
The way I understand the power/charge management is supposed to work is that, at least once the battery is in float mode, and there is enough solar power, the solar power is "routed" directly via the DC bus from the SCC to the inverter. Maybe I got this completely wrong, but if it were true, the goal of the power management would be to relieve the battery from charge/discharge current, except for short bursts due to the power management having to catch up with a changing load situation. So I would expect a relatively high rate of requests from the main controller to the SCC.

Somewhere you mentioned that you sniffed the traffic between SCC and main controller. Is this data available anywhere?
dakoal
Noobie
Posts: 5
Joined: Sat, 08 Jul 2017, 21:56
Real Name: Bernhard STrunz
Location: Europe/Austria

Re: PIP-4048MS inverter

Post by dakoal »

Thanks for the thoughts.
Yes, it seems strange to enforce the 10 A steps with the MUCHGC command. It makes sense on the LCD, but why reject commands that ask for say 37 A? It's less code to just check the minimum and maximum currents. We could add it to the existing patch, but I'm hoping the patch eventually won't be needed. Plus it's moderate work patching every new firmware that comes out, and the official update files are getting very hard to find.
That would be the best way. I will try single Amp steps in the evening on the current firmware I have on the PIP.

And for the Flash "Wearout" thing: If I retest current every minute, but let's say, flow is constant so I only have to do an change every 5 minutes I come to 12 every hour and 120 in 10 hours possible Sunlight value changes per day means (minus rainy days) 36.000 flash changes per year.
User avatar
weber
Site Admin
Posts: 2936
Joined: Fri, 23 Jan 2009, 17:27
Real Name: Dave Keenan
Location: Brisbane
Contact:

Re: PIP-4048MS inverter

Post by weber »

hennejg wrote: Thu, 05 Oct 2017, 04:03 The way I understand the power/charge management is supposed to work is that, at least once the battery is in float mode, and there is enough solar power, the solar power is "routed" directly via the DC bus from the SCC to the inverter. Maybe I got this completely wrong,
If by "DC bus" you mean the nominally 48 V DC bus to which the battery is directly connected (apart from safety devices), then you have it right. However I note that unqualified references to "the DC bus" usually refer to the nominally 400 V DC bus on the other side of the PIP's bidirectional DC-DC converter. The SCC does not connect to that. But your following point doesn't depend on where the SCC connects.
but if it were true, the goal of the power management would be to relieve the battery from charge/discharge current, except for short bursts due to the power management having to catch up with a changing load situation. So I would expect a relatively high rate of requests from the main controller to the SCC.
Yes. Although I wouldn't call them "requests", the main controller regularly tells the SCC how much current the main board is putting into or out of the battery. And occasionally tells it what the maximum current should be, and what charge mode it should be in (bulk/absorb, float or off).
Somewhere you mentioned that you sniffed the traffic between SCC and main controller. Is this data available anywhere?
No. That method was long ago superseded by coulomb reading the disassembled code, for both the main controller and the SCC, to find out what they send to each other. I'm answering these questions on his behalf because he has been busy doing that, to try to determine the feasibility of implementing a dynamic variant of the MNCHGC command that would be non-MPPT-resetting and non-EEPROM-wearing. If we attempt it, it will be a lot of work. In particular a lot of testing in various modes and under various conditions, including with multiple PIPs operating in parallel. And we don't have a file containing the 73.00 firmware, so it would still require users to downgrade to a patched 72.70. [Edit: We no longer recommend putting version 72.xx (not even our patched versions) into inverters that are supplied with 73.00. We recently discovered that this causes a problem with a relay that has been added inside these new machines, that can connect Neutral to Earth on the AC output. We now have a 73.00 firmware file.]

Clearly we're not getting paid for this. We can only justify spending the time on it if we believe it will help to accelerate the uptake of renewable energy -- to sooner win the war that the fossil fuel corporations have declared against life on this planet. So whether we feel the effort is justified depends very much on how many people would actually use such a feature. How important is it really?
One of the fathers of MeXy the electric MX-5, along with Coulomb and Newton (Jeff Owen).
andys
Groupie
Posts: 78
Joined: Wed, 20 Jul 2016, 19:44
Real Name: Andrew Snow
Location: Sydney

Re: PIP-4048MS inverter

Post by andys »

hennejg if your solar input and battery are sized optimally, I think you don't really need to worry that much about charge rate - you can set things up so that the PIP will be sending its maximum charge to the battery for say, 96% of the battery capacity. You can then switch to a lower charge rate (eg. 53.7V) for the remaining few percent.

Thats how I do it, and it is supposedly better for the battery chemistry too to spend less time at high charge voltage.

edit: In my case, the numbers are: 160Ah lifepo4 battery, Charge at 56.1V. with the usual 2 to 3 kW PV input, It doesn't actually reach 56.1V until very late in the charge, around the 95% mark.
User avatar
weber
Site Admin
Posts: 2936
Joined: Fri, 23 Jan 2009, 17:27
Real Name: Dave Keenan
Location: Brisbane
Contact:

Re: PIP-4048MS inverter

Post by weber »

Coulomb and I spent the day investigating the feasibility of a dynamic maximum charge current command. We believe we can implement it by adding new behaviour to the MNCHGC command, by using otherwise-unused values of its current argument (3 digits of amps), namely values of 500 or more. Such values will have 500 subtracted from them and will be treated as a dynamic maximum charge current, hopefully for both AC and solar charging. Values less than 500 will be treated as before.

By "dynamic" we mean:
(a) it can be sent as often as desired without wearing the EEPROM,
(b) it won't cause the SCC to restart its MPPT sweep and
(c) it will stay in force for a maximum of one minute before reverting to the value stored in EEPROM.

To do this, we have written 4 routines that we will need to call from 11 locations. Some of the complication is due to the fact that, in parallel or 3-phase configurations, the MNCHGC command lets you change this setting on machines other than the one you are talking to directly. You have to supply a machine number argument as well as the current. See the bottom of page 18 here:
uploads/293/HS_MS_MSX_RS232_Protocol_20 ... pgrade.pdf

There's still a lot of work to do to manually insert the patches into the .hex file. If we manage to get it working with a single PIP, we'd appreciate volunteers to beta-test it, particularly those with parallel setups. It will initially only be included with the LFP (lithium ferrous phosphate) version of the patched firmware.

Edited by Coulomb: added photo.
Attachments
Preparing dynamic current patch_sm.jpg
Preparing dynamic current patch_sm.jpg (253.9 KiB) Viewed 5547 times
One of the fathers of MeXy the electric MX-5, along with Coulomb and Newton (Jeff Owen).
andys
Groupie
Posts: 78
Joined: Wed, 20 Jul 2016, 19:44
Real Name: Andrew Snow
Location: Sydney

Re: PIP-4048MS inverter

Post by andys »

weber wrote: Fri, 06 Oct 2017, 19:50 Coulomb and I spent the day investigating the feasibility of a dynamic maximum charge current command.
This sounds good! I have been reluctant to flash mine thus far, since I don't ever let the PIP reach bulk charge termination. I've never tried the MNCHGC command.

I am setting up a new (parallel) system at the moment, which is not online yet, so I'd be very interested in trying out such a feature.
Post Reply