Question on assembly

oldmanplayingt
oldmanplayingt Registered Users Posts: 9
Hi im new here. If you guys could help me that would be great. I'm pretty lost

So i have a bunch of old parts that I want potentially use.
- 2 solar panels that give 20W each (16.4V, 1.22A)
- 2 solar controllers that came with the panels (the controllers say 12V, 4.5A)
- a bunch of deep cycle batteries (6)

I want to hook up my panels to be able to charge my batteries and then have my batteries power an access point. I would like the whole system to be self-sustaining also.

Im also a little confused as to what solar controllers do. Do solar controllers already have circuitry that will prevent overcharging? And they wont change my output to the 12V, 4.5A right? And also what type of cables do i need to connect 12V batteries? Do i need another solar panel?

My access point requires 12V, 6.7Amps

Yeah a lot of questions. Thanks for you help guys

Comments

  • mike95490
    mike95490 Solar Expert Posts: 9,583 ✭✭✭✭✭
    Re: Question on assembly

    You will need more panels, or run your access point only a couple hours a day.

    You have 40W of PV, that can harvest in a 5 hour solar day, 200Wh

    You have an AP, that consumes 80.4w each hour, in 2.5 hours, it will consume 200Wh

    Then there are losses in the system I've not accounted for.
    Powerfab top of pole PV mount | Listeroid 6/1 w/st5 gen head | XW6048 inverter/chgr | Iota 48V/15A charger | Morningstar 60A MPPT | 48V, 800A NiFe Battery (in series)| 15, Evergreen 205w "12V" PV array on pole | Midnight ePanel | Grundfos 10 SO5-9 with 3 wire Franklin Electric motor (1/2hp 240V 1ph ) on a timer for 3 hr noontime run - Runs off PV ||
    || Midnight Classic 200 | 10, Evergreen 200w in a 160VOC array ||
    || VEC1093 12V Charger | Maha C401 aa/aaa Charger | SureSine | Sunsaver MPPT 15A

    solar: http://tinyurl.com/LMR-Solar
    gen: http://tinyurl.com/LMR-Lister ,

  • mikeo
    mikeo Solar Expert Posts: 386 ✭✭✭
    Re: Question on assembly
    My access point requires 12V, 6.7Amps
    What kind of access point, mine requires less then 500 ma?
  • BB.
    BB. Super Moderators, Administrators Posts: 33,431 admin
    Re: Question on assembly

    First thing--get the batteries on some sort of charging cycle. If they set for months without charging, they will sulfate and die.

    Next, you really need to understand your loads. Say your access point averages 3 amps @ 12 volts and you want it to run 24 hours per day:
    • 3 amps * 12 volts * 24 hours = 864 Watt*Hours per day usage
    Battery wise, we recommend 3 days of no-sun and 50% maximum discharge:
    • 864 WH * 3 days * 1/0.50 max discharge * 1/12 volts = 432 AH @ 12 volt battery bank...
    The above is certainly conservative--If you wanted to have only 2 or 1 day of sunless energy--you could redo the calculations for a smaller battery bank (probably ~200 AH minimum).

    Next, assuming you live in a reasonably sunny area and get 4 hours of sun per day 9 months of the year (winter may be a whole other issue), your batteries are flooded cell 80% efficient and we derate the solar panels+charge controller to 77%:
    • 864 WH * 1/0.77 derating * 1/0.80 batt eff * 1/4 hours of sun per day = 351 watts of solar panels to run load
    Of course, the weather is variable, and you may make different assumptions than I would, and/or you have other requirements (summer only, backup genset, or can recharge from utility power).

    Regarding the solar charge controllers--Bascially, they pass the current from the solar panel to the battery bank until the battery reaches ~14.4 volts or so... Then the controller starts opening and closing a switch to keep the battery at 14.4 volts (reducing current flow). At night, the controller shuts down and waits until the next morning to bring the battery back up to 14.4 volts.

    Of course, that is the description of a very simple charge controller... There are much more sophisticated ones that have multiple set points (bulk charge, float charge) and other functions too. Usually, those are larger capacity controllers and much more expensive too.

    In the end, measuring your load Amp*Hour / Watt*Hour consumption and planning your solar off-grid setup to handle the loads is the usual steps to take (after you have found the most energy efficient loads you can--conservation is almost always cheaper than building out an off-grid power system).

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • oldmanplayingt
    oldmanplayingt Registered Users Posts: 9
    Re: Question on assembly

    Oh good, point i totally overlooked the math. Im definitely going to get some more solar panels then. Any suggestions/recommendations?
    (no real restrains on price, but i would like get the best bang for my buck:D)
  • BB.
    BB. Super Moderators, Administrators Posts: 33,431 admin
    Re: Question on assembly

    You might try the host of our forum. They stand behind their product and try to have good quality manufacturers on their site:

    Northern Arizona Wind & Sun

    None of us here (other than our Host/Admin Windsun) is associated with NAWS -- other than possibly being the occasional customer (including Niel and myself--volunteer moderators/spam patrol).

    You really need to define your loads and requirements before you start purchasing components. One good way to do this is an Amp*Hour / Watt*Hour DC meter. If you are going to do this with an AC Inverter--then a Kill-a-Watt meter is a great deal (also very handy for use around the home for conservation--assuming you are in North America 120 VAC 60Hz).

    Before we get into recommending specific hardware--it would really be helpful to know how much power you want to use and, roughly, where you are located. You cant 20 watt panels or 200 watt panels. 4 amp solar charge controllers or 60-80 amp controllers.

    And solar does not "scale well"... I.e., smaller systems use smaller components. Larger systems tend to use much larger hardware and it is not usually worth the trouble to "bring the small stuff to the larger system"...

    In the end, do a paper design and parts list... And figure out the costs between options/variations.

    Since you already have the solar components--you might as well hook them up and experiment.

    Remember that storage batteries (in good shape) can output 1,000's of amps into a dead short. You should always use an appropriate size fuse for the battery interconnect wiring... For example, if you use 14 awg house wring--that is good for 15-20 amp loads--so you should fuse at 15-20 amps for safety.

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • AntronX
    AntronX Solar Expert Posts: 462 ✭✭
    Re: Question on assembly
    My access point requires 12V, 6.7Amps

    Are you sure about 6.7 Amps? Did you actually measure this current with DC Ammeter connected between the power supply and the access point? This does not seem right, unless it's a desktop computer you are powering. You really need to figure out how to reduce the power demand down to 12V 0.5 - 1 A, otherwise you are looking at about 1000W of solar panels and a lot more batteries, if you want to power this thing 24/7.
  • oldmanplayingt
    oldmanplayingt Registered Users Posts: 9
    Re: Question on assembly

    So im going to borrow a amp meter to measure how much my ap actually draws. Hopefully the draw is a lot less than 6.7amps. My plan is connect my batteries straight to the ap so it will bypass the inverter, and keep everything at 12V.

    Another question I have is about my batteries. I have 6 12V car batteries. How should I wire them up? all in parallel?
  • mike95490
    mike95490 Solar Expert Posts: 9,583 ✭✭✭✭✭
    Re: Question on assembly
    ........ I have 6 12V car batteries. How should I wire them up? all in parallel?


    You will soon (weeks) have 6 dead batteries. Car batteries are not designed for deep cycle usage, and WILL fail very quickly.

    For connecting batteries in parallel, Smartgauge has a great explanation on how to connect on the Diagonal.

    Connecting more than 3 in parallel gets tricky, with fuses and such, I'd suggest using 6V golf cart type batteries, 2 in series, and the 3 series strings in parallel, to get the same wattage as 6, 12v in parallel.

    Fuses, I'd suggest the Blue Seas terminal mount fuse. ( 3 of them, 30 amp rated each)
    5191.jpg 100.2K
    Powerfab top of pole PV mount | Listeroid 6/1 w/st5 gen head | XW6048 inverter/chgr | Iota 48V/15A charger | Morningstar 60A MPPT | 48V, 800A NiFe Battery (in series)| 15, Evergreen 205w "12V" PV array on pole | Midnight ePanel | Grundfos 10 SO5-9 with 3 wire Franklin Electric motor (1/2hp 240V 1ph ) on a timer for 3 hr noontime run - Runs off PV ||
    || Midnight Classic 200 | 10, Evergreen 200w in a 160VOC array ||
    || VEC1093 12V Charger | Maha C401 aa/aaa Charger | SureSine | Sunsaver MPPT 15A

    solar: http://tinyurl.com/LMR-Solar
    gen: http://tinyurl.com/LMR-Lister ,

  • BB.
    BB. Super Moderators, Administrators Posts: 33,431 admin
    Re: Question on assembly

    And, yes, you will wire the car batteries in parallel... However, there should be a fuse on the + terminal of each battery to protect your wiring from shorts.

    Also, for long term battery life--you need to wire the batteries in parallel in a particular pattern to distribute charge current and loads evenly among them:

    Smart Gauge website

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • oldmanplayingt
    oldmanplayingt Registered Users Posts: 9
    Re: Question on assembly

    oops correction they are deep cycle marine batteries.

    can you guys help me on figuring what gauge cable i will probably need? My batteries are going to be put relatively close to each other in a rack
  • mike95490
    mike95490 Solar Expert Posts: 9,583 ✭✭✭✭✭
    Re: Question on assembly

    6 deep cycle batteries, at about 100Ah each,

    with 40W of PV, (3A) will NOT even keep ahead of the batteries self discharge.

    You need at least 10A of charge, and preferably 40A, to replace any drain on the batteries.

    Best to think this through again, and measure that access point.
    Powerfab top of pole PV mount | Listeroid 6/1 w/st5 gen head | XW6048 inverter/chgr | Iota 48V/15A charger | Morningstar 60A MPPT | 48V, 800A NiFe Battery (in series)| 15, Evergreen 205w "12V" PV array on pole | Midnight ePanel | Grundfos 10 SO5-9 with 3 wire Franklin Electric motor (1/2hp 240V 1ph ) on a timer for 3 hr noontime run - Runs off PV ||
    || Midnight Classic 200 | 10, Evergreen 200w in a 160VOC array ||
    || VEC1093 12V Charger | Maha C401 aa/aaa Charger | SureSine | Sunsaver MPPT 15A

    solar: http://tinyurl.com/LMR-Solar
    gen: http://tinyurl.com/LMR-Lister ,

  • BB.
    BB. Super Moderators, Administrators Posts: 33,431 admin
    Re: Question on assembly

    The gauge of the cables depends on the maximum current (charging or load) that you plan to take through the cables (take maximum sustained current * 1.25 for NEC safety factor).

    Also, voltage drop does matter (especially for 12 volt systems). If you have long runs and/or high surge currents (starting a well pump for example)--then you need to up size the wire/cable such that the total run from the battery bank to your load/charger does not exceed some amount of drop... For a 12 volt battery system, probably a few tenths of a volt.

    Say you want to run the battery (with loads) to 11.0 volts. And your device takes 100 amps surge with a 10 foot run and a minimum 10.5 volt operational voltage... So, you can withstand a 0.5 volt drop.

    For example, using this calculator and a little bit of playing with values (remember, for every 3 AWG change, the diameter of the wire doubles/halves). Some values (using nominal temperature conditions):
    • 14 awg = 6.1 volt drop (house size wiring for 15 amp load)
    • 10 awg = 2.4 volt drop
    • 6 awg = 0.95 volt drop
    • 4 awg = 0.6 volt drop
    • 2 awg = 0.38 volt drop
    And depending on a lot of factors, a 100 amp current can be carried on ~6 to 1 awg wire gauge minimum.

    If you want/need to design for a 100 amp load, then the wire/fuse/circuit breaker should be sized for 100*1.25=125 amps minimum.

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • oldmanplayingt
    oldmanplayingt Registered Users Posts: 9
    Re: Question on assembly

    sorry im jumping from question to question. And thanks guys for the help, i really appreciate it.

    So using an multimeter fluke i was finally able to find the 110VAC input current to be about .1amps. It goes through an inverter within my ap to 12V, does this mean it requires .1amps at 12V? Or is there another way i need to measure the current of it in dc. And this would be the current i would need to determine the load that i have right?
  • mike95490
    mike95490 Solar Expert Posts: 9,583 ✭✭✭✭✭
    Re: Question on assembly
    sorry im jumping from question to question. And thanks guys for the help, i really appreciate it.

    So using an multimeter fluke i was finally able to find the 110VAC input current to be about .1amps. It goes through an inverter within my ap to 12V, does this mean it requires .1amps at 12V? Or is there another way i need to measure the current of it in dc. And this would be the current i would need to determine the load that i have right?

    An inverter ?

    maybe a power brick step-down transformer. What are the specs on that ? It's required to be printed or molded on it. Look for the output spec. 14v 100mA or something. or even the input 115VAC @ .3A

    Or maybe even the AP has a sticker that calls out the center pin polarity, voltage and amps. Or give us the mfg & # of the AP, and we'll look it up.
    Powerfab top of pole PV mount | Listeroid 6/1 w/st5 gen head | XW6048 inverter/chgr | Iota 48V/15A charger | Morningstar 60A MPPT | 48V, 800A NiFe Battery (in series)| 15, Evergreen 205w "12V" PV array on pole | Midnight ePanel | Grundfos 10 SO5-9 with 3 wire Franklin Electric motor (1/2hp 240V 1ph ) on a timer for 3 hr noontime run - Runs off PV ||
    || Midnight Classic 200 | 10, Evergreen 200w in a 160VOC array ||
    || VEC1093 12V Charger | Maha C401 aa/aaa Charger | SureSine | Sunsaver MPPT 15A

    solar: http://tinyurl.com/LMR-Solar
    gen: http://tinyurl.com/LMR-Lister ,

  • oldmanplayingt
    oldmanplayingt Registered Users Posts: 9
    Re: Question on assembly

    sorry pardon my poor electrical vocab, it goes though some type of circuit that changes my initial AC input to DC. the sticker on the board says input is 100AC-480VAC 50/60Hz 2.0A Max and out put as 12V 6.7A (80W)

    model name is: AIR-LAP1522AG-A-K9

    Im really just trying to figure out how big my load is so i can know what panels i am going to need to buy.
  • nsaspook
    nsaspook Solar Expert Posts: 396 ✭✭✭
    Re: Question on assembly
    sorry pardon my poor electrical vocab, it goes though some type of circuit that changes my initial AC input to DC. the sticker on the board says input is 100AC-480VAC 50/60Hz 2.0A Max and out put as 12V 6.7A (80W)

    model name is: AIR-LAP1522AG-A-K9

    Im really just trying to figure out how big my load is so i can know what panels i am going to need to buy.

    This thing is a power hog. It really needs 50W min at 12vdc, so 6.7A is a real number.

    http://www.cisco.com/en/US/prod/collateral/wireless/ps5679/ps8368/data_sheet_c78-532987.html
  • BB.
    BB. Super Moderators, Administrators Posts: 33,431 admin
    Re: Question on assembly

    The reading on the meter may real (probably no DC load)... Or, you could have the meter set to measure DC amps instead of AC amps...

    When using the DC/AC amps--typically you want to plug in / configure for / the 10 amp full scale readings...

    And to make things more confusing, some meters will measure a DC current when set to AC scale (if true RMS meter) or may not (cheaper non-true RMS meter).
    • Watts=Volts * Amps (* power factor) [if PF does not equal 1.0]
    • Power = Watts = 110 VAC * 0.1 amps = 11 watts
    That is a non-trivial amount of power when used on an off-grid system (if this is the "stand-by loading" of the power brick). Usually it is a very good idea to turn things off (AC and DC) when not actually using the device for Off-Grid systems.

    For AC loads (plug in appliances)--get a Kill-a-Watt meter. Really handy and surprisingly accurate. Also great to use around your home to find power hogs when trying to conserve.

    For DC loads--if you are going to do a lot of this--Look at these DC Amp*Hour / Watt Hour meters.

    These will help you understand your power needs and off-grid system's capabilities better.

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • oldmanplayingt
    oldmanplayingt Registered Users Posts: 9
    Re: Question on assembly

    I was using one of these to measure the current:
    http://www.sperryinstruments.com/product/dsa500a

    Im going to assume that going about this project still trying to bypass the AC stuff would always going to be most efficient right? (12Vbattery to 12V input of the ap)

    I'm in a slight rush to get panels, but i will definitely look in the different meters. Judging by the numbers I have now, 12V 6.7Amp i need 80.4W/hr. My goal is to run it for 8hours a day.

    Tell me if im doing this right.
    Im going to assume my older panels (2, 40W) are only 50%efficient ~40W/hr and that I get 4 hours of good sun every day. 40*4 = 160W/day

    Load = 80W/hr *8hr = 640W per day
    640-160 = 480W
    480/4 = 120W
    So i need a panel that will give me at least 120W as a bear minimum of running the ap. Then I'm going to want some more juice so i can also be recharging my deep cycle batteries.
  • oldmanplayingt
    oldmanplayingt Registered Users Posts: 9
    Re: Question on assembly

    Scratch what I had before,

    I now have 4 135W panels, tristar mppt cc 60amps, and still my 6 100amphour deep cycle batteries. I understand I will probably still need another panel, but I would just like to get this system started.

    Planning on connecting my panels in series and put it through my charge controller, output at 12V and connecting to my batteries which would be in parallel.

    Doing the math:
    Load: 12V*6.7Amps = 80W/hr * 24hr =1929W per day (approx 2kW)
    Solar: 135W * 4panels *5hours charge *80%efficiency = 2160 (also approx 2kW)

    Is my system making a little more sense now?

    I have been reading the forums and I'm still very confused, I still dont know what size the wire connecting my battery bank together should be. It should be gauged for my solar input right? So 135*4= 540/12 = 45 and if I only use 4 batteries, about 10amps. Is my wire guage about 12-14?
  • BB.
    BB. Super Moderators, Administrators Posts: 33,431 admin
    Re: Question on assembly
    Doing the math:
    Load: 12V*6.7Amps = 80W/hr * 24hr =1929W per day (approx 2kW)
    Solar: 135W * 4panels *5hours charge *80%efficiency = 2160 (also approx 2kW)

    Just to clarify the math (watts is rate, Watt*Hours is "amount"... Watt/Hr does not make any sense in traditional use):
    • Load: 12V*6.7Amps = 80W (no "/hr"--watts is a rate--like miles per hour, etc.)
    • 80W * 24hr =1929W*Hr per day (approx 2kWHr)
    • Solar: 135W * 4panels *5hours charge *80%efficiency = 2160 W*H (also approx 2kWHr)
    I have been reading the forums and I'm still very confused, I still dont know what size the wire connecting my battery bank together should be. It should be gauged for my solar input right? So 135*4= 540/12 = 45 and if I only use 4 batteries, about 10amps. Is my wire gauge about 12-14?

    The wiring should be sized for your greatest load and acceptable voltage drop. So, for the wire from the charge controller:
    • 45 amp continuous load * 1.25 NEC safety factor = 56.25 amps for wire/fusing/breaker
    • Round up to nearest standard breaker/wire = 60 amps
    And, you can use a voltage drop calculator of some sort to estimate the voltage drop between source and load. (i.e., 12 volt battery will draw down under load to around 11.5 volts. Most inverters shut down at 10.5 volts--So, at most, you can have a 1.0 volt for your entire battery to inverter circuit path. Many times, the wire awg required to support a minimal voltage drop is larger diameter than that required for NEC/Safety for low voltage DC circuits.

    For the Charge Controller to the Battery connection--You only want to allow a maximum of 0.1 to 0.2 volt drop for a 12 volt battery bank... That voltage drop "tricks" the charge controller into measuring a higher voltage than
    the battery actually is at.

    One thing you might want to look at--if you are using a MorningStar 60 amp MPPT charge controller--they include remote battery voltage sense leads (page 23 of PDF Manual)... Just connect two light gauge wires directly from the battery bus to the charge controller sense connections and the controller can now measure "true battery voltage" (this is a very neat thing that Morning Star has on some of its solar charge controllers--I wish more vendors did that--eliminates voltage drop error when charging a battery bank).

    Also, I would recommend the Remote Battery Temperature Sensor option too for any controller that supports RBTS.

    Wiring 4 parallel batteries... That becomes an interesting issue in itself. One battery handles 100% of the current and 4 batteries each handle 25% of the current sounds nice--but frequently does not work out.

    From my experience with DC power distribution in instruments/computers--I have used a 1/n expression to estimate current sharing in DC distribution harnesses... Say I have a wire/connection that is good for 10 amps, then the two options for 5 wires are:
    • 5x 10 amps = 50 amps shared
    • 10/1 + 10/2 + 10/3 + 10/4 = 20.8 amps shared (1/n estimate)
    I have found the second equation gave me much more reliable current sharing.

    The problem is there is very little resistance in the wiring/connectors, and so minor changes in resistance (wire length, temperature, crimp area, contact pressure/cleanliness, etc.) that, frequently, it is the "good path" (low resistance) that fails first (by overheating the connector). So, first the low resistance path goes, now the current is being carried on remaining conductors/connectors and then the next weak one fails, etc.

    So--From my experience, I would avoid sharing parallel batteries whenever possible. I would over design the wire/connectors/fuses to assume that even a 4x paralleling connection (unless ballast resistance is inserted per parallel connection)--would only reduce the average load by 1/2 instead of by 1/4...

    So, for your battery bank, I personally would wire/fuse each string to carry 100% of the load--Or, at most, a reduction to carrying 50% of the load to allow for the less than ideal current sharing.

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • oldmanplayingt
    oldmanplayingt Registered Users Posts: 9
    Re: Question on assembly

    thanks, I just want to have questions on several things.
    I really would like to keep my entire system at 12V. I do not plan on going through an inverter as my load is also a 12V system.

    You mentioned that paralleling my batteries might not be such a great idea. I'm guessing that your 1/n equation is for method 4 from http://www.smartgauge.co.uk/batt_con.html
    BB. wrote: »
    From my experience with DC power distribution in instruments/computers--I have used a 1/n expression to estimate current sharing in DC distribution harnesses... Say I have a wire/connection that is good for 10 amps, then the two options for 5 wires are:

    * 5x 10 amps = 50 amps shared
    * 10/1 + 10/2 + 10/3 + 10/4 = 20.8 amps shared (1/n estimate)

    What happens to the non shared current? Does it dissipate or it just doesnt get absorbed while charging.
    BB. wrote: »
    So--From my experience, I would avoid sharing parallel batteries whenever possible. I would over design the wire/connectors/fuses to assume that even a 4x paralleling connection (unless ballast resistance is inserted per parallel connection)--would only reduce the average load by 1/2 instead of by 1/4...

    So, for your battery bank, I personally would wire/fuse each string to carry 100% of the load--Or, at most, a reduction to carrying 50% of the load to allow for the less than ideal current sharing.

    But if I follow method 4 from the smartguage site, I should have not only my two input wires guaged for 100% of the load, but the wires between my batteries also. And fuses ( hopefully using inline fuses) go on my plus terminals on every line.

    Im not really sure how I should wire my battery bank if I do not parallel my batteries..
  • BB.
    BB. Super Moderators, Administrators Posts: 33,431 admin
    Re: Question on assembly

    You have no choice other than to parallel your battery bank. You want/need 12 VDC and you have 4x 12 volt batteries.

    The issue with paralleling is that 1) you need to have "balanced" cableing and connections to have equal resistance for all current paths--which helps sharing.

    And 2) that things happen (little bit of corrosion, dirty connection, loosening connection, aging battery, etc.) which serve to unbalance the loads.

    The energy does not "go anywhere"--it just means that one or two batteries may be carrying more than there fair share of current/energy during charging/discharging cycles and the other carrying less.

    Monitoring your battery bank (balanced specific gravity between cells when discharged and charged--if you have one battery whose cells stay "charged" when the other three are discharged--you may have a bad electrical connection) and current flow for balance will help keep your system running properly.

    For checking balanced current flow--A clamp on DC amp meter is a great tool--But a quick way to check without spending $100+ for a specialized tool is to use a DVM / DMM that can read down into the 200 mVolt range.

    For example, say each battery has an identical 12" piece of cable. With a heavy load/charge current flowing to your battery bank, set your meter to 2.000 or 0.200 volt full scale and measure the voltage drop for each cable... The cable with the higher voltage drop may be carrying more current than the others--Too much difference, and you may need to clean/check the connections.

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset