If 2 12v batteries are too much for a pair of 3118's then what about a dc-dc step converter to limit the batt output to say 21 or 24 volts? I was wondering if nom12V batts when fully charged are more like 13V so exceeds 24V max? How do you charge an nominal 3x6v batt? Then would have to get a charge controller for solar panel at 18:V too?
http://www.ebay.com/itm/172015438869
This one is 1/4 the price, and surprisingly doesn't suck. It is the best of the Chinese DC-DC converters out right now that I have tested (which is over 2 dozen of them). If you want to go even cheaper there are options, but quality drops off fairly quickly with price. If you want something slightly better this is an option which is at least smaller, but now we are back up in price again.
DCDC-USB-200, Intelligent DC-DC converter with USB interface
EDIT: Fixed ebay link
This one is 1/4 the price, and surprisingly doesn't suck. It is the best of the Chinese DC-DC converters out right now that I have tested (which is over 2 dozen of them). If you want to go even cheaper there are options, but quality drops off fairly quickly with price. If you want something slightly better this is an option which is at least smaller, but now we are back up in price again.
DCDC-USB-200, Intelligent DC-DC converter with USB interface
EDIT: Fixed ebay link
Last edited:
Don't know how these are but might be worth a try for $5.
Online Shop DC CC C-D CV 9A 300W Step Down Power module 5-40V to 1.2-35V Buck Converter|Aliexpress Mobile
Online Shop DC CC C-D CV 9A 300W Step Down Power module 5-40V to 1.2-35V Buck Converter|Aliexpress Mobile
non-mobile-link:
DC CC C D CV 9A 300 Watt Step Down Power module
Dunno how good these buck-modules are but know the boost-modules made in the same manner/layout/construction. Full power is only achievable at small voltage-step-up/down steps.
I.e. stepping up from 11V -> 34V , the modules were hitting the 100W, even if sold as 150W. Thing getting really better when doing 24->34V.
(After i changed the inductors, Fet, Heatsink and Caps)
The LTC3780 modules seems a good deal/implementation with their bridge configuration. Actually implementing a battery charger based on the LTC4020, which is quite similar at the power stage.
Last edited:
So you are suggesting to spend $80 on the power supply for an amp which is under $10 for both channels? At that point, buy a name brand amp and be done with it. The mini-box unit I posted is proven reliable for over 100 hours at full load from -40C to 85C. I did the test myself and still use that board as a beater around the lab, and it's still $20 less.
I get that paying for quality makes sense in a lot of cases, but I don't think a <$10 chinese chip amp is one of them.
I get that paying for quality makes sense in a lot of cases, but I don't think a <$10 chinese chip amp is one of them.
Mord, the buck mode 150W modules perform worse out of the box than the boost mode. Can't speak for modified. The LTC3780 modules have really impressed me so far, great for running on battery too since they have an adjustable undervoltage cutoff. The LF ripple is almost non existent, though the HF noise could be better. I imagine swapping the caps would net a fair improvement.
So which unit are you guys recommending when you say LTC3780? Is that the regulator buck/boost chip used in the cheap $5 units? I agree swapping out everything but the PCB is excessive.
I also agree that an $80 converter for a $16 amp is excessive.
Interesting that the boost mode works better than buck. These could be useful 19v laptop chargers from car battery 13.8v.
I also agree that an $80 converter for a $16 amp is excessive.
Interesting that the boost mode works better than buck. These could be useful 19v laptop chargers from car battery 13.8v.
Yeah, it also pretty much depends on the input power supply impedance and inductors saturation current.
The stock configuration of the boost-modules gave up on ~30-40W. I then changed for a VER2923 i had laying around. Things went better but the fets cooling was weak. Changed that, next bottleneck where the caps ESR and current handling.
I ended up with MA5173-AE (Coilcraft 7uH), 2 Panasonic FR 820uF 35V and a 6mR FET on a big copper block. So the concept "works" with the typical tradeoffs/flaws a simple switcher got at high current + high step up ratio.
