Notice. New forum software under development. It's going to miss a few functions and look a bit ugly for a while, but I'm working on it full time now as the old forum was too unstable. Couple days, all good. If you notice any issues, please contact me.
InPhase Senior Member Joined: 15/12/2020 Location: United StatesPosts: 178
Posted: 05:32pm 23 Mar 2023
Copy link to clipboard
Print this post
I've got a big-ass forklift battery that likes almost 200 amps of charge current. The solar panels feeding it are 14x 435 watt, 76 Voc. The problem I have is that I would like to put some in series to increase the voltage, but the most I can do is 2s due to the Make Sky China controllers topping out at 160 volts.
So I've been pondering a diy high voltage controller. Maybe 250 volts. I've been reading Poida's 150 volt controller thread. What are the practical considerations ? Obviously, the switching elements need to be rated for the voltage. I have some 600 volt IGBTs on hand. I would also like to get as high of an output current as possible. So would something along the lines of a push-pull converter be better than the synchronous buck style?
Godoh Guru Joined: 26/09/2020 Location: AustraliaPosts: 458
Posted: 08:57pm 23 Mar 2023
Copy link to clipboard
Print this post
High voltage DC makes me shudder. As a retired electrician I find it scary what it can do. I have seen solar panels that have shorted across tracks and burnt. I have seen serious arcing on high voltage DC. And have blown up a couple of multimeters while repairing valve amplifiers. The stuff jumps well. I don't know about the US but in Oz working on DC above 120 volts must be done by a licenced electrician. I have my panels connected to supply 70 to 90 volts. I have 4 x 50 amp regulators in parallel. 4 different solar arrays that each have their own regulators and the outputs are paralleled to the batteries. So my system can put out 200 amps if needed. So far it is not needed as the sun comes up, panels start charging, battery voltage rises, current drops off. Build the regulator if you dare but in the words of Elma Fudd. "be wery wery careful" Pete
CaptainBoing Guru Joined: 07/09/2016 Location: United KingdomPosts: 2075
Posted: 09:21pm 23 Mar 2023
Copy link to clipboard
Print this post
Solar Mike Guru Joined: 08/02/2015 Location: New ZealandPosts: 1138
Posted: 09:42pm 23 Mar 2023
Copy link to clipboard
Print this post
A fully synchronous buck inverter is better suited to higher currents (Poida's design is asynchronous), however once output current gets over 60 amps the buck inductor starts getting quite large due to the very thick wire needed. Going to a higher voltage also requires a larger inductance, thus the inductor being the main critical limiting factor for DIY. PCB tracks will have to be 1.6-1.8mm apart to prevent arcing.
If your PV voltages are relatively fixed and the battery is constant, then there is a case to make a high frequency push-pull step down inverter around 100Khz using a large ferrite EI core. Variation of the pulse width will effectively alter the output voltage, boost - float.
Depending on whether redundancy is required, it is sometimes better to have multiple lower power charge controllers connected to groups of panels; then its possible to orientate them in different directions to get a more even solar input across the whole day.
Cheers Mike
KeepIS Guru Joined: 13/10/2014 Location: AustraliaPosts: 1679
Posted: 10:46pm 23 Mar 2023
Copy link to clipboard
Print this post
I have the a similar setup to Godoh.
4 x 60A MakeSB regulators and 400 Watt panels, only 2 panels in series in the parallel sets in the array for my 56v system.
I feel there is NO second chance with high voltage DC.
The battery bank can take 300A @ 56v, but like Godoh, they charge as the Sun comes up. The highest output from the combined regulators that I've seen is 160A, that happens when a very high load is running (5kW to 7kW) and the Solar energy is being modulated by huge fast moving thick clouds. That current is way higher than my panels should deliver.
It appears to be a combination of High Load current that also pulls a lot from the batteries during long spells of thick cloud cover, then a sudden burst of full sun as the fast moving cloud has a break causing the High load current added to the surge in Battery Charge current [LiFePO4] that's causing a problem for the Charge controllers.
I don't know if it's the array delivering peak currents higher than the Solar array rated output, or the Solar regulators being slow to Down regulate the sudden burst in Solar output, or both?
In systems running Solar input close to the regulator ratings, it can cause narrow instantaneous peak currents way over the rated Current output of the Solar regulators. I've seen that reported for all kinds of controllers in similar cloud conditions and it's possibly one of the reasons a Solar regulator blows smoke when everything appears to be within tolerance.
Once I saw those spikes on my Graph SW I was glad that I had plenty of Solar input Headroom to the regulators from splitting arrays across multiple regulators, although that was originally done for redundancy.It's all too hard. Mike.
InPhase Senior Member Joined: 15/12/2020 Location: United StatesPosts: 178
Posted: 03:19am 24 Mar 2023
Copy link to clipboard
Print this post
Just so happens, I am a licensed electrician I also have a fair amount of experience with HVDC, having installed it in telecom and industrial processes, as well as a recent 450 V solar system. The distribution and safety of the install is well understood. I do appreciate the concern.
For the most part, this is an academic exercise. I'm thinking through what I'm going to do when I build the array further away. Voltage drop is going to be an issue and the larger wire is going to be pricey. I'd rather spend the money on DIY electronics (and face failure, but learn a lot) than large wire.
Solar Mike, do you think multiple synchronous bucks at 50 amps or less would be better than a push-pull or forward converter?
Solar Mike Guru Joined: 08/02/2015 Location: New ZealandPosts: 1138
Posted: 07:45am 25 Mar 2023
Copy link to clipboard
Print this post
A high power push pull inverter is going to be more efficient than the equivalent single buck converter, however it may be easier to DIY several lower powered buck synchronous converters perhaps in a multi-phase arrangement, driven from a single control system, or slaved to a master controller for end point Bulk and float transition points.
Cheers Mike
noneyabussiness Guru Joined: 31/07/2017 Location: AustraliaPosts: 512
Posted: 11:44pm 25 Mar 2023
Copy link to clipboard
Print this post
has anyone thought of a sepic converter?? isolation, without isolation... wouldn't be difficult retro-fit to current boards..I think it works !!
poida Guru Joined: 02/02/2017 Location: AustraliaPosts: 1418
Posted: 09:21am 26 Mar 2023
Copy link to clipboard
Print this post
The mppt converter will be more efficient (less heat) the lower the conversion ratio. So this is something that you might be able to use.
What is the forklift battery voltage? Is it 150V or 200V? Most that I know of are nominal 48V lead acid types.
If each panel of yours is 76V OC then max power voltage will be good bit less, maybe 60V. Can you post the spec sheet of the panels here?
It might be that to get max power voltage from your panels you will need to put 2 panels in series since one panel might have it's max power voltage below the battery charge voltage.
IF the battery is 48V, then charge voltage is about 54 - 55 V or something. The mppt conversion will be very efficient taking the 60V down to 55V. I would make a few mppt and put them in series (with big fuses on their outputs)
Nick has driven these at well over 45 Amps, more like 75 Amps and no problems. I tested both Nick's board and Wiseguy's boards at 3,000W at 55V and all that happens is some PCB traces get warm.
14 x 435 W = 6090 W at 55V this is about 110 Amps 2 or 3 mppt will do this all day.wronger than a phone book full of wrong phone numbers
InPhase Senior Member Joined: 15/12/2020 Location: United StatesPosts: 178
Posted: 02:55pm 26 Mar 2023
Copy link to clipboard
Print this post
I got a look at the panels and I was way off. The Voc is 85.6 and the Vmpp is 72.9. This explains why I my temp setup is what it is. I currently only have 2 panels connected, leaned against a tree, while the land is cleared and the racking gets built. It's been months now and I forgot my reasoning. I originally connected two panels in series to each controller, gambling on the idea that the connected voltage wouldn't rise to Voc. It worked, but I chickened out. Here's the specs:
Vmpp: 72.9 V Impp: 5.97 A Voc: 85.6 V Isc: 6.43 A
Batt Volts: 48 V
The battery has been floating, but now it's time to get a move on. The reason I contemplate such a high voltage is that the panels will end up on a rack over 100 meters (metres) away from the battery. It would be nice and cheap to send a few amps over small wire than a lot of amps over a big expensive wire.
