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.
Standard alternator needs to turn about 2,000 rpm before it can even reach 12v and the alternator light goes out. May have to go to 3,000 alternator rpm for full rated current.
So alternator may need to turn at 12,000 rpm to charge a 48 volt battery. Max rated speed of alternator usually 20,000 rpm.
Its not going to work in a vehicle because even if you could get it to turn that fast, it would only work over a very limited engine speed range, that would make the vehicle undrivable on the road.
A normal alternator typically runs about three times engine speed. Normal car engine speed might be 800 rpm idling, 6,000 rpm flat out. Alternator x3 that, or 2,400 rpm to 18,000 rpm. works pretty well like that.Cheers, Tony.
Jacob89 Newbie
Joined: 10/09/2017 Location: AustraliaPosts: 39
Posted: 04:09am 04 Aug 2020
Copy link to clipboard
Print this post
Success! Sort of. I set the lower value at 100k as you said Tony, and it actually started switching!
All the problems I've experienced seem to stem from these cheap and nasty aliexpress transient voltage suppressors. I think they are polarised despite not having the ring marking them. I can orient them by which way up the writing on the side is, and I found by putting them the opposite way to what I have been previously they let go immediately.
So with one soldered in the "right" way, and the low voltage trip resistor at 100k it finally started working.
A few things:
The circuit does not work without the TVS in place! I cannot explain it but without it, all kinds of weird stuff happens. It some really weird really high frequency switching. 400khz - 1mhz depending on which meter I use. The mosfets get hot and I was getting hf burns from the heatsink. The input(array voltage) stays high, and the element sees much lower voltage, which varies according to the polarity I connect the multimeter in? I connected an incandescent bulb to get some visual representation of whats going on. It actually glows, but doesn't seem generate any appreciable heat and nor does the element.
When the circuit is working properly, it cycles quite slowly. I can see the bulb flashing at around 2-3 times a second. Using the 100k lower and 220k upper means it doesn't turn off until the voltage drops to around 100v and turns on again around 220v. Today has been intermittently cloudy which is perfect for testing this. When the sun comes out the array is big enough that the voltage doesn't drop low enough and it won't switch off until another big cloud comes. In full sun it will stay over 200v but with partial cloud it will dip right down to 110v and still not turn off.
So I added a 47k resistor in series with the 100k one. This worked briefly but with the much narrower on/off window, the switching frequency was much higher and the TVS doesn't seem to be able to cope with the faster switching and let the smoke out again.
I think this explains why the first two units kept blowing the voltage suppressors, with the higher and closer trip points the switching would have been much faster and they couldn't take it.
Nonetheless I'm pretty pleased to see it run, and I'm going to order some better transient voltage suppressors and hopefully I can better optimize the switching points.
Jacob89 Newbie
Joined: 10/09/2017 Location: AustraliaPosts: 39
Posted: 04:24am 04 Aug 2020
Copy link to clipboard
Print this post
They will output much higher voltage though if you feed higher voltage to the field. I was able to get over 60v out of a 12v alternator by bypassing the regulator and feeding 12v straight to the field. Spinning about 5000rpm.
Granted, that was open circuit no load voltage though. I don't know what you could expect with a load on it. And I don't know if the field is saturated at 12 volts, or if you could put more in it, or if its already way past saturation at that point.
I seem to remember an old thread by Oztules where he'd made an external regulator for an old 12v alternator so it could be used to charge a 24v bank.
A 24v alternator might be a better candidate for modifying to charge a 48v bank though.
Whether or not you could actually charge a 48v bank from a modified 12v alternator, I'm not sure, but I am sure it would be quite inefficient, and noisy, however you're going to get it spinning at 6000+rpm.
Warpspeed Guru
Joined: 09/08/2007 Location: AustraliaPosts: 4406
Posted: 04:27am 04 Aug 2020
Copy link to clipboard
Print this post
Great stuff there Jacob
The transient suppressor does exactly that, reduces any very high voltage spikes which may be getting back and confusing the control system.
Definitely sounds like you have the dc type polarised suppressors, if they burn up when turned around.
With possibly fake Chinese parts you can never be quite sure what you are actually getting. The part number on it is not always a guarantee of whats inside.
If the suppressor runs hot, which it may well do if the switching frequency is very high, you could parallel up some suppressors, or add more electrolytics to slow the whole beast down. Those suppressors are very effective, but can only handle a watt or two. As you may be switching kilowatts, it does not take much of a voltage spike to heat them up.
If you spread the trip points further apart, that will slow it down as well. Its really a case of experimenting. Anyhow you are making real progress and that's the main thing.
As you have found, plenty of sun, the load will stay on continuously, and that is a good thing. If you add on extra load it will start to switch again.
Once the two voltage trip points are set appropriately, it should self adjust and "do its own thing" under all solar conditions.Cheers, Tony.
Davo99 Guru
Joined: 03/06/2019 Location: AustraliaPosts: 1578
Posted: 06:34am 05 Aug 2020
Copy link to clipboard
Print this post
I'll have another play with mine and pay particular attention to the resistors and set up a bank of panels to match as close as I can.
I asked the local Auto elecs about using a 12 V alternator for higher voltages a while back. They do rewinds and both the father and son in the business are older than god so should know something.
I suggested spinning the alts a lot faster and they said the input voltage and Current would need to be a LOT higher as well. They said this would cause overheating. They said you could lower the amps to keep the heat down however time you got it to a workable level the amps would be very low.
I have seen a white paper where an external Reg was used to get a decent 24V output from a 12V leece Nevile alternator. The external regs are used a lot in marine applications and are stepped charge I think the term is, which is a lot more efficient than the built in regs. A big bonus with the external regs is getting something like 80% output at engine idle speeds. There is one controller I saw which has a temp probe and will do 100% output for a time at least as the alts tend to heat up due to lack of airflow so the controller backs off the output when the alt gets too hot. Still be a big jump on charging a depleted battery ASAP at low engine speeds.
I also read a number of times car alts themselves are in fact very efficient as they are of the same Claw pole type construction type as power station alts. The thing with car alts is the internal regulation controlling them is basically crap. One of the reasons given is the extreme conditions a car alt works in with under bonnet temps and widely varying RPM. A stationary application is almost always a lot easier on the alt because they work in free air not air that can easily be 100C before it is asked to cool the alt. With an external Reg a lot more control can be applied and the efficiency can take a huge leap as well as the variance in application.
I have played a LOT with the Mitsubishi Enclosed ( twin) fan alts and they are virtually unbreakable. I have run them for hours upon hours on my lister engine at absolute FULL power and had them on a battery as well as Direct connected to a load with no batt which is supposed to be a Big No No. Hasn't seemed to worry them a Bit. Besides Running them dead shorted, I have put them through about every torture test I could think of to try and blow the things but haven't succeeded.
The Twin enclosed fan alts I think are a Huge step up from the external fan type. On the twins, one fan cools the windings and one fan cools the electronics. You can feel the Temp coming out of them like a fan heater but they seem to run hot but stable even flat out which the ones I have tested all output slightly more , 5-7A, on an 80 A alt all day.
I have only run them stock with the internal regs and they are pretty inefficient with the mechanical input power. I'd guess they are pulling about 3 Kw ea for an 80A or 1000W load. I have pulled them apart and it seems tricky to isolate the regulator and the Diodes with the way they are built but no doubt a smarter person could manage it OK.
They do seem to be tough as nails though and I would think that if one were going to over clock one, The twin fan ones would definitely be the first choice as the Cooling seems very good indeed.
nambuat Newbie
Joined: 19/03/2022 Location: AustraliaPosts: 5
Posted: 07:14am 20 Mar 2022
Copy link to clipboard
Print this post
Hi Tony. Would this work for a Hardy 250 Litre HWS? I have 9 panels available to use as a stand-alone set-up to run the HWS. The panels are rated : VOCV-47.8, ISCA-5.57, VMP-38.3, IMP-5.23 Rated @ 200w. Cheers,Don
nambuat Newbie
Joined: 19/03/2022 Location: AustraliaPosts: 5
Posted: 10:18pm 20 Mar 2022
Copy link to clipboard
Print this post
Hi Tony. Would this work for a Hardy 250 Litre HWS? I have 9 panels available to use as a stand-alone set-up to run the HWS. The panels are rated : VOCV-47.8, ISCA-5.57, VMP-38.3, IMP-5.23 Rated @ 200w. Cheers,Don
Revlac
Guru
Joined: 31/12/2016 Location: AustraliaPosts: 1017
Posted: 12:09am 21 Mar 2022
Copy link to clipboard
Print this post
Hi Don, welcome to the forum, Tony is not here at the moment. It should work for you. What is The wattage of the heater element? I have an 1800W heater element, have run it with lower voltage direct from the panels, just to see how much power I could get away with, at 130vdc it takes most of the day but plenty hot enough. I have the pcb's made but haven't put it all together yet to see how it works.Cheers Aaron Off The Grid
nambuat Newbie
Joined: 19/03/2022 Location: AustraliaPosts: 5
Posted: 04:55am 21 Mar 2022
Copy link to clipboard
Print this post
Thanks Aaron. It's a Hardi-Dux 250 Litre so I imagine it's 3.6Kw. Any chance you might sell me some pcb's so I can try it out and report the results please? Cheers Don Edited 2022-03-21 15:19 by nambuat
Revlac
Guru
Joined: 31/12/2016 Location: AustraliaPosts: 1017
Posted: 08:07am 21 Mar 2022
Copy link to clipboard
Print this post
I have 2 spare, will send you a PM.Cheers Aaron Off The Grid
Davo99 Guru
Joined: 03/06/2019 Location: AustraliaPosts: 1578
Posted: 10:53am 21 Mar 2022
Copy link to clipboard
Print this post
The wattage shouldn't matter, you just have to set the resistors accordingly. I couldn't get mine to go and then with convid I couldn't get parts either and lost concentration although it has been on my mind as I really would like them to work.
I have been playing with direct connection most of the summer and that works OK but is pretty efficient and limited. I can get around 1Kw usually a bit less into a 3600W heater out of 5x 250W panels but I'd like the controller so I could get more power into a given element size.
I was mucking around with using an arduino to do a set switching rate of panels straight connected to a set of caps and then switched into a MOSFET into the element. That worked pretty well too. As long as the caps aren't engaged too long and fill up, I don't think the switch time is all that critical as whatever you have still gets sent to the element. I kept the dump to to 50 Hz as that is mains frequency and should allow use of the normal thermostat without frying it.
My main limitation is the lack of skills in laying things out. Tried mounting the setup it on a board and it was all a dogs breakfast and not secure. Have to revisit that again too and maybe buy a hot glue gun to stick things in place. Having Tony's design on a PC board was an asset in itself.
Another of his generous Gifts to the electronics Dilettantes like me.
Revlac
Guru
Joined: 31/12/2016 Location: AustraliaPosts: 1017
Posted: 01:55am 24 Mar 2022
Copy link to clipboard
Print this post
Had another look at the Hot water system (1800W) and its 150L, when we moved here we built a boiler with stove top and connected to the HWS, its wood fired and very handy to have. Cooked a lot on that stove and the excess heat went to the HWS, good to have after days of bad weather and little solar. The boiler holds maybe 80L of water and would be the reason it takes a bit longer for the solar power to heat it all upCheers Aaron Off The Grid
Revlac
Guru
Joined: 31/12/2016 Location: AustraliaPosts: 1017
Posted: 07:56am 19 Apr 2022
Copy link to clipboard
Print this post
A little progress, found an old Telecom 48V power supply, will use it to house all the parts for the hot water circuit board. The capacitors will be ganged together and hopefully covered, so nothing can enter and get toasted. The on/off switch will be cut into the cover on top and the heat sink will be at the back underneath. I might even put an old analogue volt meter on the front just for something to look at.....If I can find it....it was on a bench somewhere.
Cheers Aaron Off The Grid
cam1447 Newbie
Joined: 14/09/2024 Location: AustraliaPosts: 1
Posted: 02:33am 29 Sep 2024
Copy link to clipboard
Print this post
Hi,
Is anyone able to share me the gerber files for this design. I have tried reaching out to warpspeed but he may be no longer active in the forum.
Thanks!
rogerdw Guru
Joined: 22/10/2019 Location: AustraliaPosts: 852
Posted: 10:02am 29 Sep 2024
Copy link to clipboard
Print this post
He's fairly active over on the diysolarforum ... do a search on his name there perhaps.Cheers, Roger