Topic: Update Large review of Finksy/J4bbrwock server psus compared to ATXpsu-photos up - page 11. (Read 20897 times)

thanks for spacer I have been told to not double post as people perceive it as advertising my signature. sometimes I want a new post to show info more clearly.


[/quote]

Happy to serve. 

Unless a mod warned you, I'd tell those people to you know what.

Are you... talking to yourself? LOL 
[/quote]

thanks for spacer I have been told to not double post as people perceive it as advertising my signature. sometimes I want a new post to show info more clearly.

Are you... talking to yourself? LOL 

No but I have one more a two hour test

and it is pretty much the same as the 1 hour test.

2880 watt psu tests

30 min

amps 9.1
volts 241
dc  12.20
dc  12.10
dc   12.00
hash 6879
watts 2193
watts per gh 0.3188

1 hour

amps  = 9.0
volts   = 241
dc      = 12.20 at the one un - used pcie cable
dc      = 12.10 at avalon 1 software reading
dc      = 12.00 at avalon 2 software reading
hash  = 6872
watts = 2169
watts per gh = 0.3156

2 hour

amps  = 9.0
volts   = 241
dc      = 12.20 at the one un - used pcie cable
dc      = 12.10 at avalon 1 software reading
dc      = 12.00 at avalon 2 software reading
hash  = 6849
watts = 2169
watts per gh = 0.3166

I would do two more sets of tests  a 4 hour and a 12+ hour

I did pul the 2980watt apart and fans are mounted differently but are delta fans use 1.68 amps each or about 5amps at 12 volts for all three = 60 watts.

they are mounted better I took a lot of photos will post them. in the morning.

have you finished your 2880 watt psu tests?

they are not quiet but the 3 pack of fans is different  I did 1 20 minute test of the 2980watt.  I am finishing the longer tests on the  2880.    I am going to try taking the fans of the 2980watt apart as they are different mounting then the 2880.


To me a guy that is home mining 3 2980 watt psu's with kilo's fan mount will be the most I am going to run.

Thanks for the early results. Looks like I was right in regards to efficiency. Are the 2980W-ers as loud as the 2880W-ers? Should require less airflow since they're running more efficiently thus less heat waste to inefficiency.

FANTASTIC! I love what you've come up with! I'll surely copy your setup when I need more PSUs and pick up some 2980W-ers. Noise was the biggest limiting factor with these IMO. This should take care of that!

I'd definitely remove the gap between each PSU. You want as much air going through the interior casing of those PSUs as possible, going through the channels is wasting cooling potential.

Will you be able to load all 3 units to 100% to see if cooling is sufficient? Might be that there isn't enough air flow and they shut down due to over temp protection at full load.

my pleasure.

more results on the  2880 watt

1 hour

amps  = 9.0
volts   = 241
dc      = 12.20 at the one un - used pcie cable
dc      = 12.10 at avalon 1 software reading
dc      = 12.00 at avalon 2 software reading
hash  = 6872
watts = 2169
watts per gh = 0.3156

I will give 4 more readings on the 2880w   the garage is cold as cold weather has arrived in NJ

I will go back and do  that same on the 2980

Some caveats the  fan packs are a bit different.
I have yet to pull the 2980watt fan pack apart.

so while I know that the 2880 watt is using 60 watts in its fan pack  I do not know that the 2980watter uses  60 watts

so while it appears to be 125 watts difference in the early tests I think the difference may be more like 90 to 100 after all tests.

And do not forget  kilo17's  3x 2880 watt  fan mod  is kick ass  as it should save  150 to 160 watts for 3 psu's

Philip,

you always find the best deal, last time I checked the lowest price was $150. Thanks for doing this.

okay my first of 3 2980watt platinum psu's arrived.

I hooked it up to my 2 avalon sixes and I am using 2000 watts to do 6800gh


My 2880 watt psu uses 2124 watts to do 6800gh


this is a very fast down and dirty comparison.    this is about 72 kwatts a month

at 10 cent power 7.20 usd a month  so if you find cheap 2980watt psu's buy them  


I have 2 psu's on order from this seller.  If they are are good as the one that arrived today these are a good price.

he sold me 2 at 80 each.

http://www.ebay.com/itm/161937378976?_trksid=p2057872.m2749.l2649&ssPageName=STRK%3AMEBIDX%3AIT

now back to rating the 2880 watt vs the 2980 watt  

2000 watts vs 2125 watts     if this is a true measurement  it is 6% different  putting the 2880 watt like a silver and the 2980 watt like a platt.

I will do more tests with a lot of photos and proof.

One issue with tests is volts provided need to match.

ie I know the draw id setup the same two avalons at freq 500 fans at 40%.

but if one psu provides 12.22  and one psu provides 12.19   my results would be off.


BTW the seller has 8 of the 2980 watt plats left.

longer test below

30 min

2880 watt

amps 9.1
volts 241
dc  12.20
dc  12.10
dc   12.00
hash 6879
watts 2193
watts per gh 0.3188

EDIT:  Never mind... Thanks for the education, VirosaGITS.  Much appreciated!!!  

I realize now the daughter board and the BB board combined are what we call the "controller" for the ant miners.  Bitmain is only selling the BB boards and not the daughter boards.  I might buy a couple just in case any of mine for the S7's go bad.
Any idea of where I can get a Daughter Board for the ASIC's and Fans to plug into?

The BB board is like a RasPi board, but you need the daughter board, the other board you plug into the BB that has all the ASIC and fan connectors, to actually use the miner. I'm not sure that ordering the BB there is going to include a daughter board hat you would need.

But i guess that if its the BB failing, you can re use the daughter board. I could not tell, however.

What is a BB board?  Is that a control board?  

Bitmain is selling BB boards, 18 pin (120mm) signal cables, fans, etc...

BB Boards:  https://www.bitmaintech.com/productDetail.htm?pid=00020160105102242316JqapyMif0705

18 pin signal cable:  https://www.bitmaintech.com/productDetail.htm?pid=00020160105105043457FPP8X92I0616

120mm fan:  https://www.bitmaintech.com/productDetail.htm?pid=00020160105104607947aoAkWsJK06D0

You can tell I'm doing FAN and Exhaust testing from that picture with so many fans eh?  Haha.  In my garage, I'm very limited to options as I don't really want to cut out the walls as the exterior is Brick.  I'm blowing out the Garage door thru the Windeevents, hence the small 600 CFM Centrifugal fan.  That works well but it's too little power to handle a 24x20x10ft Garage which is 7200CFM.  Assuming that the manufacture even does half their rating which makes it 300CFM, that's 25mins per full air exchange.  This is way too long of a time and heat would just build up too much during that time.  This is only ok when the outside temperature is 70 or below but definitely not when temps are in the 70+.  The bigger 7000 CFM which as you stated, are no where close to the manufactured stated output.  Let's assume it's 3500 CFM, is powerful enough to quickly push hot air out on hotter days where I would just raise the garage door up 6-12".  It does the trick but I'm working on running Exhaust thru the attic, it's the least efficient way but it would offer the cleanest looking as my wife does not like keeping the garage door and my garage is street facing so I really shouldn't be adding more than 4 Windeevents where HOA will start knocking on my door.   Those Tube Axial fans are very cool but I would probably only consider them once I start moving out to a warehouse, which may be soon if my Shed / Garage solution draws too much unwanted attention.

P.S I ordered both the big Fan that Kilo used as well as more Ultra Kaze fans to test a 4 PSU stacked configuration.  I think with the new 4K board, I can taped the 4 Ultra kaze fan on or use a  FAN mount to the side and it will plenty of CFM while offering a very clean / sexy look.  But it still doesn't hurt to try that 1 big fan method too as its' cheaper at $40 (170CFM) vs $68 for the 4 Ultra Kaze (130CFM x4 = 520).   These will arrive before Finksy board so I will test them next Monday. 

your photo

I would think you just simply connect it to the 3Pin FAN Controller on Finksy / J4bberwock board since you have that board.  The b/o board made for the 2880 / 2980 PSU does not have a 3 or 4 pin fan jack.

As for CFM on exhaust fans, you cannot always trust the CFM declarations made by fan manufacturers.  Especially, if the fan is not inside of a tube axial type enclosure.  I've found the longer (deeper) the tube axial is, the better CFM will be with the same size motor.  

I've had a little bit of experience over the years with exhaust fans for wood/furniture shops when staining, painting, etc...  My father always had a wood/furniture shop built on his property where ever they bought a house.  I also worked in a furniture factory in Berlin, New Hampshire when I got out of the Marine Corp in 1988.  My father and the owner of the furniture factory both said the tube axial design is the way to go if one can afford it and has a proper place to mount it.  The tube axial design has more suction power than one that has a rinky dink box or cage built around it.

Here is a link for example of Tube Axial Fans with the cheapest price.  You will not find cheaper prices on the internet for fans of this quality that are new:  http://www.ebay.com/sch/i.html?_from=R40&_trksid=p2047675.m570.l1313.TR0.TRC0.A0.H0.XTUBE+AXIAL+DUCT+FAN.TRS1&_nkw=TUBE+AXIAL+DUCT+FAN&_sacat=0

Also, if you can afford or find an affordable tube axial fan, it is best to get a belt driven tube axial than a direct drive.  If something should ever go wrong with the motor of a direct drive model, you will have to disconnect the blade from the direct drive motor.  THAT can be a pain in the ass!  If you have an issue with a belt driven motor, you can easily take the belt off the wheel of the motor and do maintenance or replacement much easier without having the frustration of removing a fan blade.

NOW THIS IS A GOOD IDEA!  Triple or QUAD 2000PSU with 1 FAN.  Muahhahahah.  DO IT!!  I'm building with many 2K PSU.  If I can stack more than 2 in the same space since I have plenty of height space, it would be a huge space saver!  I'm going to order 1 right now for fun and compare it against 4 x 2KW PSU stackout on top of each other with 4x Ultra KAZE Fans blowing from the side.

See my RACK.  I can easily stack 3-4 x 2KW PSU with this FAN Mod.  It would also be much more efficient cable management wise with the new 4K Board then my single boards which have to be spread side ways and waste too much space.  I put 2 x 2KW PSU combo opposite as they spread too wide sideways so the other other pair of PSU using single boards will be running PCIE from the back to maximize space.  With the 4K boards I have coming next week, I won't have this issue and I can stack 4 PSU on top of each other with no issues and cabling will be much easier and neater.

http://imgur.com/Wntew7g

Please quote so the picture will show.  Please don't mind the 3 fans.  I'm doing Heat stress test and exhaust test with 3000CFM, 7000 CFM Axial fan and 600 CFM Centrifugal.

@Kilo17 - How do I cut /connect the cable or connect for the 182 version that is 3PIN with 1 manual controller for FAN speed?  Or can I just connect this to the 3Pin FAN Controller on Finksy / J4bberwock board?

It is a 2 wire out so I simply spliced it into a PCIe and plugged in it.