Results 1 to 18 of 18

Thread: SeriTek/2eEN4 External 4 Bay Enclosure

  1. #1
    Join Date
    Oct 2005
    Location
    Toronto
    Posts
    3

    Default SeriTek/2eEN4 External 4 Bay Enclosure

    Hi
    I am moving my data over to a Hot-Swap SATA drive system and MacGurus has been terrific in improving my knowledge. Currently I back-up only to Firewire drives with no real system in place - that is all going to change. I have narrowed my search between the Granite Digital 4 bay, the Burly and this latest 4 bay box from FirmTek with its "direct-to-backplane" design. Currently, I am running on a 867 G4 but intend to finally upgrade to a new G5 in the next few days or whenever the new Macs arrive. I am a photographer so I work on large files and I fill up HD's quite quickly.

    I am leaning towards the FirmTek enclosure bundle including the SeriTek/1eVE4 host card looks appealing. However, (and I realize these are very small factors) I would prefer a LCD display plus FirmTek doesn't offer any transport cases for the trays. I suppose I could use the padded carrying case sold here or at Granite Digital?

    I am sure all three enclosures will do a good job. I would be interested in hearing what others think about this new FirmTek SeriTek/2eEN4 enclosure compared to the other two box enclosures mentioned.

    Thanks in advance.
    Clive

  2. #2
    Join Date
    Nov 2004
    Location
    Germany
    Posts
    2,352

    Default

    Hello cbarber,

    welcome at MacGurus Tech Forums!

    I would prefer the burly cuz I think the airflow in this case is better and it is using a larger fan wich don't needs to run at high speeds like little fans.

    Regards

    Nicolas
    Custom Configurations! Rad Hacks and Mods!

  3. #3
    Join Date
    Aug 2001
    Location
    Grangeville, ID USA
    Posts
    9,124

    Default

    We won't sell the Firmtek case. It has a 150 WATT power supply, minimal for 4 of todays drives and I believe we will see them failing pretty often.

    The case is small and tight and uses a high speed fan that makes a lot of noise. I have not tested drive temperatures in them because of the other issues, but I would bet that they are not very efficient at keeping the temperatures stable.

    The hotswap trays are skeleton trays in the Firmtek box and they do nothing to protect the drives. We use the best quality aluminum full protection trays we can find. they will keep temperatures at a stable low and protect the drive even carried loose in a briefcase. And yes, the Granite soft cases will hold them.

    All in all the trend towards small 'pretty' enclosures bodes ill for the life of hard drives. The bigger the drive, the power it uses and the ,more heat they generate. Our 4 Bay Burly has 300 WATT power supply that will outlive the heatdeath of the universe, steel case you can stand on, and 2 big 80mm fans turning at 2700 RPM for a lot of air without much noise.

    Best to you,

    Rick
    molṑn labe'
    "I am a mortal enemy to arbitrary government and unlimited power. I am naturally very jealous for the rights and liberties of my country, and the least encroachment of those invaluable privileges is apt to make my blood boil."
--Ben Franklin

  4. #4
    Join Date
    Oct 2005
    Location
    Toronto
    Posts
    3

    Default

    Nicolas and Rick, thank you for responding.

    This is the kind of information I am looking for. I was aware the FirmTek trays did not cover the drive completely and was concerned about the protection and portability. It was this review of the FirmTek enclosure that swung my opinion.

    The Burly is the box I have been coming back to again and again in my search. I want a card that is bootable and is eSATA. I do not believe the Sonnet Tempo is bootable yet, is this correct? Do you sell the SeriTek/1eVE host card? Do you consider this to be a good card?