I did some effort into this in the past with simulations here:
10/30/50/100W LED application driver (UC3843A) - #360customs
Putting some Litium-Ion in front at a 6-8s configuration would be optimal.
So for finding a good buck/boost, input switching current is the primary limiting factor to look for. More phases are always welcome. (Like TIs 4 phase automotive boost board)
@xrk,
the $5 buck/boost modules are implemented by UC3843A or TL494 mostly in current-mode-control.
The LTC3780 boards are well priced at $15.
While this:
http://www.ebay.com/itm/DC-DC-Conve...matic-Step-UP-Down-Power-Module-/111580295460
seems to have some LC-filtering at in/output, changing the fuse would be an adventure.
Like Jensen said,
this design looks good:
http://www.ebay.com/itm/LTC3780-DC-...-Down-Regulator-Charging-Module-/322049388287
Even if through hole caps would have lower ESR and higher current handling.
The stock configuration of the boost-modules gave up on ~30-40W. I then changed for a VER2923 i had laying around. Things went better but the fets cooling was weak. Changed that, next bottleneck where the caps ESR and current handling.
I ended up with MA5173-AE (Coilcraft 7uH), 2 Panasonic FR 820uF 35V and a 6mR FET on a big copper block. So the concept "works" with the typical tradeoffs/flaws a simple switcher got at high current + high step up ratio.
I did some effort into this in the past with simulations here:
10/30/50/100W LED application driver (UC3843A) - #360customs
Putting some Litium-Ion in front at a 6-8s configuration would be optimal.
So for finding a good buck/boost, input switching current is the primary limiting factor to look for. More phases are always welcome. (Like TIs 4 phase automotive boost board)
@xrk,
the $5 buck/boost modules are implemented by UC3843A or TL494 mostly in current-mode-control.
The LTC3780 boards are well priced at $15.
While this:
http://www.ebay.com/itm/DC-DC-Conve...matic-Step-UP-Down-Power-Module-/111580295460
seems to have some LC-filtering at in/output, changing the fuse would be an adventure.
Like Jensen said,
this design looks good:
http://www.ebay.com/itm/LTC3780-DC-...-Down-Regulator-Charging-Module-/322049388287
Even if through hole caps would have lower ESR and higher current handling.
Last edited:
This is the LTC3780 module I am discussing.
1pc LTC3780 Automatic Lifting Pressure Constant Voltage Step Up Down 10A 130W | eBay
The LTC3780 is the controller chip which controls the FETs to do the buck/boost conversion. Most of the Chinese DC-DC modules DO NOT use a name brand chip like the LTC, but instead rely on generic chips, often scratched off.
1pc LTC3780 Automatic Lifting Pressure Constant Voltage Step Up Down 10A 130W | eBay
The LTC3780 is the controller chip which controls the FETs to do the buck/boost conversion. Most of the Chinese DC-DC modules DO NOT use a name brand chip like the LTC, but instead rely on generic chips, often scratched off.
Yeah, in Boost mode your output power will be limited by the input current and your input voltage. In buck mode your output power is limited by the output current and voltage. Of course efficiency and component selection all plays into this big time.
I have taken my LTC3780 module to its limits and it has performed admirably out of the box. Increase cooling if you want to run at the 10A limit (in or out) for more than a few minutes though. 6A and below is fine stock.
EDIT: Mine says it is a "LIAOWHANGZE", but I have seen various brands on what is essentially the same PCB. There is a version of this PCB that has the heatsink mount on the back so you can bolt on a nice big sink to it, usually costs about $5 more for that version, but if you want to push it hard might be worth it.
EDIT AGAIN: Link to the other version (same thing, different heatsink layout)
http://www.ebay.com/itm/LTC3780-Aut...612320?hash=item2104a17aa0:g:GY0AAOSw9mFWLZdv
I have taken my LTC3780 module to its limits and it has performed admirably out of the box. Increase cooling if you want to run at the 10A limit (in or out) for more than a few minutes though. 6A and below is fine stock.
EDIT: Mine says it is a "LIAOWHANGZE", but I have seen various brands on what is essentially the same PCB. There is a version of this PCB that has the heatsink mount on the back so you can bolt on a nice big sink to it, usually costs about $5 more for that version, but if you want to push it hard might be worth it.
EDIT AGAIN: Link to the other version (same thing, different heatsink layout)
http://www.ebay.com/itm/LTC3780-Aut...612320?hash=item2104a17aa0:g:GY0AAOSw9mFWLZdv
Last edited:
Well..., heck yes! I have used other sino converters and they just do not last here in FL, on the saltwater, in boats...6 months max then corrosion effs them up. I just thought sailorman might want a ready to go solution that will be working long after you all are moved on to whatever new DIY chip is in vogue...or not...mine is still working flawlessly three years after purchase.
Last edited:
On a boat with saltwater is a bit out of what most people do, although I didn't notice the name was sailorman. Is it going on a boat? Also, have you tried any of the models from SurePower (cooper bussmann)? I ask because I recently tested some at work, and they performed very well, but we don't do any salt spray testing or anything.
Mord:
I have seen those other LTC3780 boards and will try one out at some point. Didn't notice the additional filtering, thanks for pointing that out. I needed the CC functionality for battery charging when I got the first one anyway though.
Last edited:
from sailormanbigd: "I was wondering if nom12V batts when fully charged are more like 13V so exceeds 24V max? How do you charge an nominal 3x6v batt? Then would have to get a charge controller for solar panel at 18:V too?"
So I am assuming he is on a boat and charging batteries from solar panel(s). Since I am doing that I thought the Victron that I use would be ideal as sailormanbigd would not have to change his controller or batteries. Just hook up the Victron to battery (ies) and to his amp(s) and enjoy the music. The saltwater off grid environment is different than living up north in apts or houses
anyway...back to the tunes.
So I am assuming he is on a boat and charging batteries from solar panel(s). Since I am doing that I thought the Victron that I use would be ideal as sailormanbigd would not have to change his controller or batteries. Just hook up the Victron to battery (ies) and to his amp(s) and enjoy the music. The saltwater off grid environment is different than living up north in apts or houses
anyway...back to the tunes.
The boards are the same, just different heatsinks. Having the heatsink at the bottom isn't optimal due to missing thermal vias, having it on top, isn't optimal due to case thermal resistance. So both not optimal.
But if it works, why not.
Fets are ~8-10mR:
http://www.aosmd.com/pdfs/datasheet/AOD4184A.pdf
t-head: I have some solar panels on my roof that charge up a bank of LiFePO4 cells, and I live nowhere near the water (OK, not really that far, but not on it either). So I didn't think anything of that. But with that username, and that scenario, you may well be correct. Although he will still need to DIY waterproof the amps, so could just stick any DC-DC board into whatever enclosure they will be in, and assume the whole thing will eventually corrode away
Mord: I should have mentioned, main reason the sink on the bottom will help, is because you can still put a sink on top as well, so you get 2x crappy cooling = slightly less crappy cooling
Thanks for pulling up the data on the fets.
Mord: I should have mentioned, main reason the sink on the bottom will help, is because you can still put a sink on top as well, so you get 2x crappy cooling = slightly less crappy cooling
Thanks for pulling up the data on the fets.
Jensen: I should have mentioned, main reason the sink on the top will help, is because you can still put a sink on bottom as well, so you get 2x crappy cooling = slightly less crappy cooling
Btw. Tracking says you haven't picked up the pcbs from your local post office yet.
@xrk,
for pure boost, you may go for those:
www.aliexpress.com/item/600W-DC-DC-...ge-9-60V-turn-12-80V-48V-72V/32507014533.html
(Still 10A input current limit)
Last edited: