Anotherpower.com Forum
Renewable Energy Questions/Discussion => Automation, Controls, Inverters, MPPT, etc => Topic started by: birdhouse on February 19, 2012, 03:07:59 pm
-
i bought a charger from ebay for my 24v bank. it was sent straight from hong kong, and i believe it is designed to be the power source for LED lighting. it has an incoming voltage switch to go from 115 or 230v. there's also an adjustment screw for DC voltage out and a thermostatically controlled fan on top of the all aluminum case.
for $35 USD shipped, it seemed to be the best bang for my buck to be able to manually charge my battery bank with a generator running while at the same time running the generator for other uses.
i've also got a morningstar TS-60 in diversion mode hooked to the bank for my wind turbine, so that "should" make this operation fairly foolproof. not that i'm too worried about it, as i've got flooded batteries, and 15A going into a ~450Ah bank can't do that much harm.
Oz mentioned switching out a resistor on the voltage adjustment screw (trim pot??) to increase the voltage being produced by the charger. i wired it all up and hooked a volt meter to it while unloaded, and turned the screw. i got reading from below 19V and maxed the screw at 28.3V. i thought that was pretty good! it would be nice to hit 31V if possible to be able to use this charger for equalization as well.
questions: if i've got this charger running at 28.3 and say hook it to my bank at 23.5, i presume this will clamp the charger down to the lower bank voltage, but will this hurt the charger?
anyone know how to mod this? (Oz?) to get 31VDC
thanks for any input!
adam
(http://i995.photobucket.com/albums/af78/birdhouseadam/solar/15Acharger.jpg)
-
Hi Adam,
Different board all together to the one in mine from 7 years ago.... but your looks a simpler design ( mine had double sided and was delicate from memory).... no matter (looks the same from the outside though... well the box is the same at least)
The VR1 is your voltage control. It is simply part of a voltage divider network. It will have a series resistor going to it, and even perhaps one going away from it. The trim pot is just used to give you a small change to the resistance of one side of the divider.
I edited this part because it was cryptic:
As a simple start get some resistors about 10 time the value of the original in circuit ones close to the trimmer, and carefully place them across the resistor on the board, noting any change in voltage. The 10 times will change the value of the board resistor by about 10%.
I chose to vary the original value by 10% (by adding 10 times the original in parallel), because it will not hurt anything, but will give you an indication of which ones are in series with the pot...
Thankyou Rossw ( see rossw notes below)
end edit
When you find the one that pushes the voltage up or down, you know your in the voltage sense area. If it goes down, then use 10% less resistor in it's place, and you will probably be able to sweep the voltage up to 30v... if it goes up same thing, but you can just piggyback it on the original.
Does that make sense?
The other way is to pull the thing out of the box, and then trace where the lines go to and from the pot. Then you will know exactly which resistor to fiddle with..
Generally they pull one side of the voltage divider to ground, and as the chip sees the divider voltage fall, it tries to compensate by changing the pulse width (wider pulse) this pushes the voltage back up so the divider voltage it sees is comparable to it's reference voltage (usually another divider from the vref pin on the pwm chip.).
This style of supply has current regulation of generally 2 types. The first is an error amplifier that behaves like the voltage system above. It just gets a read of the current (usually by a single wire shunt) and feeds this to the error amp2. It then decides if the current is within spec, or too much. If too much, it backs off the pulse width to keep it in spec. So it will be held to batt voltage ( error amp v1 will remain out of the picture until the voltage gets too high) and the error amp (current) will take control.
Which ever error amp needs to over ride the pulse width will. So it will use the over voltage one if there is no load, and will use the current one if current exceeds a set amount.
The second one will shut the thing down for either a set period when it will try again, or until the load is removed... third type sets an scr, and need to be powered down before restarting at all (computer power supplies do this one. .... till you rip it out)
I have found they design these things with a lot of head room, and you can push them pretty hard .... particularly if you rewind the transformer with a bigger core, and change the current settings. I have pushed over 400 watts out of 150 watt ones without problem this way.
but I'm evil ;)
..............oztules
-
It would seem "likely" that the 1.2K (R41) and 1.5K (R40) are in series with the trimpot to limit its adjustment range.