    Thanks again.
    Clive
    http://barefeats.com/hard58.html

  5. #5
    Join Date
    Aug 2001
    Location
    Grangeville, ID USA
    Posts
    9,124

    Default

    The Firmtek cards are great. They are compatible with just about everything. George, who writes the firmware, also wrote all of Apple ATA firmware in days past. He is a bit crazy, but like a fox. The guy is a genius and can make a host card sit up and beg. I bow in his direction as the master of the trade. You can use his cards with the full confidence that they will do the job. Here's the link for them on our site: Seritek/1VE4

    Here's all the SATA Host cards

    Rick
    molṑn labe'
    "I am a mortal enemy to arbitrary government and unlimited power. I am naturally very jealous for the rights and liberties of my country, and the least encroachment of those invaluable privileges is apt to make my blood boil."
--Ben Franklin

  6. #6
    Join Date
    Oct 2005
    Location
    Toronto
    Posts
    3

    Default

    Sorry for the error in my last post regarding the link to a review. I'll have to figure out what I did wrong.

    I saw the three 4 port Seritek host cards on your site but didn't see the SeriTek/1eVE card. Am I missing something?

    Thanks.
    Clive

  7. #7
    Join Date
    Aug 2001
    Location
    Grangeville, ID USA
    Posts
    9,124

    Default

    Ah, I thought it a typo. I think that Firmtek is keeping those to sell their new enclosure. I had heard that near the end of the month we were scheduled to receive an eSATA connector version of their 4 port card. I didn't know that its part number was 1eVE. I mislead you.

    No difference in performance with the TypeA cards at least. But we won't have the eSATA until later this month.

    Rick
    molṑn labe'
    "I am a mortal enemy to arbitrary government and unlimited power. I am naturally very jealous for the rights and liberties of my country, and the least encroachment of those invaluable privileges is apt to make my blood boil."
--Ben Franklin

  8. #8
    Join Date
    Feb 2003
    Location
    ex-Cupertino, CA, USA. Now HU+SRB (Europe).
    Posts
    163

    Default

    Quote Originally Posted by ricks
    We won't sell the Firmtek case. It has a 150 WATT power supply, minimal for 4 of todays drives and I believe we will see them failing pretty often.
    With all due respect, I have to disagree. FirmTek 4-bay enclosure power supply that rated at 150 Watts continous output power is more than powerful enough for 4 high performance SATA disk drives.

    For example, the average power consumed by a Hitachi 500 GB disk drive are listed below:

    Randow R/W (avg power): Single drive 13.6 Watts : 4 drives 54.4 Watts
    Idle (avg power) : Single drive 9.6 Watts : 4 drives 38.4 Watts
    Unload Idle: Single drive 7.5 Watts : 4 drives 30 Watts
    Low RPM Idle: Single drive 5.1 Watts : 4 drives 20.4 Watts

    A single Hitachi 500 GB drive when consumes 29.95 watts maximum during startup/spinup; When 4 are powered on at the same time they consume 119.8 Watts for only several seconds.

    The power supply also has excellent transient response & can provide energy to easily meet the peak power requirements of 4 drives performing random R/W.

    Quote Originally Posted by ricks
    The case is small and tight and uses a high speed fan that makes a lot of noise. I have not tested drive temperatures in them because of the other issues, but I would bet that they are not very efficient at keeping the temperatures stable..
    Rick, you would lose the bet! Would you like to try? Independent reviews of SeriTek/2eEN4 may be found at: http://www.barefeats.com/hard58.html In this review the author states:
    "We measured the decibels at 6 inches to be in the low 40s. The G5 Power Mac was louder even in "resting" mode. Similar enclosures have been measured as high as 65 decibels."

    Quote Originally Posted by ricks
    The hotswap trays are skeleton trays in the Firmtek box and they do nothing to protect the drives. We use the best quality aluminum full protection trays we can find.
    Rick, with all due respect you did confuse FirmTek trays with trays found in a product sold by an other company. I did see quite a few "skeleton" trays, but not a single one sold by FirmTek. BTW: FirmTek drive trays are not aluminium but stainless steel and were adopted from enterprise class storage arrays. These trays also help protect against static electricity when removing & inserting a drive into the enclosure; they are designed to conduct static from your body to the frame ground of the enclosure.

    Last but not least: the SATA signal integrity is a very difficult issue to achieve. There is a good reason, why SeriTek/2eEN4 is a "latecomer" - it is a product of substantial work. At this moment FirmTek's box is one of the very few which does have a passive backplane and did pass the SATA signal integrity test.

  9. #9
    Join Date
    Aug 2001
    Location
    Grangeville, ID USA
    Posts
    9,124

    Default

    OK, so I'll back off on that. I disliked the 2 bay enough that I let myself get disagreeable on the 4 bay. I still believe that 150 is too small of a power supply. I dislike open trays. I apologize for missing the other details. Won't happen again.

    Rick
    molṑn labe'
    "I am a mortal enemy to arbitrary government and unlimited power. I am naturally very jealous for the rights and liberties of my country, and the least encroachment of those invaluable privileges is apt to make my blood boil."
--Ben Franklin

  10. #10
    Join Date
    Feb 2003
    Location
    ex-Cupertino, CA, USA. Now HU+SRB (Europe).
    Posts
    163

    Wink

    Quote Originally Posted by ricks
    OK, so I'll back off on that. I disliked the 2 bay enough that I let myself get disagreeable on the 4 bay. I still believe that 150 is too small of a power supply. I dislike open trays. I apologize for missing the other details. Won't happen again.

    Rick
    Rick the user should care less, is that 150 W enough for drives or not. For the beginners - there is only one moment (in theory) which is of concern, namely the moment of the drive spin-up. An enclosure featuring a mere 70 W (!) power supply is, however, enough. There is a nice thing called "staggred spin-up". All modern SATA drives do feature it. In other words, a drive spins up only at the first moment it is REALLY accessed, not at the moment you turn the device on or insert the drive. This can be pretty much a life-saving thing for the subsystem... IF and it's a very big "IF" the driver can guarantee, it never will spin-up the drives at the same time. Unfortunately EVERY driver but ONE particular does can't do that by design. Apple's "open source" driver cannot guarantee it and every ATA driver which derives from it suffers from the same fate. Someone has to have balls to design and introduce a true ATA driver which is on one side ATA (not a SCSI emulation, so some feaures are not lost) on the other side it's not Apple's ATA. Now, there is one (to be precise: one driver family) which works and quite a few people are using it on MacOS-X since good old UltraTek/66.

    The 150 W power supply was made with those unfortunate souls in mind who use a non-Firmek card.

    The most important fature of the box by far is it's compliance with the specs of www.sata-io.org. It's - as it is now REQUIRED! - a cableless design with a backplane. In the good old days of parallel SCSI and relatively low-speed FireWire people could assemble a box from off-self parts. Even the first generaton SATA-150 deals about TWICE fast frequency than the fastest FireWire. The second generation has 4x frequency compared with the fastest FireWire. It's important to understand, that how important is the frequency. On SATA, regardless, what is drive's effective speed the signal on the cable travels with the same 1500 Gbps or 3000 Gbps. The 3000 Gbps is around the corner to entirely replace the 1500 Gbps. If both the controller and the drive do negotiate the link being at 3000 Gbps, all boxes with internal cabling are in trouble. Even for 1500 Gbps I see the boxes with internal cabling fail more often than I wish.

  11. #11
    Join Date
    Feb 2003
    Location
    ex-Cupertino, CA, USA. Now HU+SRB (Europe).
    Posts
    163

    Talking

    Quote Originally Posted by ataMan
    Rick, with all due respect you did confuse FirmTek trays with trays found in a product sold by an other company. I did see quite a few "skeleton" trays, but not a single one sold by FirmTek. BTW: FirmTek drive trays are not aluminium but stainless steel and were adopted from enterprise class storage arrays. These trays also help protect against static electricity when removing & inserting a drive into the enclosure; they are designed to conduct static from your body to the frame ground of the enclosure.
    Forgot to mention. I cannot force you to love FirmTek trays - but than again, if you don't, please take a look at XServe trays. You should not love Apple XServe trays either!

  12. #12
    Join Date
    Feb 2001
    Location
    on the landline, Mr. Smith
    Posts
    7,787

    Default

    Interesting stuff.

    ATAman: So....one cable is OK to attach the external array to the host, but two (the second being internal in the array case as you mention) is too much for optimal performance and reliablility? Is the issue the second cable, the overall length of all the SATA cabling, too many plugs/connections, or a combination of all of the above?

    Just curious.

    Yes, seems many arrays have skeleton trays these days, which I only see as a problem for those who need to pull and transport drives. All of ours at work are so we can swap a drive on the fly, without pulling a machine out of service, out of the rack, and cracking it open. "Good" or "bad" depends on the users needs I suppose. I always assumed the open trays would allow heat to move away from the drives better. I could see many SOHO users wanting to backup to drives and store the whole tray, which I think is Rick's primary concern - protection in transit....right Rick?

    Any thoughts about PCI Express on the new G5s?
    "Imagine if every Thursday your shoes exploded if you tied them the usual way. This happens to us all the time with computers, and nobody thinks of complaining." -- Jef Raskin

  13. #13
    Join Date
    Nov 2004
    Location
    Germany
    Posts
    2,352

    Default Sorry to chime in

    The new RocketRaid card is using internal 4xSATAII and external one Infiniband to a special 4bay enclosure from them.

    I don't like when "standards" get bypassed cuz you are bound to a brand, I think that is the goal for them.

    Regards

    Nicolas
    Custom Configurations! Rad Hacks and Mods!

  14. #14
    Join Date
    Aug 2001
    Location
    Grangeville, ID USA
    Posts
    9,124

    Default

    ATA-Man,

    SATA-IO.org pushes a bunch of crap most times. They swap specifications around like tokens at a casino. Companies like Seagate push their weight around and 'create' out of new whole clothe entire new realms of specifications that are newly blessed by the io-gods as the 'only' right way. You and I both know that as fast as the spec changes none of us can count on things not being different again in a matter of weeks. And don' t preach to us about how Firmtek follows the SATA-IO.ORGs every little specification. You guys avoided eSATA for over 8 months. Cost basis and reality need to set in here as well. I have yet to ever see reality used as a basis for standardizing to no internal cables in an enclosure. Hocus pocus used to scare the masses into a more limited market. Give me a break. You cannot cherry pick your specs and then use them to beat up on other manufacturers.

    Rick
    molṑn labe'
    "I am a mortal enemy to arbitrary government and unlimited power. I am naturally very jealous for the rights and liberties of my country, and the least encroachment of those invaluable privileges is apt to make my blood boil."
--Ben Franklin

  15. #15
    Join Date
    Feb 2003
    Location
    ex-Cupertino, CA, USA. Now HU+SRB (Europe).
    Posts
    163

    Default

    Quote Originally Posted by ricks
    ATA-Man,
    SATA-IO.org pushes a bunch of crap most times. They swap specifications around like tokens at a casino.Rick
    .... and these guys who push a bunch of crap are running around in a typical SATA Plugfest with equipment worth a mortgage. C'mon Rick, be realistic. As far as I know, the ONLY major issue was the external eSATA spec. The shielded "Type A" connector was just fine, but FoxConn had a different agenda.

    Quote Originally Posted by ricks
    Companies like Seagate push their weight around and 'create' out of new whole clothe entire new realms of specifications that are newly blessed by the io-gods as the 'only' right way. You and I both know that as fast as the spec changes none of us can count on things not being different again in a matter of weeks.
    You and I know, FoxConn made some funny twists with the cable between the shielded "A" and the "B". However, they invested so much in cables, I can uderstand, they had some separate agenda to push.

    Quote Originally Posted by ricks
    And don' t preach to us about how Firmtek follows the SATA-IO.ORGs every little specification. You guys avoided eSATA for over 8 months.
    To my best knowledge, eSATA is being sold now. The first SeriTeks with eSATA begun to appear about 4 month after it became standard. Perhaps some time was needed for the change. Fortunately, FoxConn uses the same material in eSATA as it used in the FirmTe-influenced shielded type-A. No wonder, even if the shielded type-A is not eSATA standard and shall go the way of bird dodo, it luckily STILL PASSES the "EYE PATTERN TEST". So no harm done.

    Quote Originally Posted by ricks
    Cost basis and reality need to set in here as well.
    Right - but please do not use a solution which makes this mortgage-worth-equipment complain loudly. To my best knowledge, "Burly Box" never was tested for the eyepatternt test compliancy. Please make sure, it does pass.

    Quote Originally Posted by ricks
    I have yet to ever see reality used as a basis for standardizing to no internal cables in an enclosure. Hocus pocus used to scare the masses into a more limited market.
    Rick, this is not "hocus pocus". The SATA-IO.org membership enables the access to technology what otherwise many companies cannot afford - think equipment worth 200K. How many times you can see a complain like "the controller cannot see the drives"? At least, with that equipment the developers can see, why it's happening. There are quite a few blames regarding off-shelf built enclosures, with over-engineered power supplies to offset the poor performance. Yes, we DID see enclosures with 300 W power supply where MOST of the power went into heat due poor choice of components.

    Quote Originally Posted by ricks
    Give me a break. You cannot cherry pick your specs and then use them to beat up on other manufacturers.Rick
    As soon as eSATA was feasable to support, it became the standard. The pressure from the customers in bad need of external controllers was very high and after ensuring, they pass the eye pattern test they were realeased before the eSATA became standard. In the fact, FoxConn used the experience with SeriTek/1SE2 to create the eSATA. So much about "cherry picking". I am not beating up on any given manufacturer. All my "guilt" was to correct few incorrect things mentioned here, saying that SeriTek/2eEN4 does fully comply with the current spec and trying to explain, why this spec makes sense. The 1.5 Mbps is not a piece of cake and the 3.0 Mbps is dead serious thing.

    Here is a $1,000,000 question for you. Why is the FireWire speed only a half of SATA at best?

    a) because the high-frequency serial signal transmission is no joke - even for Apple's and TI's budget
    b) because FW800 is the end of the road for the FireWire

    Using "patch cables" inside of an enclosure is messing with the high-frequnecy serial signals. Here is what happening with a few solutions custoers did complain about. There is the innocent controller and either the eSATA or (in worst case) the shielded "type A" cables going between the controller and the case. The case has external (optimally eSATA) connectors. Therefore everything in the theory should be fine, the controller with the cables did pass the eye-pattern test. However, as it turns out, these are essentially "fake" eSATA connectors because on one side they look good, on the other side they feature either a regular "type A" connector (that is the worst case) or just a molded-in unshielded cable. Both solutions are breaking the logic and not just some out-of-blue-sky idealistic SATA-IO.org spec. SATA is very strictly a point-to-point protocoll, without "breaks" in the cabling being allowed. Such breaks mean extra impedance, xtalk, etc. has to be taken care of. As I did mention, the SATA packets are always running at maximum speed negotiated between the points. If the controller and the drive did negotiated a 3 Gbps, that's much worse case.

    Quote Originally Posted by unclemac
    Interesting stuff.

    ATAman: So....one cable is OK to attach the external array to the host, but two (the second being internal in the array case as you mention) is too much for optimal performance and reliablility? Is the issue the second cable, the overall length of all the SATA cabling, too many plugs/connections, or a combination of all of the above?
    Too many plugs essentially break the point-to-point logic. Typical case- for instance, a very troublesome from KingWin: the card is qualified, the connectors on the card are qualified, the cable from card to the box is qualified, the connector on the box which introdices also an internal cable is not qualified. The solution of KingWin *** USUALLY *** does work *** MOST *** of the time. However, the drives on Sonnet Tempo 4+4 disappear and the SeriTek card complains at random that "SATA configuration changed" - which are meaning basically the same thing: the SErr register of either Marvell chip or the SImage or Intel chip got an odd value and the authors of above firmware/drivers have no better idea, than to interpret what happened as a quick "remove-reattach" event. However the change in SErr can also mean SATA communication problem - and in the case of KingWin case it, sure does.

    Of course, KingWin can either re-engineer their cabling by putting some extra compensation at the "joint" point or jusr use a backplane. In many cases backplane is cheaper to do. And of course, after putting extra components for the compensation, the entire thing could pass the "eye test" and imprve the reliability.

    Rick, I am not a SATA-IO.org zealot - for me there is nothing wrong with having mousetails for SATA cables and having to power the box with powered by a hamster running in the wheel - as long as the ENTIRE thing passes the SATA "eye pattern test". To achieve that goal requires some engineering with a backplane, a littele bit more engineering with "patch cables" and "double-end connectors" and quite a bit of engineering with mousetails and hamster. The tendency of today's society (think: 65 mph limit on freeways) to use the least common denominator. We, of course, know, Nicki Lauda in Lamorghini is different to judge than a 90 year old grandma in her Ford Fairmont. Think of SATA-IO.org as of MHP (Montana Highway Patrol). As long as you drive over 70 mph, you could face a judge - who will decide. It well could be, the "Burly Box" is fine at 120 mph, just make sure the eye-pattern test is fine.

  16. #16
    Join Date
    Feb 2001
    Location
    on the landline, Mr. Smith
    Posts
    7,787

    Default

    One question, kinda off topic -

    If FW is indeed at the end of the line, please tell me that external SATA (or something!) will be both bootable and support something like Target Mode. One of my biggest concerns about the upcoming Intel switch, as an admin that relies on Target Mode to clone all sorts of different Macs almost daily, is: will there will be anything as easy and reliable in the future if FW does go away?

    Any clues you can give?
    "Imagine if every Thursday your shoes exploded if you tied them the usual way. This happens to us all the time with computers, and nobody thinks of complaining." -- Jef Raskin

  17. #17
    Join Date
    Feb 2003
    Location
    ex-Cupertino, CA, USA. Now HU+SRB (Europe).
    Posts
    163

    Default

    Quote Originally Posted by unclemac
    One question, kinda off topic -
    If FW is indeed at the end of the line
    Any clues you can give?
    Of course, it's not at the end. My guess is, Apple will proceed with FW1600 in two years. Just the high-speed serial I/O is a tough thing, thus the delay. I don't think, "target" mode will go away. Currently it's built-in to use FireWire, but it could be anything.

  18. #18
    Join Date
    Feb 2003
    Location
    ex-Cupertino, CA, USA. Now HU+SRB (Europe).
    Posts
    163

    Default

    Just making sure, there is no misunderstanding: from two answers

    "a) because the high-frequency serial signal transmission is no joke - even for Apple's and TI's budget
    b) because FW800 is the end of the road for the FireWire"

    one - and only one - is correct, the other is deliberatly wrong. I asked to pick the correct one. This was to illustrate, the high-frequency serial signal is a bit more, than some folks in some dot-orgs playing politics. If FW800 had such a difficult birth and FW1600 is obviously not an easy one - please make sure the signal quality is OK. Nothing more but nothing less. As I explained, if you use hamsters and mice - but the thing still passes the eye pattern test, I am OK with that even if it has nothing to do with what SATA-IO.org wants. The standard was made simply because it's technically easier to pass if you stick to strict point-to-point shielded SATA cable with backplane rather than smart devices based on hamster's running power.

    http://www.hamsterhideout.com/galleryreaders43.html

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •