*Official* Norco data storage products thread

EnderW said:
anyone using the RPC-450 or RPC-470 as their main desktop chassis? I'm getting a rack and would like to put everything in it, so considering moving my main computer from my trusty P182 to a 4U RPC-450 or RPC-470

would have to get a new CPU cooler though and I just spent $90 on a HR-02 :(

also concerned about noise, it doesn't look like there is a 120mm fan bracket for the RPC-470 like the 4220 and 4224
Click to expand...

i guess you can just use a 120 fan wall , the chassis 470 , 4220 , 4224 are just about the same , it might be nessasairy to just dril a couple holes if they are not alined but i think this is no havy job.
Also its easy to modify the wall itself , cutting extra holes if you need to.
I using now 120mm Scythe fans are nice quite much much better than 4x 80 mm yetengines :D
 
I just got a Norco 4224 case that I'll be using to combine my NAS and main desktop PC into a single computer. I'm puzzled why everyone says that these cases are loud with the stock fans. I couldn't even hear the Norco running with the stock 80mm fans over my other two computers. I just got the 120mm fan wall and the 120mmx38mm fans that I'll be installing will be much louder than the stock fans. Of course those fans also have to do double duty by also pushing air through a radiator for my water cooling setup so they have to move more air through more restrictions than the stock fans.
 
my Norco has origional Delta fans 3000 RPM 31dB , the AFB0812H they really loud so i think you maybe have other fans ?
 
Hello everyone,
I have a norco 4220. I currently have a standard 850w power supply installed.

My norco's mission is about to significantly change.

Can anyone point me to redundant power supplies (650+), that come in the "standard" W x H power supply form factor? The depth doesn't matter. But I would like to be able to simply bolt it into the existing power supply area.

Thanks.
 
simplesam said:
Hello everyone,
I have a norco 4220. I currently have a standard 850w power supply installed.

My norco's mission is about to significantly change.

Can anyone point me to redundant power supplies (650+), that come in the "standard" W x H power supply form factor? The depth doesn't matter. But I would like to be able to simply bolt it into the existing power supply area.

Thanks.
Click to expand...
hmm i dont know if you get things clear but redunant PSU are 2 or more units
but they are smaller , they would fit into a standard ATX space but need another bracket to fix them to the case.
Than you can use from both PSU`s molex 4 pin cabels to the backplanes and than they are powered double so when one fails it still get power from the other PSU
If using only one PSU there`s no need to fill both 4-pin Molex connectors on the backplanes

edit : you would need something like this ; http://cybershop.ri-vier.nl/seasonic-oem-520w-2u-eps-ss520h-2u-80plus-white-modular-p-53.html
2 of them fit into 1 ATX space
 
Last edited:
simplesam said:
Hello everyone,
I have a norco 4220. I currently have a standard 850w power supply installed.

My norco's mission is about to significantly change.

Can anyone point me to redundant power supplies (650+), that come in the "standard" W x H power supply form factor? The depth doesn't matter. But I would like to be able to simply bolt it into the existing power supply area.

Thanks.
Click to expand...

Norco has a line of redundant power supplies in multiple wattages. However, they aren't cheap:
http://www.ipcdirect.net/servlet/Categories?category=Power+Supply&s=9

When I asked around a while back, the general consensus was instead of purchasing these, it was a better move to start with a Supermicro 24 bay chassis. Since you already have a built Norco, your circumstances are a bit different. If you do decide to try the Norco/IPC power supply, please report back with your impressions.
 
axan said:
ditch your norco or keep it for another project and pickup supermicro CSE-846TQ-R900B http://www.supermicro.com/products/chassis/4U/846/SC846TQ-R900.cfm
Redundant 900W psu + a lot better built quality + it comes with sliding rails all for ~$900
Click to expand...

With the price of the power supplies, I was thinking the same thing.

I was throwing this together for a friend's business, for storing backups, but with the power to run a few VM's. Now the scope has changed to "I think we'll use it to replace our server" and "we must have redundant power"...

Anybody in metro detroit want a 4 day old norco 4220? PM me if so.
 
The four fans that came on the fan wall were model AFB0812M. The two on the back of the unit are model GNF802512HS.
Parcifal: You mentioned that these fans are only 31db. Maybe I'm just used to louder fans in my bedroom as I have several mid 40db fans in each of my cases. I have two 56db fans in one case but I normally run them in series to keep the noise down and only put them on full blast when doing extended stress testing. Even running in series they are louder than the AFB0812M fans that came with my Norco case. In fact, I plan on running both in my Norco plus some additional over 40db fans.
I found that with the 120mm fan wall, I couldn't put three 120mm X 38mm fans on the HDD side of the fan wall like I was planning to. The sas cables can't make a 90 degree bend fast enough. I found that I can get away with a normal 25mm think 120mm fan. Since this fan will also be forcing air through a radiator it was hard to find a fan with enough CFM and static pressure under 45db.

Since I don't think I've bought a fan that is less than 35db, I guess these 31db fans seem quiet to me.
 
simplesam said:
With the price of the power supplies, I was thinking the same thing.

I was throwing this together for a friend's business, for storing backups, but with the power to run a few VM's. Now the scope has changed to "I think we'll use it to replace our server" and "we must have redundant power"...
Click to expand...

Note that the TQ model has 24 sata ports, probably not what you want. The A model is most similar to the Norco 4224 in that it has 6 SFF8087 ports. It seems that the A model is only available with a 1200W Gold redundant PSU, but that is a nice PSU anyway. Or you could get a model with a built-in expander, but that opens a whole new can of worms, and I am already far enough off topic for this thread.
 
Indeed the Norco fans should be far over 31db i read wrong somewere i guess
but for redunant powersupplys be prepair they using 40mm yetengines they most worse !
i have had an 1U chassis that have 40mm Delta fans ... this makes it not usable into home areas.
 
grogthegreat said:
The four fans that came on the fan wall were model AFB0812M. The two on the back of the unit are model GNF802512HS.
Parcifal: You mentioned that these fans are only 31db. Maybe I'm just used to louder fans in my bedroom as I have several mid 40db fans in each of my cases. I have two 56db fans in one case but I normally run them in series to keep the noise down and only put them on full blast when doing extended stress testing. Even running in series they are louder than the AFB0812M fans that came with my Norco case. In fact, I plan on running both in my Norco plus some additional over 40db fans.
I found that with the 120mm fan wall, I couldn't put three 120mm X 38mm fans on the HDD side of the fan wall like I was planning to. The sas cables can't make a 90 degree bend fast enough. I found that I can get away with a normal 25mm think 120mm fan. Since this fan will also be forcing air through a radiator it was hard to find a fan with enough CFM and static pressure under 45db.

Since I don't think I've bought a fan that is less than 35db, I guess these 31db fans seem quiet to me.
Click to expand...
I know the SAS cables are hard to bent i have one its nearly too short to reach the controller , i cut of shrink for the first 10 cm near the connectors and on that way they are much more flexible , with some care it wont hurt the cable.
 
Update; i put the leds into the front , very easy to take off the front panel thats make the work easyer.
pic.php
 
Seems Norco use some different types for the 80mm fans on fanwal
for peeps who interested in the specs ;
pic.php
 
Hey treadstone,

treadstone said:
@DarthRay: I'll give you an explanation/function description of the J1 pins tomorrow. I just returned from a one week vacation and I'm rather tired right now. I can tell you that the I2C connections from the white 3 pin header only go to the non populated sideband controller chip on the backplane. So connecting anything to it will do absolutely nothing for you!
Click to expand...

Have you had a chance to look at the J1 pins? I'm really curious about the possibility of driving the error LEDs. Have you been able to do it?

Thanks!
 
@DarthRay: Sorry for the delay, I am rather busy...

I have a schematic diagram I put together a while back, but it's rather messy and not sure how much help that would be to you. I can give you a brief explanation on how Norco implemented the error LED:

This is for a single error LED (D8 in this example):

J1/8 (J1 pin 8) is labeled as the positive (+) input terminal and connected to pin 9 of the HC14 (schmitt trigger hex inverter input) along with a 4k7 pull down resistor (R84). Pin 8 of the HC14 (schmitt trigger hex inverter output) is connected to a 470R current limit resistor (R8) that is in turn connected to the error LEDs cathode (D8/R). The error LEDs anode is directly connected to the +3.3V supply (provided by U6 in the case of D8). Also, the HC14s output (pin 8) is connected to J1/7 and labeled as the negative terminal (-). The HC14s output is also connected to U7/21 (the SGPIO sideband backplane controller chip that is NOT populated). I assume that if U7 was populated, the HC14 (forgot what the reference designator for this chip is) and J1 would not be populated as they are not required.

To control the error LED:

Connect the HBAs error LED output to J1 pin 8 to control the backplanes error LED. If the signal is driven high, the inverter will pull it's output low and turn the LED on. If the signal is either driven low or left floating, the inverter will drive it's output high and turn the LED off.

I assume that the idea for the positive (+) and negative (-) connections on J1 is so that you could use J1s positive (+) input if the HBA has an active high error LED status output or use J1s negative (-) input to drive the error LEDs direct if the HBA has an active low error LED output. However, since the HC14 is not an open collector/drain output type inverter, you would essentially be pulling the output stage of the HC14 low via the HBAs error LED output if you try to drive the error LED direct via the negative (-) input! This is NOT recommended as this would cause excessive currents to flow through both the HC14s and your HBAs output stages!

If your HBA has active low error LED outputs, I would suggest you remove the HC14 from the backplane and drive the error LEDs direct via the negative (-) marked pins on J1. Otherwise, just use the positive (+) marked pins of J1 and you should be all set...

Hope this helps.
 
@treadstone: thank you very much for the detailed description! Even though that explanation went above my head a bit, I believe I got what I needed to make a decision. I knew you were busy, sorry if I seemed impatient there. ;)

Thanks again!
 
Many tnx Threadstone for the good work !

i have been busy with it and i could light up the red led by put the pin 8 via 4k7 resistor to + 5 volt
but i couldnt get the negative input work
my controller seems to have negative output as my homemade leds for the data activity are working well as they lead via resistors to + 5 volt.

so it could be hard to get it work too for the errors leds without modify the backplanes ,
properly the best way is what you mention disconnect the backplanes led from the HC14 and than lead it to the controller.
But .. i dont know if error leds are already working via the SAS cables ?
as my SAS cables are wired up with 2 x 4 wires flatcable already.
Is there a easy way to check out the error leds if they do work via the controller by SAS ?
or wait for a drive is fail and see what happen ...

It would be not nessasairy to modify the backplanes and wire up things to the controller if the error leds option is already be able to work.
 
@Parcifal: my impression is that our options to drive the LEDs are:

1. side-band via the SFF-8087 cables (SGPIO, I believe)
2. I2C
3. individual LED headers

I believe #1 and #2 are out given that the backplanes are not populated with the right components. So that leaves us with the individual LED headers.

Do you know of any other way to drive the error LEDs?
 
DarthRay said:
@Parcifal: my impression is that our options to drive the LEDs are:

1. side-band via the SFF-8087 cables (SGPIO, I believe)
2. I2C
3. individual LED headers

I believe #1 and #2 are out given that the backplanes are not populated with the right components. So that leaves us with the individual LED headers.

Do you know of any other way to drive the error LEDs?
Click to expand...

indeed option 1 wont work with the backplanes it seems even while the SAS cables has the wiring for it .
option 2 lost as well
so left is 3 , and try to use the led headers on the controller for it
i really think that many controllers work with negative output , so we need to change
the backplanes, there might be another complication than as the blue en red led is a dual color one , need to be very carefully not mixing voltages and blow up something .

when the bleu and red are light together it wont be fine , the red one really needs to get light when there is error and not reverse .
I dont know yet for a solution, not for the backplanes anyway .
Left is a try to use extern leds , homemade as i did for the activity leds.
maybe just use 1 simple error led for all controller outputs , driven by diodes .... dont know if thats posible ,
need solution that if one of the error outputs is given negative power ( if case there is an error) that one single led connected to all outputs is glowing, if its output 1 or 16 ...
That would be fine too, you know than there is an error on some disk and can find it in the managment which disk is in fail.

Or... using optocouplers to isolate the controller outputs but than we can drive leds on the optocoupler`s output circuit to positive or an negative source
but this option is hard to make , it needs lot of soldering and that PCB would be not really nice small.
I havnt seen yet 16-way optocouplers in 1 single chip but that would save some work.
 
Last edited:
Is there such a thing as an inexpensive circuitry to convert form negative output to positive?
 
@DarthRay: You are most welcome.

@Parcifal: If your controller only has a negative active output (pulls the output to ground when active), you should remove the HC14 from the backplane and simply use the negative (-) marked pins to control the red error LEDs. There is no conflict between the red and blue LEDs other than that the color will not be read or blue but rather a mixture of them (not sure what that color is since it also depends on the intensity of each).

Due to the fact that the blue LED has a Vf of around 3V and IIRC, the supply is at 3.3V and the transistor that Norco uses to drive the activity LED has a VCE(sat) of around 0.3V, that 470R current limit resistor is way to big, hence the low light output of the blue LED. They should have either selected a smaller current limit resistor or used a higher supply voltage for the LED (e.g. used the +5.0V instead of the +3.3V)!

I am going to replace the current limit resistors on my backplanes since it is next to impossible to see any of the activity LEDs when the room is bright enough. The only time I can really see them is when the room is dark...

Since the red LED has a lower Vf, I think they should be ok (should be bright enough), but if necessary you could change the current limit resistors for the red LEDs to make them brighter too.

Since I don't have any specs on the LEDs that Norco used on these backplanes, I wouldn't go beyond 5mA for each LED to stay on the safe side...
 
indeed so such a thing as the HC 14 should be posible to add on a pcb
for me its maybe best way to cut the circuit between the red led cathode and the 470R resistor , and take the led pin with wire to my controller via a resistor , if the red led is powered by + 3,3 volt on the anode on the backplane.
Than it should just working like the other mod i did for the activity leds they leading also to + of the PSU , wouldnt be much different if i take the 3,3 or 5 volt i think , but the + 5 volt is just on so much molex connectors that its the most easyest way.

Treadstone; if you have some free time and posibility to draw something of the circuit it might be much more clearly , i think i got it right but not really so sure lol
would be perfect also for new peeps here on forum.
 
Last edited:
Just wonder have you been able to fix the resistors voor de bleu activity leds on the backplanes ?
i dont know how far to go with it cant messure things on that.
Need to put out all the backplanes and thats a big job with all the cables lol .

For the error leds i maybe have another idea , and glue some red leds on top of the led housings on the backplanes , solder them to + 5 volt and other side to resistor and forward to the controller but wiring up is a big job.
 
hmm project seems on a death path, i dont have more ideas
if anyone can could do some input ....
I actually google around to find a way to blink led outputs by command
as some professional applications have the option to localize the server/ drive which is in error or needed maintenance
but it seems more regulair controllers dont have this option or it is just not added in the Gui software managment .
I would like so see a tweak / program to get this option but i guess its hard to find for windows machines.
 
Blue ACTIVITY LED Update:

I modified my backplanes to make the blue activity LED brighter. First, let me say that the Norco engineer that designed the circuitry obviously didn't design this properly. Starting with the 470 Ohm current limit resistor for the blue LED to the transistor he used to buffer the activity signal from the HDD...

Anyway, long story short, I replaced the current limit resistor and the transistor with an actual logic gate to properly drive the LED. I used parts I had on hand which in this case was a single AND gate (74LVC1G08GW) in a TSSOP5 package.

The original design had around 200uA to drive the blue LED. Which is the reason for them being so dim. I increased the current to just under 8mA (around 7.8mA to be exact). Since I don't have the specs on the LEDs, I tried to keep this low just be to on the safe side.

The LEDs are now nice and bright.

One more change I may do is to remove the capacitors the Norco designer put on the drive activity signal. It effectively stretches the length of the active period the LED is on, which when there are really short burst on the drive appear as a lot longer than they really are. It looks like the drive is accessed continuously when in reality the drive is being accessed in a lot of short bursts!. So I may remove those capacitors to have the LEDs on for as long as the actual drive turns them on!


@Parcifal: The identify a drive or identify an enclosure command you are referring to should be part of your HBAs firmware! My Areca RAID controller has it and it works quite well on my Chenbro enclosure (through the two HP SAS expander cards). The RAID controller treats each SAS expander as an individual enclosure, so even though both are in the same chassis, when I select to identify enclosure #1, the first 24 HDD in the chassis (upper half of the chassis) are identified by blinking the red error LEDs and when I select to identify enclosure #2, the second 24 HDD in the chassis (lower half of the chassis) are identified by blinking the red error LEDs. When I select to identify an individual drive, it blinks the appropriate red error LED for that particular drive. The chenbro enclosure uses a blue LED to indicate that a drive is present and powered, a green LED to indicate activity and a red LED for the error status. The green and red LEDs are on the same light pipe. In comparison, the Norco 4224 has the blue LED as activity and green as power.
 
@Trepidati0n: Sorry, I may have worded that wrong. Driving an LED without a current limit resistor would be a bad thing, unless the LED has a build in current limit resistor (only a handful do).

I used a 49R9 0603 resistor to limit the current.
 
You`ve done hard work on it !
well i will see if i can get some parts like transistors and resistors

seems like 49,9 ohm type http://nl.farnell.com/panasonic/era3aeb49r9v/resistor-49r9-0603-0-1-0-1w/dp/1809605

and the gate : http://nl.rs-online.com/web/search/...ts&searchTerm=74LVC1G08GW&searchTermKey=|PREV
Thanks for the tip for the controller , i dont think i can change that myself into the firmware
and it needed to change the managment too with that option.
Hopefully Highpoint does fix that in next release
 
Last edited:
@Parcifal: 0.1% is total overkill for the 49R9 resistor! Even the 1% resistors I used are not necessary.

I'll post some pictures when I find the time. I also had to add a 100k 0603 resistor as a pull-up on the input of the gate since the 10k pull-up resistor on the board is connected to the +5V supply that get''s turned off when the drive is removed. This caused the activity LED to turn on, which wasn't something I wanted, so the 100k pull-up resistor keeps the blue LED off if the drive is not present or the signal is left floating.

If you are buying parts to do this modification, I would suggest you actually buy a buffer. I only used the AND gate since that's what I had handy and the particular gate I picked is able to drive up to 32mA (some of those low power single gates have a limited current drive capability). Also, since the MMBT3906 transistor that's on the board is a SOT23 package and the gate I used is a TSSOP5, they don't exactly fit and the gate requires a supply voltage which isn't present on the existing footprint, so I had to wire that up to the gate. I used the +3.3V supply of the Activity/Error LED which is available close to the location of the transistor. Please also keep in mind that these parts are fairly small and unless you have proper tools, this modification is going to be difficult to do.
 
treadstone said:
@Parcifal: 0.1% is total overkill for the 49R9 resistor! Even the 1% resistors I used are not necessary.

I'll post some pictures when I find the time. I also had to add a 100k 0603 resistor as a pull-up on the input of the gate since the 10k pull-up resistor on the board is connected to the +5V supply that get''s turned off when the drive is removed. This caused the activity LED to turn on, which wasn't something I wanted, so the 100k pull-up resistor keeps the blue LED off if the drive is not present or the signal is left floating.

If you are buying parts to do this modification, I would suggest you actually buy a buffer. I only used the AND gate since that's what I had handy and the particular gate I picked is able to drive up to 32mA (some of those low power single gates have a limited current drive capability). Also, since the MMBT3906 transistor that's on the board is a SOT23 package and the gate I used is a TSSOP5, they don't exactly fit and the gate requires a supply voltage which isn't present on the existing footprint, so I had to wire that up to the gate. I used the +3.3V supply of the Activity/Error LED which is available close to the location of the transistor. Please also keep in mind that these parts are fairly small and unless you have proper tools, this modification is going to be difficult to do.
Click to expand...

Its indeed a very difficult modification , not only one part to change but several more and that on all backplanes... it really need the right tools , i soldering over 25 years now but i haven`t done much with SMD parts , i did modificate some cpu`s that was a difficult job to and lots of risk with a very thin copperwire into a quite normal solderiron lol.
I would like to see some pictures from your mods, than we can get a better idea how its done and if we should do this too lol , i guess you have much more experiance.
If it were "normal" sized parts i was started already tonight but SMD is maybe a bit too far for me.
 
Hi Guys,

Got my Norco 4020 today and looking at the connections area for the HDs, I noticed it's all sata plugs and no sas connections. I also purchase the LSI 9260-4i and Intel RES2SV240. I know the intel comes with a bunch of SAS cables not they're all sas=>sas, not breakout cables like I'll be needing now.

Is the the correct cable I'll need? http://www.newegg.ca/Product/Product.aspx?Item=N82E16816115064 Going to need eventually 5 of them to get all the connections setup.

Or am I looking for a 'reverse-breakout' cable ?

TIA
 
A suggestion to norco for the 4216 product, please make the 5.25 bays use a standard cutout, don't use the silly swing out doors like some 10 year old HP consumer pc. Not everyone uses those for DVD drives, trying to use a 4-in-1 2.5" drive cage or anything else mean hacking the thing up.
 
You must log in or register to reply here.
Back
Top