Warpspeed and I have been talking a bit about a low frequency (500 Hz?) forward converter. But forward converters are generally used up to a few hundred watts. I'm also pondering the idea of just going full bridge through a transformer. I've got plenty of quarter bridge boards left from my modular inverter design. Basically, just turn the PV voltage into AC and put it through a transformer to be rectified at battery voltage. I'm not sure about efficiency, but the trade off is probably worth it compared to the cost of wire and multiple SCCs or a couple of expensive Victron or Morning Star high voltage SCCs. Plus I can service a DIY rig myself.
Godoh Guru Joined: 26/09/2020 Location: AustraliaPosts: 458
Posted: 09:02pm 26 Mar 2023
Copy link to clipboard
Print this post
There is another way to do what you want. Why not just build a small shed or plant a half shipping container next to the panels. Put the batteries, controllers and inverter there at the panels. Run AC from there to the house. Pete
phil99 Guru Joined: 11/02/2018 Location: AustraliaPosts: 2135
Posted: 10:02pm 26 Mar 2023
Copy link to clipboard
Print this post
" a half shipping container next to the panels. Put the batteries, controllers and inverter there at the panels."
Commercial remote area systems are often like that. Presumably it is the most cost effective.
InPhase Senior Member Joined: 15/12/2020 Location: United StatesPosts: 178
Posted: 11:17pm 26 Mar 2023
Copy link to clipboard
Print this post
It's geographically untenable, but you're right, it would be ideal. This controller I am thinking of may never work and I might blow up a bunch of IGBTs. I may be forced to dynamite a mountain side in order to put commercial controllers next to the panels. But much like the inverters I have built (piles of dead MOSFETs as evidence), I learned so, so much. And I thank all of you that came before me to murder MOSFETs and record their findings here. When I first came across the OzInverter video on YouTube, I remember thinking that there is no way I'll ever be able to make my own inverter. In some way, I still haven't, because it is just some boards screwed to plywood, but it works a treat!
This might not work, but I'll document my failures for you all to see.
poida Guru Joined: 02/02/2017 Location: AustraliaPosts: 1418
Posted: 12:50am 28 Mar 2023
Copy link to clipboard
Print this post
InPhase, I would not worry too much about efficiency, that is copper losses over 100 Meters.
The most frequent situation we see here is the battery is partially charged and for only a few hours does it consume high charge current due to it getting to absorb voltage and then current is reduced by the mppt controller
I find in my case with a 400 Ah 48 Volt battery, I run it down about 50% of it's capacity each day and I replace that the next day. That is, a 10 kW.hr discharge/recharge cycle and that is a heavy use day. Mostly it's more like 6kW.hr
I have 6kW peak panels and 1/2 does a little work early in the morning and the other 1/2 does the main part during the day. The ratio is about 1/4 for the East facing array, 3/4 for the North array.
Of course things will be different for you.
I particularly want to point out that the charge current in my setup rarely gets above 45 Amps in total. 40 Amps at 50V = 2kW and all I need is 2 hours of that plus the rest of the day at absorb voltage to put the 6 or 10kW.hr back in.
here is a typical sort of sunny day. cloudy in the morning. 8.1kW.hr was used by the inverter for that 24 hours (Brown line) 7.4kW.hr was got from the panels and mppt controller (Dark Blue line)
We can see the battery spent from 12 until 6pm in absorb mode, where there was more than enough solar power to keep it there.
There is no need for more power (or higher current capability) It's built big enough for my battery and use case.
I doubt you will be able to use 200 Amps charge current for long and not often either. So consider designing for 100 Amps charge current. This will make it possible to run long lengths of maybe 25mm2 cable the 100 Meters needed. We do not need to follow design rules when choosing large current cables. Some heat generation is OK, depending on the circumstances, that is how long high currents are being transmitted, how high the over current situation persists...all that.
Where I live we have a lot of cloudy days. That means peak panel power will never occur.
here is a cloudy day:
there was not enough sun to even get the battery to absorb voltage. The inverter used 9.3 kW.hr only 5.8kW.hr could be got from the sun
No point in having 200Amp capable wiring and controllers when the power is not there.
But, it's very good to have as many panels you can get helping with the charge since they are all running at a small fraction of rated peak power.
Maybe you could make good use of 200 Amp charge current if 1 - it's a huge battery, 800 Amp.hr at 48V or bigger 2 - the battery is discharged well below 50% remaining capacity 3 - the sun is out long enough to get to absorb voltage and stays out for that time again 4 - you expect dismal weather in the next few days.
otherwise, just cycle the battery 20-30% down and back up again and 100 Amp charge system will do fine.
This is my thinking. no need for any complex new designs.
put the panels facing South (for you in the Northern hemisphere) wire them in parallel this means 6 Amps x 14 panels or 84 Amps to convey the 100 meters. find used or scrap 25mm2 or bigger cable. one for + one for - run that 100 meters to the charge controllers, both set to 50 Amps max. output put the battery near the controllers and the inverter.
The cable will carry 84 amps for a hour or two but it won't need to do it for longer. You could run them separated by a foot or so, just put in the soil with no air so the soil can take some of the heat away.
But I am a hacker, not an engineer.
if you are interested, you can browse this data at http://1eq2.cc/mppt/nudatah.php Edited 2023-03-28 15:02 by poidawronger than a phone book full of wrong phone numbers
poida Guru Joined: 02/02/2017 Location: AustraliaPosts: 1418
Posted: 05:08am 28 Mar 2023
Copy link to clipboard
Print this post
Also, at max power, those 14 panels will do 84 Amps at 73V With a conversion efficiency of about 94% (maybe more like 92%) you will get 0.94 x 84 x 73 Watts or 5760 Watts This is the most those panels will be able to feed into the battery.
If the battery is 48V, then the mppts output will be 84 x (73/48) or 128 Amps This is not 200 Ampswronger than a phone book full of wrong phone numbers
nickskethisniks Guru Joined: 17/10/2017 Location: BelgiumPosts: 458
Posted: 11:13am 28 Mar 2023
Copy link to clipboard
Print this post
Would be interesting to build a HV unit no?
We build our own 48V--> 240VAC inverters so why not try a HV MPPT charger? Yes HV DC is dangerous, but so is AC and low voltage DC.
If you choose for high quality components and a metal box to contain everything ( including the smoke and sparks ) This will be as safe as a grid tie inverter. But those HV cables need to be well protected, I think this is the weak and most dangerous part of the system.
BUT, I would use/suggest you to use a grid tie inverter for safety. Only need some circuitry/software to prevent the battery from over charging.
But I have the feeling you want to build something, no?
You can split the panels in 3 strings: 1x4 series panels and 2x5 series panels, the 2 strings of 5 series panels can be paralleled. So you need 2 HV controllers. I would not start with 2 strings of 7 series panels ( higher voltage igbt's, paralleled capacitors, etc..).
I would prefer using a full H-bridge and a hf switching transformer, a push pull can give troubles in the control loop, because of the leakage inductance in the transformer. You probably need to use parallel transformers above 2kW to keep using standard cores and coilformers. I think Peters software can be implemented without changes. Edited 2023-03-28 21:14 by nickskethisniks
InPhase Senior Member Joined: 15/12/2020 Location: United StatesPosts: 178
Posted: 01:46pm 28 Mar 2023
Copy link to clipboard
Print this post
Yes, I can't give the battery 200 amps with what I have now. But the plan is always more panels when I can get them. The battery will live with less, but some day I'd like to stir it up with something closer to C/10.
And yes, Nick, I'm just kinda itching to build stuff! I have plenty of unfinished projects, so this is just a preliminary feel-it-out stage. For the immediate future, I will just have to tough it out at a lower voltage and deal.
Warpspeed has been using an analog set power point controller with great success. I've studied the circuit and it makes good sense. It is immediately adaptable to a full bridge transformer drive. He uses the TL494 PWM chip that I already had in mind to use (as I have a few already). A pot sets the rated Vmpp and then the TL494 adjusts the PWM to control loading on the circuit. No need for software and very fast. Perhaps not quite as efficient in the long run, but very simple. So I have the IGBTs, I have the PWM controller, I just need to work out the transformer, something I've never really studied much of in HF power supplies. Edited 2023-03-28 23:47 by InPhase
Notice. New forum software under development. It's going to miss a few functions and look a bit ugly for a while, but I'm working on it full time now as the old forum was too unstable. Couple days, all good. If you notice any issues, please contact me.
Print this post
) This will be as safe as a grid tie inverter. But those HV cables need to be well protected, I think this is the weak and most dangerous part of the system.