 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. |
Forum Index : Electronics : Variable input voltage bat charger
Page 1 of 8   |
|||||
| Author | Message | ||||
| Janne Senior Member  Joined: 20/06/2008 Location: FinlandPosts: 121 |
Hello! Could anyone recommend a battery charger, that could be attached to a variable voltage / variable frequency AC source? The source in mind is an axial flux wind generator, with output voltage ranging from 80-300VAC, and frequency of about 10-35Hz. The system has a PLC, that would switch the charger on at 100V with a contaktor, and switch it off at 80V.. the values could be changed some if the charger won't accept voltages that low. The battery bank is 24V, and the suitable power for the charger is about 200-300W. I've asked a few manufacturers for a few chargers that i suspect would work well(switcing mode chargers), but no-one will confirm that their charger will work. It is also possible to use a rectifier with smoothing caps before the charger, if the low frequency is a problem, but the charger will accept DC. If at first you don't succeed, try again. My projects |
||||
sPuDd Senior Member Joined: 10/07/2007 Location: AustraliaPosts: 251 |
Assuming the wind generator is big enough to support the charger load without stall, then it should work. You could try it. Find say a, 30-50W 13.8V SMPSU charger and a spare car battery and give it a go. Just make sure the charger will do 80-240VAC. If it blows - at least you've learned something. If the windmill is not capable of the full 300W load, you may have to interface to the charger and dial the output voltage/current up and down with wind speed. Here's an example of this from bjblaster sPuDd.. Edit: At 300V you'll probably blow the charger - you will have to avoid this one way or another. It should work ...in theory |
||||
| Janne Senior Member Joined: 20/06/2008 Location: FinlandPosts: 121 |
Thanks spudd for the idea. We might consider modding a psu for the purpose. One bad thing is that most psu:s are for voltage range of about 12 to 16 volts, and we need a charger for the 24V battery bank. Also, the psu would need some modifications to prevent the battery from overcharging. If the charger can't handle the 300V max voltage, the PLC can be configured to drop the charger out from the circuit before that. The main load on the turbine are the heating elements, so losing the battery charger won't cause much of a load loss. If at first you don't succeed, try again. My projects |
||||
| Dinges Senior Member Joined: 04/01/2008 Location: AlbaniaPosts: 510 |
Janne, When I was modifying PC PSUs I also ran a test with a variac to see at what voltages it started (and stopped) working: http://www.fieldlines.com/comments/2007/10/18/16953/116/38#3 8 If you can accept the fact that the PSU starts to work at 110Vac instead of 80Vdc, and has full-rated output at 140Vac (those were the results for the one PSU I tested; others may behave differently, of course), then a PC PSU could be the way to go. It's easy to modify them to give up to 28Vdc output. Note that not all PSUs can achieve 28Vdc; most attain 26Vdc easily, but some I couldn't push beyond 24-25Vdc. Bit of 'luck of the draw', I suppose. If you use a PC PSU whose AC input can be switched over from 110Vac to 230Vac with a relay and some control circuitry that would give you an even wider input voltage dynamic range. I doubt the doubler (when PSU is set for 110Vac input) would work properly at 10Hz though. It's relatively easy to control the TL494 too (limit charge current). I recall Commanda/Amanda having designed a circuit (MPPT?) that should be relatively easy to interface to a PC PSU; I believe Oztules mentioned that fact somewhere in that thread too. Your requirements are quite heavy (10-35 Hz, 80-300Vac); the PC PSU likely can't cope with much above 250-260Vac (limited by the voltage rating of the input electrolytic capacitors, usually 400Vdc) Peter. |
||||
| Janne Senior Member Joined: 20/06/2008 Location: FinlandPosts: 121 |
Hi Peter. Yes the computer PSU would be a very cost efficient solution. It might work better if we just bypassed the rectifier and other AC components in the PSU, and build our own rectifier with larger input caps to smooth out the low frequency. The loss of power with only being able to use it 110V and upvards would imo be minimal, and would only affect performance on very low wind days. The 115/230V relay could be controlled with the contactors controlling the heating elements, or just use the auxiliary contacts on those contactors. It remains to be researched however, how the PSU will respond to changing that switch in "flight". (I know it will make a loud bang if you switch it to 115V while connected to 230V supply :D ) I noticed all your images on fieldlines are missing :( .. at least the stories are still there. I'll keep you all posted how this one turns out. As this one is not for my own use, I prefferred buying a ready made unit, but it now seems we can't find a suitable charger off the shelf. If at first you don't succeed, try again. My projects |
||||
oztules Guru  Joined: 26/07/2007 Location: AustraliaPosts: 1686 |
Janne, Peter is correct, the psu's from the computers are very tough to beat. I would use 2 in series (making sure you isolate the secondaries from earth/chassis. Use a rubber grommet between the circuit board and the chassis (screw) to achieve easy isolation. On the fly voltage switching should present no problem at all to the tl494. They will follow a sine wave if they have to (no filtering on the input)... so surges will not present a problem. The caps in the psu's will be 200v because of the doubler capability. If you rescue some old monitors, you will find 400vdc caps. Use these and some higher voltage transistors, and you may find you can direct inject the DC and skip the ac input section. With 2 in series, your power levels will be easily reached, and they will run cool at your 300w@28vdc A single unit will do it, but reliability will be marginal for pushing 300w through the transformer for long periods, and the rectifier diodes would run hot. .......oztules Village idiot...or... just another hack out of his depth |
||||
| Dinges Senior Member Joined: 04/01/2008 Location: AlbaniaPosts: 510 |
Thanks for the corrections, Oztules. Yes, 200Vdc caps, not 400Vdc. Though I'm not as sure as you yet that a PC PSU won't go defect after (many) transients from switching to 110<->230Vac. I'd test it out a few times under load first if I were Janne, before I'd start modifying anything. (Peter is a pessimist, Oztules is an optometrist) One thing I forgot: the asymmetry in operating voltage (PSU needs at least 110Vac to start to work, but once it's working continues operating down to 50Vac) is probably caused by the starting up (bootstrapping) of the TL494? So, if we'd inject external power (12Vdc or so) into the TL494 power pin, it may already start to work at 50Vac? Just a guess, but I think it'd be worth a quick test. Another thought I'm having now, after reading your remarks (or maybe you meant this in the first place, not sure): couldn't Janne use one plain (unmodified output, i.e. 12V output) PC PSU in series with one modified PSU? I.e. just one modified-output PSU + a stock one as opposed to two modified ones. The modified-output PSU could have adjustable voltage (28-12= ca. 16V) and current limiting. Any controlling of current and voltage could be done in that PSU, while the other, unmodified PSU, would need no current limiting (they're in series, so the 16V-PSU would do the current throttling) and voltage adjustment? Think personally I'd prefer to use one single ca. 500W PC PSU (they exist), modify that (perhaps even rewind the transformer for slightly higher voltage, 32V unloaded) and modify that for adjustable voltage and current. Still, it would be quite an undertaking to do all these modifications. Series connection would be much easier, but that gives another problem with grounding/earthing. And then there's the legal liability thing; do the CE/TÜV/KEMA (EU Underwriter Laborities) certificates still apply to such a heavily modified contraption... Good thing you and I don't live in a litigious society as the USA, I wouldn't even *think* of doing a project like this there with its sue-happy population (no offence intended to any Americans, except RonB). Peter. |
||||
| oztules Guru Joined: 26/07/2007 Location: AustraliaPosts: 1686 |
Dinges, You are quite right I think. Only one needs to be modified for this to work. The converted one will provide the current limiting and the voltage boost required... good thinking. The earthing is not an issue. Only the isolated 12v side of the board needs to be isolated.... so not an issue. The mods are also only on the isolated side of the transformer, so presents no problems that would not be encountered with a linear supply after the transformer. The input (mains side) is only seeing earth from the perspective of some filter caps (tiny), but they don't need to earth the dc. It can be all floating. The chassis will all still be earthed. .........oztules ps (Dinges, it is yet to be empirically determined if Ronb is either ........... No...I'd best be careful I know he has my co-ordinates  Village idiot...or... just another hack out of his depth |
||||
| Dinges Senior Member Joined: 04/01/2008 Location: AlbaniaPosts: 510 |
Yes, we'd better be careful indeed; he might sue....  |
||||
GWatPE Senior Member Joined: 01/09/2006 Location: AustraliaPosts: 2127 |
The computer SMPSU offer current lmited outputs. When used on a windmill, the battery loading will attempt to fully load the input while providing a voltage regulated output up to output current limiting, at which point the input voltage will be able to rise. The input voltage will be effectively held down until output current limiting occurs, or output loading decreases. These units will probably not be ideal impedance matching devices for a windmill. The windmill rpm may be held at the rate just above where the SMPSU converter starts to operate. The windmill will have to be powerful enough to create a current limit in the output before the alternator will become unloaded and rpm can increase, at which time the SMPSU will be at contstant power output. This is not the correct loading for a windmill. The equations will be different if there was current limiting offered at the SMPSU input as well. The fact that these SMPSU take a bit more volts to fire up and continue to operate down to lower voltage levels will compound the problem. Gordon. PS edit: If there was some way of having a voltage controlled current limiting, like I have on my windmill maximizer, then this would improve the SMPSU performance on a windmill. The current limit has to increase with input voltage so this should not be too much a problem for someone to test. Maybe Oztules or dinges has already modified units this way. A doubling of the current limit for a doubling of the input voltage will make a vast improvement to allowing the windmill to track the wind energy. A cubic relationship is more difficult, but squared is a good start. become more energy aware |
||||
| oztules Guru Joined: 26/07/2007 Location: AustraliaPosts: 1686 |
Gordon, The current limit is already using a voltage controlled error amp... very simple implementation. There is also no reason why you couldn't use the voltage error amp as a current error amp as well. One for A max, and one for A variable. It is very very flexible. You don't need to control V in this instance, as the batt will clamp it according to current limit. A max is necessary to stop the non converted unit going over current and switching in the scr clamp on the dead time control (pin 4 If I remember). If you set the gain on the amp carefully, you should be able to approximate the required curve if your keen, and allow it to operate as you suggest.... probably rpm:error voltage would do for a start. ......oztules Village idiot...or... just another hack out of his depth |
||||
| GWatPE Senior Member Joined: 01/09/2006 Location: AustraliaPosts: 2127 |
Hi oztules, I won't be putting my hand up for this. I was merely suggesting that a straight current limited SMPSU will require more than a mod to the regulated output voltage setting to be of use. There should not be any need to measure rpm as well for control purposes. In a system that exhibits close to optimum performance, the output voltage will closely follow rpm anyway. Someone familiar with SMPSU operation could quickly knock up a working test unit. Gordon. become more energy aware |
||||
| Janne Senior Member Joined: 20/06/2008 Location: FinlandPosts: 121 |
Hello agian. Could you offer some clue on what kind of psu's one could find the TL494 based circuits? I've already popped open a bunch of newish atx psu:s, but they all are based on various different PWM(?) ic's, and I can't seem to be able to find the datasheets to any of them.. Without datasheets my chances vary from slim to none. Even more odd is the 386 psu i also cracked open, that only has discrete components and one logic IC in it.. Still it seems the most interesting of the bunch, and I'm going to fiddle around a little more with it, mayby it will prove useful in some other project than this charger(too low power for this application) If at first you don't succeed, try again. My projects |
||||
Gizmo Admin Group  Joined: 05/06/2004 Location: AustraliaPosts: 5078 |
The early ATX and late AT power supplies often used a TL494. You can tickle the output voltage by playing around with the resistors on Pin 1. The resistors form a voltage divider, and if you decrease the value of the resistor connected from pin 1 to ground, you will increase the output voltage. But you have to be careful not to let the output voltage go over the voltage rating of the output filter caps, so make small changes first to get your bearings. Some old monitors also used a TL494, but the UC3836 or similar was more common, a 8 pin IC. Glenn The best time to plant a tree was twenty years ago, the second best time is right now. JAQ |
||||
| Dinges Senior Member Joined: 04/01/2008 Location: AlbaniaPosts: 510 |
Janne, Oztules is the residential PSU specialist, but I'll give it a go. What IC markings are there in your PSUs? It might well be that it's a direct equivalent to the TL494. I know that the KA7500(B) is much more common (at least I find a lot of those, and only rarely a TL494), for example. The KA7500 is a direct equivalent of the TL494. There are a few other ICs that are a direct equivalent too (e.g. DB494). And one other IC number (very strange number; can't even recall it now) that, after checking it out, turned out to be a TL494 in disguise as well. I haven't (yet) found any Unitrode UC... ICs in AT(X) PC PSUs; I did find them in other PSUs though (both consumer electronics and professional power supplies). Try posting your IC numbers here and maybe it'll ring a bell for someone. Peter. |
||||
| oztules Guru Joined: 26/07/2007 Location: AustraliaPosts: 1686 |
Janne, Dinges has said it all really. Try and find an old AT power supply from a clone. They seem to have used a fairly universal size and topology. Simple to unravel, and easy to work on. The name brands had some exotic shapes and circuits. Nearly all the supplies I have had contact with were clones and were TL494 based/KA7500. I found the 494 devices easy to work with, and the AT's easy to work in.... plenty of room. The ATX were a bit heavily populated.... and a bit smarter in some cases, which can be a nuisance some times. .........oztules Village idiot...or... just another hack out of his depth |
||||
| GWatPE Senior Member Joined: 01/09/2006 Location: AustraliaPosts: 2127 |
By the time ATX PSU were in, energy star ratings were in vogue. The old AT style had the mechanical switch, so if it was OFF, it was really OFF. The AT supplies were only rated 150W-250W, in the same pack size. More space inside. The only issue I see with the varying of current limit with input voltage would be isolation. This could be achieved with an opto-coupler. The 3-5mA at the full input voltage will present a bit of power though, that will not end up in the load. I know someone with schematics for a PC ATX type PSU. I will have a quick check to see what can be done. I was hoping that someone with experience in modifying a PSU for output voltage, would already have experience in the current limiting side as well. Gordon. become more energy aware |
||||
| GWatPE Senior Member Joined: 01/09/2006 Location: AustraliaPosts: 2127 |
I had a SMPSU from an old PC. this was 200W with a tl494 PWM. I was able to find website power supply mods and this gives some instructions on mods to use as radio power supply, 13.6V output. This used a model ptp-2008 cct design, of which I had the same. I read a bit further and decided not to go further, when the important value 60% efficiency at max output came to light. Seems a lot of power goes to waste. Approx 100W wasted to give 200W output. Here is the text. 
There would be a better way to do things, so I would explore other options. Gordon. become more energy aware |
||||
| sPuDd Senior Member Joined: 10/07/2007 Location: AustraliaPosts: 251 |
Gordon, do you think that’s normal efficiency for an ATX, or due to the mods done on it? ie; unbalanced output. I modded a recent ATX for 48V output, but didn't check before & after efficiency. sPuDd.. It should work ...in theory |
||||
| GWatPE Senior Member Joined: 01/09/2006 Location: AustraliaPosts: 2127 |
I was told once on this forum that these units were efficient. There were claims of modded units better than 95%. This guy has made some measurements and come up with 60%. I cannot confirm these numbers. Mods will affect efficiency. Maybe VA was measured on the input and the power factor was not 1. I suppose, now there will be a host of readers checking for answers. Gordon. become more energy aware |
||||
| Page 1 of 8 |
|||||
 Print this page |
