Puppeteer

Author Topic: More panels!  (Read 29760 times)

vdubnut62

  • Full Member
  • ***
  • Posts: 211
    • View Profile
Re: More panels!
« Reply #15 on: March 03, 2018, 02:14:17 AM »
I'm not scared of the dark, I'm scared of the things that live in the dark..................

Stef

I think I am more afraid of what I'm afraid might be living in the dark that I'm afraid of.  ;D :o


No really, Glort would you mind terribly if I PM'ed you with some really stupid questions?
Ron
« Last Edit: March 03, 2018, 02:16:55 AM by vdubnut62 »
"The penalty good men pay for indifference to public affairs is to be ruled by evil men."   Plato.

"Remember, every time a child is responsibly introduced to the best tools for the protection of freedoms, a liberal weeps for the safety of a criminal." Anonymous

BruceM

  • Hero Member
  • *****
  • Posts: 3054
    • View Profile
Re: More panels!
« Reply #16 on: March 03, 2018, 04:04:29 AM »
My power addition project is still slowly moving along.  With the additional 1500 watts, I decided my old linear PV current regulator needed a bit more power.  I did a redesign and nearly doubled the capacity, 7 instead of 4 power transistors.  I just got the new PCBs today.
It should easily handle up to 3000 watts, but when I checked on the big heatsink (12x10x1.3 inches) with a thermal model it just didn't have the guts for the 300 plus watts the new board can do.  So I ordered a new heatsink, 24x10.1x2.8", with about 4 times the capacity.  It's the downside of linear regulation- a $160 heatsink!  Could have stuck with the old one with a thermal disk controlled small fan but I wanted to keep it outside the power shed and small muffin fans and outdoors don't mix well.  Nothing to fail with a big hunk of finned aluminum, also.

I'll assemble the new PV current regulator board on the first good day for soldering outside.  Hopefully I didn't make any big goofs on the PCB layout. 


BruceM

  • Hero Member
  • *****
  • Posts: 3054
    • View Profile
Re: More panels!
« Reply #17 on: March 03, 2018, 04:12:35 AM »
Hey Glort-  I was wondering if the Autopot valves are also spendy there in Sydney?  In the US the valve alone (for on pot or tray) is about $30 US.  The whole setups are also very spendy- $300 for a measly 6 pot setup with feeder tank.  Too steep for serious vegetable growing.


BruceM

  • Hero Member
  • *****
  • Posts: 3054
    • View Profile
Re: More panels!
« Reply #18 on: March 03, 2018, 04:48:30 PM »
"I'm a big fan of forced air."

I ran a couple different free heat sink thermal models a few days ago before deciding what to do for the power upgrade and even the most pathetic, barely moving airflow TRIPLES a heatsink's capacity compared to still air convection/radiation.  I WILL be using a modest cooling fan for my low EMF inverter! 

Thanks for the info on Autopot in Au.  It's a great product but as you suggest, the same principle can be done in other ways. 



BruceM

  • Hero Member
  • *****
  • Posts: 3054
    • View Profile
Re: More panels!
« Reply #19 on: March 04, 2018, 02:16:53 AM »
When the difference in temperature between the heatsink and the ambient starts getting small, then blowing more and more air (with an exponential increase in fan power consumption) will do less and less...that's the point you noticed experimentally where more air doesn't seem to help.

I got my new PV current regulator assembled and soldered today.  Now for a rough night with a whopping headache from the solder fumes.  I thought I dodged them but apparently not.



BruceM

  • Hero Member
  • *****
  • Posts: 3054
    • View Profile
Re: More panels!
« Reply #20 on: March 04, 2018, 04:47:16 AM »
This new PV current regulator  is really a simpler version of the one I've had in service for the last 8 years or so.  It's just a big linear series regulator. It's set up now for 0-5V current as control (through an opto-isolator) = 0-16 amps output (limited by the max output of the PV arrays).  It uses seven 350V NPN Darlington power transistors (BU323) that are tough buggers designed for automotive ignition systems.  It is a "low side" regulator, the PV negative is below the 0V of the 120V battery bank. 

So it's really 3 different boards to do the charge regulation;
1. A Battery Regulator board; float, bulk and equalize, temp compensated 12V shunt regulator for each battery (10 total), with shunt current feedback.  This lets me charge a large series string but baby each battery individually; that means adding only a gallon of water to the set of 10 every 2nd year.  It's designed for AGMs, which I may use for my next battery set depending on prices at the time.

2. PV linear current regulator; just a linear series regulator for the 120V nominal array (120-220VDC).  0-5V in for 0-16 amps current.  The previous design did fixed net charge selections of 0.5 amps (float), 2.5 amps (absorption), and max (PV limited to about 5.5 amps).  The new BBC (see below) charges with a continuous adjustment of current as the batteries come up to full charge so will charge faster. Much faster with larger PV and AGM batteries; the cheap marine batteries I'm using now have moderate internal resistance and can't really absorb high charge currents as AGMs could.

3. Battery bank charge controller (BBC).  This is a 3 year old newer all analog (op amps) design.  that computes the real time minimum charge current to keep all the battery shunt regulators just barely regulating without exceeding the 3 amp shunt current limit and automatically transitions from bulk to absorption to float.  It does this smoothly despite fast shifts in load or PV charge (sun). It also does the equalize timing. 

Other projects in the works- updating my generator- DC charger to a fixed voltage  (150VDC) 10 amp design that feeds into the PV regulator.  The current design uses a homebuilt motorized variac and adjusts itself via PICaxe controller as the charge current tapers off.  Charging is so infrequent that I may simplify this and ignore the minor loss of fuel efficiency of regulating via the PV charge regulator.  I only charge about 2-3 hours per winter, total. With the new PV capacity that will lkely be zero.

Lastly, the newer (3 year old) analog battery bank controller board has room for expansion with an added embedded microcontroller.   I may add that to continue to be able to use my remote terminals (LCD plus buttons) in the shop and house to monitor and control.  I'd use this controller to for my low EMI inverter control and power management like pumping the well or pumping air when sun is available and demanded loads are low.   

I'll take some pictures of the new boards on my bench tomorrow.







 


BruceM

  • Hero Member
  • *****
  • Posts: 3054
    • View Profile
Re: More panels!
« Reply #21 on: March 04, 2018, 03:29:00 PM »
I wish used panels were available here; I'd add one more set of 5 vertical panels for late in the day "sundown" power.  A set of 175's would be just dandy. 

BruceM

  • Hero Member
  • *****
  • Posts: 3054
    • View Profile
Re: More panels!
« Reply #22 on: March 04, 2018, 06:46:38 PM »
Here's some photos of my new boards for my custom 120VDC off grid power system. 

The power regulation is all linear, both on the Battery Regulators, and the PV current Regulator. 

The new Battery Bank Controller is also all linear but has the option to add a shielded processsor module via filtered DB25 connector.  It uses Microchip 5v micropower op amps so it's all up power use is about 6 ma. The processor module will add another 5ma.  We're talking tiny, tiny power use.

The PV current regulator uses 7 parallel BU9323 Darlington power transistors.  So for my intended max of 16 amps at 146V (2336W), each would carry only 2.2 amps.  It will be mounted on the back of a large 24x10.1x2.8" finned aluminum heatsink on the outside wall of my power shed.  The heat losses during throttling of PV output is does not affect charge or direct use efficiency.

The Battery Regulators (one for each 12V battery) can shunt up to 3 amps of current but in practice, that never occurs for a well matched set of batteries.  Repeated cycling with individual battery regulation makes the batteries match very closely so the Battery Regulators tend to serve more as a very small correction for battery impedance matching. The battery regulators each have a external thermal sensor that is attached to it's battery negative post.

Definitely not the way things are normally done these days but phenominally low EMI and quite efficient.  Bulk charge losses are very low- diode drop plus 0.8V at 20A power.  Winter PV efficiency is down around 17% due to mismatch of panel array voltage (higher than necessary).  Size and cost of throttled (linear instead of rapid on/off switching) charging is certainly significant but that was lowest on my design priorities.

Sorry for the "soft focus" photos- I'm a shaky today but was too lazy to get out the tripod.



« Last Edit: March 04, 2018, 06:48:13 PM by BruceM »

BruceM

  • Hero Member
  • *****
  • Posts: 3054
    • View Profile
Re: More panels!
« Reply #23 on: March 04, 2018, 06:55:37 PM »
Here's a photo of my prototype system- some while still in test and development in 2009.  I'm still using the same boards now.  Time flies.


carlb23

  • Hero Member
  • *****
  • Posts: 527
    • View Profile
Re: More panels!
« Reply #24 on: March 05, 2018, 12:31:55 PM »
Glort,

How many Kw of solar do you have now? 

ajaffa1

  • Hero Member
  • *****
  • Posts: 1725
    • View Profile
Re: More panels!
« Reply #25 on: March 12, 2018, 11:45:54 PM »
An interesting development, my neighbour is building a homemade off grid solar set up. He has been gathering second hand equipment for months. I have volunteered to assist in the hope of learning something useful.
It appears that he has got his hands on a couple of battery packs off forklift trucks, should provide good storage.
With his permission, I will try to post some photos and information once we get started.
Bob

BruceM

  • Hero Member
  • *****
  • Posts: 3054
    • View Profile
Re: More panels!
« Reply #26 on: March 13, 2018, 04:03:07 PM »
I'm very skeptical about DIY internet claims for lithium battery banks that have no cell management.  Even minor corner cutting on series cell voltage management as the Nissan Leaf has done has resulted in many premature battery failures, and that's starting with new matched cells.  As you know, Tesla is no cutting edge, new tech company; electric cars are 100 years old and their car has no "new" technology other than fancy electronic displays and other ancillary computer features.  Instead they have been wise adopters and refiners of proven, existing technology. They use the Texas Instruments lithium cell voltage management IC's so that no series cells are left voltage-uncontrolled.  Reports are that they do have very reliable battery life because of this. 

Even 48V wet lead acid systems could have better battery life by incorporating individual battery management, but no one does because it would add cost and people are more interested in today's cost and not the expense of battery replacement.  In fact, just talking about battery replacement cost would be bad for business.
« Last Edit: March 13, 2018, 04:12:44 PM by BruceM »

ajaffa1

  • Hero Member
  • *****
  • Posts: 1725
    • View Profile
Re: More panels!
« Reply #27 on: March 14, 2018, 02:24:23 AM »
Hey Bruce. when you say that individual battery management is not used because of cost, how complicated/expensive would it need to be to significantly improve life expectancy? Is this something that could easily be cobbled together out of cheap or second hand components?
Bob

BruceM

  • Hero Member
  • *****
  • Posts: 3054
    • View Profile
Re: More panels!
« Reply #28 on: March 14, 2018, 07:43:37 AM »
This Aussie guy doing his own DIY powerwall is using full (parallel) cell level monitoring, as far as I can tell.  His presentation glossed over the technical details but those packs look like all series cells are monitored for over/undervoltage.  Alas, it appears his charging system isn't automatically balancing the cells since he's talking 2 days to equalize his cells.  That's NOT how cell management is supposed to work.  It should divert power from charged cells to the slower cells and thus keep them well matched on every charge cycle.  If his system can't do that, and is requiring regular extended equalization that's not so good. Since he's selling stuff, it may be hard to get an honest answer about service life of his cells/packs.

I would not touch lithium right now myself...I'll wait until the price beats lead acid for stationary applications where light weight is not important.   By then, you'll likely see cell management integrated into the packs and standards will emerge so that things are not so proprietary and expensive.  Hopefully other more enlightened nations will pick up the ball and run with it.  It's hard to imagine that racks of 1000's of tiny laptop batteries are really the best we can do...though obviously for now, it is.

I'm sure if you google around you can find more in the way of lithium cell management systems.  I'd have to look hard at the TI IC solution...I don't like dirtying up my power with a bunch of unshielded embedded processors scattered throughout the battery bank but given the fan out and total numbers involved, their bussed serial communications are ideal.  Many of the newer AVR and PIC processors have multiple I2C serial ports available to help with the fan out. 

It's not a trivial piece of engineering, but given the price of a few thousand lithium cells, it's obviously worth it.


« Last Edit: March 14, 2018, 07:55:34 AM by BruceM »

BruceM

  • Hero Member
  • *****
  • Posts: 3054
    • View Profile
Re: More panels!
« Reply #29 on: March 14, 2018, 09:59:50 PM »
How "modules" or "packs" are assembled regarding some combination of parallel and serial cells is relatively arbitrary.  You would prefer to see management at the module/pack level to cover every group of parallel cells. The bottom line is that cells don't all have identical internal resistance for charging and discharging plus those individual cell charge/discharge rates are affected by temperature and age.  Lithium cells are voltage devices so that 100 cells in parallel can be treated for charge/discharge management purposes as a single larger cell.  Overcharging cells reduces their life dramatically, as does dropping the voltage too low.  I don't recall the common cell failure modes but I expect some means of disconnecting and bypassing a module with some sort of failed cell(s) might be necessary in a car to allow you to limp home. Better to keep the cells from failing by good charge management but it's all a trade off for cost, complexity, and performance.

I expect that the most cost effective battery management system for cars will be found only after many years.  The high current charge/discharge cycles of EV use proved very challenging for AGM batteries too, and so various methods were tried until settling on management at the 6 or 12V battery (3 or 6 cells) level, since going lower wasn't practical.  I'm happy to sit back with my wet lead batteries and wait for the bleeding edge of new battery tech to mature and get affordable.

A true breakthrough in batteries would certainly be an industry changer.  Every crackpot thinks cold fusion or free energy schemes are real and suppressed, and that a magic carburetor once provided 150 mpg in a 8 cylinder pickup truck.  Everyone of these schemes I've looked at closely was at best a gross error in measurement, most are outright frauds or incompetence.  I'm not saying conspiracies don't exist, just that for the stuff I've looked into there was no need for one except in the minds of the promoters.  At one time I was reviewing and rating over 150 (per year) Internal R&D projects of big companies doing cost plus contracts (advanced development) with the Air Force.  (I was just one of many reviewers.)  After a while you get good at spotting technical baloney.