Undoing the screws and checking where the tracks go on the board, from the trimpot (VR1) would confirm that pretty quickly.
If it is as I surmise, then whichever resistor is connected to the "high" end of the pot (ie, when it's adjusted to maximum voltage - logically the fully clockwise direction) could be slightly reduced to get you to 31V.
The least butchery way would be to tack a buffer resistor across it from the top. To reduce the 1.5K resistor a little (say, 10%), you could solder a 15K (brown, green, orange in the old 3-band resistors) across the 1K5 to reduce it to 1.363K.
Re-test your outputs and tweek the value if its not quite to your liking.
Edit: Damn, beaten by a minute. Yeah, what oz says.....
-
Sorry Ross :)
One more thing..... DO NOT REVERSE THE BATTERY LEADS TO THESE THINGS..... it won't be nice (done that too.)
If you think there is any likelihood of you doing an oztules, and something stupid (an oztules), then place a big diode across the + and - terminals with the cathode on the + and the anode on the - and an inline fuse in series with your leads.
That way when you do the inevitable, the fuse will blow and no one will know ;D
................oztules
-
As a simple start get some resistors about 10% of the ones close to the trimmer, and carefully place them across the resistor on the board, noting any change in voltage. The 10% I chose, because it will not hurt anything, but will give you an indication of which ones are in series with the pot...
Just a little clarification here. I know what oz meant... but the way he's written it could let out magic smoke....
When he says "Get some resistors about 10% of the ones"... he means "to CHANGE the combined value by 10%"
So if it's 1K, parallel 10K.
If it' 1.5K, parallel 15K
It's not perfect, but it's close.
The other thing - if you choose to solder them on, turn off and unplug the supply. If the iron is grounded (and it should be), bad things will probably happen if you touch a grounded iron to a live circuit. :)
-
first off, thanks a bunch for the help!
i'm pretty dumb in this realm!
so i just looked at the color coding chart on wiki. i finally understand what all those colors mean!
i did take the hint that oz meant 10% plus or minus of the actual rating. by either adding 10% or finding a resistor that is 90% of the original. now what i don't get, is:
So if it's 1K, parallel 10K.
If it' 1.5K, parallel 15K
wouldn't 10% of a 1000 be 1100? and 10% of a 1500 be 1650? i don't get where you get 10k and 15k respectively??
there's a radio shack three blocks away. i may just walk down there and buy a bunch of them and go from there. i think they're pretty cheap!
adam
-
i did take the hint that oz meant 10% plus or minus of the actual rating. by either adding 10% or finding a resistor that is 90% of the original. now what i don't get, is:
So if it's 1K, parallel 10K.
If it' 1.5K, parallel 15K
wouldn't 10% of a 1000 be 1100? and 10% of a 1500 be 1650? i don't get where you get 10k and 15k respectively??
Ahh, and that's *EXACTLY* the problem, and why I posted my comment.
If you have two 1K resistors (1000 ohms each) and put them in parallel, you end up with a combined resistance of....
.... 500 ohms! Ie, two identical resistors in parallel makes one effective resistor of only 50% the resistance.
If you have a calculator handy, the formula is R(effective) = 1 / ((1/R1) + (1/R2))
So using a resistor thats approximately 10 TIMES the value of the original, will reduce the value of the pair to about 90% of the original value.
Does that make sense?
-
so, again. very noob here with all this. yes, your last post makes complete sense and i now realize i didn't understand your post before that!
so the colors of R40 is brown green red. so 1500 so i should try adding a 15000??? ie brown green orange?
the color of R41 is red violet red. so 2700 so i should try adding a 27000??? ie red violet orange??
i really appreciate the small electronics for dummys course!
adam
-
Success!!!
by adding a 15K to the R40 spot, the voltage changed from 22.2 to 23.8!! (i had the pot turned down some while experimenting.
now to do some soldering!
thanks again!
adam
-
Thanks Ross..... I didn't think to explain what I was thinking.... good catch. Sounds like he's on the way now.
Adam, as a rule of thumb in these things if you change almost any electrical component by 10%, bad things don't normally happen ( unless it is mission critical stuff).
Ross explained it right, and I was thinking one thing and obviously writing another thinking you would know what I was getting at..... must remember to take my smart pills. sorry.
Keep us informed as to the outcome (final). get a little more headroom and aim to get to 31 or 32v at the upper end perhaps ( may want to scare up an equalise one day)
You can also change the trim pot and the resistors divider to get a much wider sweep range. These things are pretty forgiving ... ie end up with 24-34v
Ross can you change the original screed to reflect that in case someone does not read this far before playing please.... I no longer can.
...............oztules
scratch that Ross I was looking at the recent post page.... can't modify from there. It's fixed
-
thanks goes out to Oz and Ross. i'm not going to take a picture of my solder job, because i'm pretty sure you would all have a good laugh! now it tops out at 30.9V. perfect for equalize even when it's really cold out!
i've modded all sorts of carbs, engines, suspensions, chainsaws, ect, but this is the FIRST small electronics thing ive ever modded. couldn't have done it without the help from this board!
adam
-
I am not exactly sure whats going on here, but I am glad it works.
If the 15A charger gets up to 'whatever' voltage, the TS-60 will divert the extra power to keep the bank at the voltage the TS-60 wants, right?
Oz and Ross,
Anybody concerned the ICs beta will decrease the circuit's hysteresis?
Probably not a worry for this, just something that popped into my head.
Might be a good thing since the TS-60 is PWM.
G-
-
Ghurd-
i think i get what you're asking. the TS-60 should act as a controller of the "new" charger while in use. i've got a 50A dump load hooked to it, so that should be fine. i totally understand that while the bank is in absorb and float mode, i'd be dumping most of the power i just created via a genset.
not ideal by any means, but in the winter, especially during hunting season, i run a generator to power up a 32' airstream trailer that has so many inefficiencies it's crazy. so i figure, might as well charge up the bank since a gen set is running anyways.
sometimes, i run the trailer off the inverter for the night, and wake with a battery voltage somewhere around 23.8 (loaded). at this point, it is going to be really nice to fire up the genset, remove the trailer plug from the inverter, and into the genset, AND charge my batteries at the same time!
as time goes on, i want to replace all the lighting in the trailer with 24v lights. remove a few things, and get a much smaller/more efficient 120 to 12v converter to power the things that are a pain to change. IE: furnace blower/ignition, exhaust hood fan for cooking, oven light, ect. right now, the 120 to 12v converter is larger than a loaf of bread, and with NOTHING running is very warm to the touch. i might look into a computer power supply as its replacement... still thinking on that...
adam
-
Not sure I understand the question Ghurd.
in the op amps? or feedback loop? or......??
The output is stable to within probably 50th of a volt. set it where ever you want.... no need for a divert, it just closes up the pulse width at set V
Not sure I follow
There will be no discernible ripple on the output.
elucidate please? :)
...............oztules
-
"elucidate"?
Not ripple.
In my mind, the voltage divider current seemed kind of high given the values of the fixed resistors.
I did not consider a feedback loop.
I only considered the voltage drop in each half of the voltage divider when the output state changed. And with a large voltage divider current, changing something much (beta) could have a significant effect.
Seems kind of backwards thinking as I type it out.
Adam,
NOTHING is ever ideal!
If this modification is better than it was before, then it is closer to ideal.
32' Airstream, hunting season, and 50A dump load... I hope the dump load is inside the Airstream.
Makes me cold just thinking about it.
G-
-
"elucidate"?..... must have found it on a bubble gum wrapper in times past.
With these supplies, they need a current flowing in the transformer circuit all the time, so they need a minimum load, or the book keeping part has no knowledge of whats going on, and so can't make sensible decisions. This is done by a big resistor on the output that just soaks up current, and keeps the pwm working into a load, some of that load (tiny )can also be in the divider. Too high an impedance and noise can occur in the feedback loop, and instability is the result.
In the series of these I played with, they use a zener in series with the divider, so the currents are probably not what you expected.
The slope of the gain of the comparator is set in a feedback loop at the pwm chip, so is not likely effected....
The clincher is that they give you a trim pot to modify the voltage .... so they don't consider this an issue either if you think about it. ;)
Adam,
I was thinking slightly higher setting ability, because we don't know the slope of the comparator. It may ease off a bit fast as it gets close to Vmax, so if you want to push it hard in equalise to stir up the electrolyte, a bit of headroom may be useful to frighten em a bit ... or not. You'll know when you try it... or try it at home.
Just figure if your burning fuel to charge , you don't want to see half an amp for the last hour etc.... But testing is knowing.
Anyway it is easy to change as we have seen.
.................oztules
-
Ghurd-
it's actually quite comfortable. there's a brand new "furnace" in the airstream, ducted, and set with a thermostat. the dump load sits below my battery bank in hopes of taking the chill off a bit, though i think that's wishful thinking...
Oz- i started thinking of that last night. the charger will go to 30.9 right now, but that's unloaded. IE: open volts. i'm thinking it won't go quite that high under full load. i guess time will tell. thinking it's gonna be atleast a month before the snow levels allow me to drive in there.
adam
-
i was able to get up to my place last weekend, though couldn't drive in. got two trucks with 35" tires stuck. 4 wheeler did pretty good though!
i hooked the charger to a honda EU 1000 and the bank and fired it up. you could hear that it made the honda work quite a bit. battery volts started to rise immediately. (440Ah @24v). i ran the charger for around three hours. during that time, it wreaked of burnt electronic components, but the charger just kept chuggin away. no magic smoke that i saw.
then the TS-60 started being the smart end of the rig, and dumping power. they played nice together. it was hard to tell, but even though the charger would go to 30.7v unloaded, it seemed to taper off at 30v in use. the solar was still connected through a different mppt controller, so that is why it was hard to tell.
all in all, it was the best $35 charger i could imagine. just hoping that burnt elec smell is maybe just things breaking in?? kinda like new woodstove paint??
adam
-
I would guess that its just the new baking away as you suggested.
Time will tell.
You should post some pics of that fun road into your place. ;)
-
Some of the electrical insulation on products from India & China are from recycled rubber that is blessed with a creosote-burnt rubber smell.
You need to visually inspect the device for scorched circuit board area(s) and components near the hot spot(s) being affected even if the hottest places were within design specs...
-
Nice to see the inverter charger worked.
My experience with pwm power devices, is they work or they don't. The difference is usually measured in milliseconds or less.
Cant understand the electronic smell without total failure, unless the load resistor had foreign matter on it, as it will get hot.... not roasting hot, but too hot to touch. The output inductor can get toasty too, but will not fail under 200c. The next hottest thing is usually the diode block, but less than 100c, and finally the transistors.
The power transformer may heat a bit, and give off a hot lacquer smell, and I can't think of any other bit that will heat , and still function.
These things live on the edge of destruction every 30 thousandth of a second or less, they make the right decision and live, or the wrong decision and blow up ...... every cycle.....
After 3 hours, it all appears that the feedback loops are working, and your "burning" off all the wax and lacquer aromatics... hopefully.
I find them very reliable, and even use one to charge the electric car (bomb) when the wind don't blow.... and have done for years. It does 15A@43v.
...............oztules
-
This is an informative post in so much as it..
Shows that adjustable switch-mode power supplies do make acceptable cost efficient single stage chargers.
That the adjustment range can be changed.
The posts on achieving said increase in voltage range were good, multiple posts tend to stick in my mind better than a single flawless post.
It demonstrates the difficulty in explaining a subject at a level that assumes no previous knowledge, in this post it was achieved well by several posters.
It highlighted the possible pit falls, and gave solutions.
I also used one in the same circumstances as Birdhouse, while I was happy to have mine top out at 27.5V in the knowledge that it could not overcharge my unregulated lead acid batteries, I did wonder about raising the top voltage, now I would be happy to make adjustments.
Well done to all parties.
Allan
-
Shows that adjustable switch-mode power supplies do make acceptable cost efficient single stage chargers.
Not always.
I (go figure) have smoked them, almost always.
Mine were removed from something else, and designed for other purposes.
14~28VDCout?, to 7AH 12V battery, usually.
I guess the ones I tried were not current limited?
They were however VERY cost effective (free). ???
Nice to see the inverter charger worked.
My experience with pwm power devices, is they work or they don't. The difference is usually measured in milliseconds or less.
Agreed.
And agreed.
G-
-
G I suspect the problems you have experienced were due mainly to your location together with inadvertently selecting a southern hemisphere model.
Using the wrong model can cause the guyrencryptamonitor to malfunction, if you must use a southern hemisphere model turn it upside down, but don't impede the cooling fan air flow.
One can usually detect the intended area of use by looking for an 'S' or 'N' in the serial number, better luck next time.
Allan
-
Ghurd..... Alan is right.... its that hemispheric thing again... :o
Cheap pwm supplies )free by the thousand now I suspect where you are) come from computer power supplies.
They are fun to convert, and when you do, you have precise current and voltage control.
This one I use as a power supply for all my circuits under development, as I can limit the current and not blow up transistors.... any where from .2A to 18A and from about 7v to 28v...... and it only delivers the preset current if you short it out.... does not care.
[attachimg=1]
[attachimg=3]
Here is another in battery charging/maintenance mode, set for 13.9v max 5A. It drives the back up battery for the "in testing" solar hot water system.[attachimg=2]
You can tell from the spider webs on the current control opening, that this experiment has been running for a long time.... the board you can see has pwm for the motor as well, so in poor light, the pump runs slow, and as the temp in the panels gets a greater differential, it speeds up to full speed. Must do a write up of the system and circuits one day . :-[
................oztules
edit. Just noticed the bright spot on the scope..... am fiddling with a circuit for 8hz sine wave..... scan rate is very slow so the spot is NOT that bad in reality......my excuse and I'm sticking to it.... ::)
-
edit. Just noticed the bright spot on the scope..... am fiddling with a circuit for 8hz sine wave..... scan rate is very slow so the spot is NOT that bad in reality......my excuse and I'm sticking to it.... ::)
You must have felt me CRINGE when I saw that. I was going to write you a nastygram - but you owned up already!
-
wow-
never thought there would be this much interest in a chinese charger :D
It demonstrates the difficulty in explaining a subject at a level that assumes no previous knowledge, in this post it was achieved well by several posters.
this was key for me! i'm kinda an idiot with small electronics, yet the folks here basically held my hand through the mod, and it came out great! so thank you oz, and ross, and any others i forgot.
i'm thinking it was some chinese pcb's or something burning off in there. i mean the charger basically went from shelf, to FULL MODIFIED duty for three hours. the fan ran the whole time, but the air coming out wasn't all that warm.
next up is installing the boost circuit i bought for my wind turbine. that will take some hand holding for sure. :-\
and for wolv, since he asked:
(http://i995.photobucket.com/albums/af78/birdhouseadam/ranch/2012-03-25_14-26-37_815.jpg)
(http://i995.photobucket.com/albums/af78/birdhouseadam/ranch/2012-03-25_14-26-37_815.jpg)
and what i was working on: roof electrical:
(http://i995.photobucket.com/albums/af78/birdhouseadam/ranch/roofelec.jpg)
adam
-
G I suspect the problems you have experienced were due mainly to your location together with inadvertently selecting a southern hemisphere model.
Using the wrong model can cause the guyrencryptamonitor to malfunction, if you must use a southern hemisphere model turn it upside down, but don't impede the cooling fan air flow.
One can usually detect the intended area of use by looking for an 'S' or 'N' in the serial number, better luck next time.
Allan
That was the problem. The malfunctiong guyrencryptamonitor sent an upside-down spike and fried the rest of the circuit. The serial numbers verify the semiconductor components are intended for use in a metric hemisphere.
Oz,
The ones I played with (meaning smoked) were far smaller than PC power supplies.
Think like 12V to USB or newer cell phone charger kind of circuit sizes.
G-