best cheap/free scores 1.1

[Dan81]

Re: best cheap/free scores 1.1

If it's a half-bridge PSU or uses CM6800 chip, just remove/disable APFC circuit and carry on.

That one uses TL3842P it seems, along with a LM339. Guess it's half bridge?

[RJARRRPCGP]

Re: best cheap/free scores 1.1

I asked the guy that gave me the Pegatron IPMTB-TK what he did with the rest of it....I saw the smashed case sitting out in his garage.... He bilked the 2x SSD's for some laptops he had, but he said the rest was junk, and I was welcome to it.... I love what some consider 'junk'...so let's see:

1x Xeon W3690 CPU (3.46GHz hex-core)
6x 2gb sticks of PC3-10600E
...and the pick of the litter: a PNY Quadro M4000 8gb GPU!

That CPU is awesome! But I'm afraid the Asus P6T Deluxe won't have support for it.

[Topcat]

Re: best cheap/free scores 1.1

That CPU is awesome! But I'm afraid the Asus P6T Deluxe won't have support for it.

I have the X-version (X5690) of it in a dual rig, and they function nice....infact its what I tested the board with, I have one in my test CPU bin. The W3690 is basically the same as the X5690, except it can't be run in SMP (dual CPU boards) and it maxes memory support @ 24gb.....which is the max supported by the chipset anyway (x58), so being a single board with the x58 chipset, neither of those limitations are factors anyway.....I'd hate to imagine what some poor soul paid for this workstation.... A quick check, these GPU's actively sell on ebay for ~$300-ish....the CPU and board is cheap though....so is the RAM....but free is wonderful. The case OTOH looks like someone tied a rope around it and dragged it from the bumper of a car for a few miles down the highway....it boggles my mind how it got so demolished, and the guy I got it from didn't know either....it was just in a bunch of crap that someone unloaded on him......which was mainly a bunch of HP laptops.

[dmill89]

Re: best cheap/free scores 1.1

A Sony STR-DH130 Frankenreceiver. I pieced this thing together from parts from 3 different scrap receivers, 2 of which had been dropped (1 had a broken main board and the other had a broken VFD, both had case damage) and one somebody spilled soda in (did a number on the electronics but had a good case).







Man are modern low to mid end receivers underwhelming, at least this one has descrete transistors and some of the caps are decent (some Chemi-Cons but also some Lelon), Sony claims 100W per channel into 8 ohms, but I wouldn't bet on it actually doing more than 50W per channel (at least sustained). It is almost hard to believe they actually charged $150 for these things when they were new. The only weight this thing has is the transformer if you remove that it feels like an empty tin box, and the metal the case is made from is so thin you can easily bend it with your fingers (and not allot of pressure).

Granted some newer receivers are reasonably well built but generally only at the high end, for comparison here's the insides of my Pioneer Elite VSX-84TXSi, of course the Pioneer cost ten time as much as the Sony did when new (and is 7.1 channel vs 2 channel). I realized I never posted an interior shot of it when I originally posted it a little while back:

[ratdude747]

Re: best cheap/free scores 1.1

Made a potential pawn shop score. While poking around a pawn shop in TN today (on a solo business trip), found a 3.5" hdd buried in one of the display cases... it's a 4TB Seagate Ironwolf. Has been there for over a year (drive was made in February 2017)... the shop offered $20, and of course I took the offer.

Bozo here didnt bring his SATA adapter this time, so I won't be able to test it until I get home... but even for a "small" 4TB drive (they go up to 14TB per drive) $20 is nothing. Cheap enough to be worth the gamble.

I needed a drive about this size to upgrade my server's internal backup drive (so I can store more than just 1 week of images at a time)... not to mention that the server's main RAID array is 1.75TB but the current backup drive is only 500GB... not too good

Got her home... she does work, but there's 100 pending sectors and 96 reallocated. Still OK to use for a server backup image drive? It's only a 6000RPM spinner, so it's not good for much else.

[Curious.George]

Re: best cheap/free scores 1.1

Got her home... she does work, but there's 100 pending sectors and 96 reallocated. Still OK to use for a server backup image drive? It's only a 6000RPM spinner, so it's not good for much else.

Yikes! I toss drives at 1 remapped sector (have to have some criteria!)

You can reason that a backup is less likely to see any real use...

Or, can reason that when it does see use it's because you're already up against the ropes!

Have you tried using it and watching performance across the surface? I "time" each write() operation as I wipe the drive and watch for individual write()s that take longer than expected to tell me when the drive is "hiccupping".

[ratdude747]

Re: best cheap/free scores 1.1

Yikes! I toss drives at 1 remapped sector (have to have some criteria!)

You can reason that a backup is less likely to see any real use...

Or, can reason that when it does see use it's because you're already up against the ropes!

Have you tried using it and watching performance across the surface? I "time" each write() operation as I wipe the drive and watch for individual write()s that take longer than expected to tell me when the drive is "hiccupping".

She's bin food... tried to make a new partition, cyclic redundancy check city. Only 100 power on hours too... no wonder it was pawned off.

At least it's only $20... Probably can't RMA it... I wish, lol.

[ratdude747]

Re: best cheap/free scores 1.1

She's bin food... tried to make a new partition, cyclic redundancy check city. Only 100 power on hours too... no wonder it was pawned off.

At least it's only $20... Probably can't RMA it... I wish, lol.

Nope, still under warranty until Febuary 2020... so RMA'ing it. :p

We shall see how this plays out.

[TechGeek]

Re: best cheap/free scores 1.1

Scored a 2002 Onkyo HT-R500 for free yesterday. Works fine, just lots of rust. Could've been used on contractor jobsited, IDK. Was pulled from a repossessed vehicle.

[ChaosLegionnaire]

Re: best cheap/free scores 1.1

wow... at 100 hrs poh and its already racking up bad sectors! shows u how well built new high density drives are!

[ratdude747]

Re: best cheap/free scores 1.1

wow... at 100 hrs poh and its already racking up bad sectors! shows u how well built new high density drives are!

I don't know it's past history... was it dropped? who knows. Hell the RMA might get rejected... no idea.

[mikay786]

Re: best cheap/free scores 1.1

Carboot was closed today due to weather so went to a smaller one that’s located under a bridge but probably only around 20 sellers.

Got an archos 705 WiFi 80gb - £5
iPhone 5c - £7.50

iPhone doesn’t turn on, going try switching batteries from a working 5s
Archos, don’t have power adapter and battery may be dead. When I connect to pc and with a random 5v phone charger it flicks on for a second. Screen looks fine.

[Topcat]

Re: best cheap/free scores 1.1

The hardware rescuer in me bought an X6DAL-XG with IO shield for $35 (cheap for these) on ebay last night....it may take a while but I'll find a home for it.

[Dan81]

Re: best cheap/free scores 1.1

I got these for free:

-lots of laptop SATA HDDs, in different stages of bad sectors (2 boxes full of 'em)
-unknown brand slotket
-ASUS WiFi@Home card (can post pictures tomorrow if anyone wants to see it)
-DeepCool VGA heatsink w/ fan
-Raidmax Sagitta case
-Samsung SH-224 SATA DVD-RW
-Geforce 2 MX200
-S3 Savage4
-Geforce 2 MX400
-Geforce 4 MX440SE
-Geforce FX5200 128MB
-Radeon 7500 64MB
-Gecube Radeon 9550 128MB AGP8x - ironically the older version I had of this had Sanyos, while this newer one was RIDDLED TO THE BRIM with bulged Sacon FZs (same goes for a Inno3D GF6200 I have)
-Geforce FX6200 128MB (both AGP and PCI-E)
-Geforce 7200 256MB
-Geforce 7300 128MB AGP
-Geforce 7900GS 256MB DDR3 (PCI-E)
-Geforce 8400GS 256MB
-Geforce 8500GT 512MB
-Winfast Geforce 8500GT 256MB
-BFG Geforce 8600GT 256MB
-Inno3D 9800GT 512MB DDR3 (definitely going to reflow this - it only has minor artefacts anyways, so nothing unfixable)
-Geforce 210 512MB
-Sapphire Radeon HD4650 512MB (this one shorts out and triggers PSU protection)
-Winfast TV Pro II VGA TV Tuner
-generic 1280x768 LCD TV Tuner
-MSI K9N4 Ultra w/ 2x512MB DDR2-533 GeIL RAM/Athlon 64 3800+ 2.4GHz
-nifty Turbo display

[Topcat]

Re: best cheap/free scores 1.1

Picked up another cheap BP6 in untested condition. These are starting to fetch decent prices these days....so I snap them up when I can get them cheap....many times sellers that don't really realize what they have and stick low BIN prices on them, or 'make offer', which is a golden opportunity to lowball.

[momaka]

Re: best cheap/free scores 1.1

I asked the guy that gave me the Pegatron IPMTB-TK what he did with the rest of it....I saw the smashed case sitting out in his garage.... He bilked the 2x SSD's for some laptops he had, but he said the rest was junk, and I was welcome to it.... I love what some consider 'junk'...so let's see:

1x Xeon W3690 CPU (3.46GHz hex-core)
6x 2gb sticks of PC3-10600E
...and the pick of the litter: a PNY Quadro M4000 8gb GPU!

LOL. Nice!!!
Some people just don't know what they have. If it ain't the latest, it's trashy and slow.
On that note, I wonder if the person who this smashed case came from is a "gamer". No disrespect to anyone who identifies as that, but I have to say that most PC gamers I've seen tend to be total slobs when it comes to caring for their stuff. I got a subwoofer for parts a few months back for cheap, and the case was just in pretty bad condition - all scratched up and stained. Speaking of which, what is it that people/gamers spill on their cases that it always leaves such stains/residue that I cannot remove with anything. I noticed the same thing on the Antec case I posted about here some pages ago. Is it Mountain Dew? Monster/energy drinks? Whatever it is, I cannot wash it off with anything. Tried soapy solution, window cleaner (with ammonia), Chlorine-based wipes... even 99.9% IPA does nothing.
Anyways, I just can't seem to understand how some people end up trashing their stuff they way they do. I can only do that if I actually tried.

That one uses TL3842P it seems, along with a LM339. Guess it's half bridge?

Not exactly. 3842 is current-mode PWM controller, which implies single-transistor forward topology. However, it could be that the 3842 is used for the APFC circuit and there is another controller for the main PS. You need to find all of the controllers in that PSU to determine what it uses. I seriously doubt the 3842 and LM339 are the only ICs there.

A Sony STR-DH130 Frankenreceiver...

Haha, awesome!
So did you have parts from three identical STR-DH130 receivers? I can't seem to find any "funny handiwork" in there.

Man are modern low to mid end receivers underwhelming, at least this one has descrete transistors and some of the caps are decent (some Chemi-Cons but also some Lelon), Sony claims 100W per channel into 8 ohms, but I wouldn't bet on it actually doing more than 50W per channel (at least sustained).

Well, the power consumption on the back of the unit says 200 Watts max. If that is indeed correct, then this makes it 100 Watts per channel - but that's with the amplifier consumption (i.e. losses) included. Given that class AB amps have a typical maximum efficiency of about 65%, I guess we can expect a maximum of 65 Watts / channel.
So yeah, your guess of 50 W/ch probably is a very good approximation.

Good that it does have discrete output stage, though. You'd be hard-pressed to find an amp today that doesn't use a class-D chip/module of some sort. Almost all do. :\

Got her home... she does work, but there's 100 pending sectors and 96 reallocated. Still OK to use for a server backup image drive? It's only a 6000RPM spinner, so it's not good for much else.

Hard to say. Put it in a PC and start using it as an every day "backup", just to give it some normal reading/writing to do. If the bad sectors don't increase in a few months of use, it's probably OK now.

Yikes! I toss drives at 1 remapped sector (have to have some criteria!)

That's because you have a lot of drives to choose from / deal with.
If you didn't, I'm sure you could let that slip a little.

Also, 1 bad sector doesn't mean that the drive is more likely to fail than a drive with zero bad sectors. The two are related, sure, but not entirely linked.

Like anything else in manufacturing, there are always variations and imperfections. Same applies for the disk surface of an HDD. And because of this, sometimes you may get an HDD that passes factory testing but really is marginal on certain surfaces on its disk. It's only with time and use that these imperfections may eventually come out and make certain sectors unreadable. However, that doesn't mean the HDD will necessarily fail. Yes, if you have data on those sectors and they do become unreadable, that could technically be considered a failure. Either way, it's always possible that once all of the "bad"/marginal surfaces have failed or have been remapped that the HDD would then only have "good" sectors left to write to. In which case, the HDD could be considered "stable" from that point forward. I actually have many HDDs like that - i.e. HDDs with a few bad sectors, but where I haven't seen their bad sector count increase anymore after that. Some of these have served me for more than 5 years now!

So bad sectors (within reason) doesn't necessarily mean bad/failing HDD.
It just means, use it with caution (if you dare), and if it doesn't fail/get worse after a long time, it's probably stabilized and is reliable enough to use.

-Raidmax Sagitta case

Not bad.

-Geforce 7300 128MB AGP

Nice. I got one here too.
Put a nice big cooler on it and you should be able to overclock it close to the core clocks of the PCI-E 7600 GT (which IIRC is 560 MHz core). I haven't seen a 7300 GT with a good stock cooler anyways (and that's actually an understatement, as many simple come with more of a joke than an actual cooler.)

-Inno3D 9800GT 512MB DDR3 (definitely going to reflow this - it only has minor artefacts anyways, so nothing unfixable)

Good luck! I had the same happen on a PNY 8800 GT XLR8 and couldn't revive it. After 2nd reflow, actually, GPU went short-circuit between all 3 power rails. RIP!

-Sapphire Radeon HD4650 512MB (this one shorts out and triggers PSU protection)

Usually that is bad news for an ATI card if it's any of the high-end models. But for a mid-range, you might be lucky and it might just be a shorted FET without the GPU chip being bad.

[dmill89]

Re: best cheap/free scores 1.1

Haha, awesome!
So did you have parts from three identical STR-DH130 receivers? I can't seem to find any "funny handiwork" in there.

Yep, it is pieced together from 3 different receivers I acquired for parts. Two of them I got years ago back when I was selling electronics in college (both shipping damaged items they wrote off and were going to throw away, this was very common with these units as the metal the case is made from is so thin it is easy to bend and crack a board if dropped/kicked around) then threw them in a closet and forgot about them (basically saving them if I ever needed to strip parts for something else) and the third I recently found sitting on the curb (the one with liquid damage) and managed to find enough good parts from the three units to make one good receiver. They weren't exactly identical (some of the PCBs were slightly different color, and some components on the boards were from different manufactures but everything plugged in and worked correctly so Sony should have been sourcing parts from multiple suppliers). And of course I got a pile of caps and relays as well as some transistors from what was left over.

[Curious.George]

Re: best cheap/free scores 1.1

Yikes! I toss drives at 1 remapped sector (have to have some criteria!)

That's because you have a lot of drives to choose from / deal with.
If you didn't, I'm sure you could let that slip a little.

But, you still have to have SOME criteria -- whether you've explicitly codified it, or not. When do you consider a drive to have "too many" bad sectors?

Also, 1 bad sector doesn't mean that the drive is more likely to fail than a drive with zero bad sectors. The two are related, sure, but not entirely linked.

As I said, "have to have some criteria!". "One" is mine. "Zero" wouldn't make sense. And, it's too hard to decide how much higher than "one" you should set your cutoff!

Does 2 bad sectors more accurately reflect an impending failure? What about 3? Or 47?

"After their first reallocation, drives are over 14 times more likely to fail within 60 days than drives without reallocation counts, making the critical threshold for this parameter also one." -- google

[There's actually been some research that questions the ability of SMART to act as an accurate predictor, at all!]

You can ask yourself the same question about ECC memory... when do you start seriously wondering about UNDETECTED errors? If your ECC corrects an error today, should you be leary that it is more likely to be missing some errors, as well??

Like anything else in manufacturing, there are always variations and imperfections. Same applies for the disk surface of an HDD. And because of this, sometimes you may get an HDD that passes factory testing but really is marginal on certain surfaces on its disk. It's only with time and use that these imperfections may eventually come out and make certain sectors unreadable. However, that doesn't mean the HDD will necessarily fail. Yes, if you have data on those sectors and they do become unreadable, that could technically be considered a failure. Either way, it's always possible that once all of the "bad"/marginal surfaces have failed or have been remapped that the HDD would then only have "good" sectors left to write to. In which case, the HDD could be considered "stable" from that point forward. I actually have many HDDs like that - i.e. HDDs with a few bad sectors, but where I haven't seen their bad sector count increase anymore after that. Some of these have served me for more than 5 years now!

So, if the grown defect list kept increasing for the first 4.9 of those years, how would you decide to stick it out for 5 years in the hope that the number of grown defects would magically stop increasing?

"have to have some criteria!"

So bad sectors (within reason) doesn't necessarily mean bad/failing HDD.
It just means, use it with caution (if you dare), and if it doesn't fail/get worse after a long time, it's probably stabilized and is reliable enough to use.

There's a lot of "lore" around disks (and other components) that doesn't seem to hold up, in practice. Have a read at some of the larger reliability studies that have been published to see where the lore falls down. (google has a great study of a few gazillion drive-hours from their server farms)

"One of those we thought was most intriguing was that drives often needed replacement for issues that SMART drive status polling didn't or couldn't determine, and 56% of failed drives did not raise any significant SMART flags (and that's interesting, of course, because SMART exists solely to survey hard drive health)"

Ooops!

[momaka]

Re: best cheap/free scores 1.1

But, you still have to have SOME criteria -- whether you've explicitly codified it, or not. When do you consider a drive to have "too many" bad sectors?

See, that's the thing: you're trying to make an exact science out of this, whereas I think of it more of an art of sorts ... or rather a balance between various factors, but no exact criteria.

Why?

Well, you answered that question yourself with the Google study (part of which I also read a long time ago) - it's because SMART isn't the absolute verdict when trying to establish the reliability of an HDD. Sure, an HDD is more likely to fail if it has bad sectors, and especially if it keeps racking them up quickly with everyday normal use. But there are still also many cases... at least with older drives... where an HDD may rack up some bad sectors, then stop or go very slowly at it... at which point, one could consider that HDD "stable enough" for general use.

With that said, my criteria is *generally*/*roughly* that:
- if an HDD has no bad sectors at all and at least 5k hours on the clock, it's very likely to be very reliable drive, and I will trust it with my data (but still keep a backup - as I always do!)
- if there are only a few bad sectors, but the HDD has 5-10k hours on the clock and neither the BSC (Bad Sector Count) nor the CPS (Current Pending Sector) are increasing much over time (less than 5 sectors per year), it's an OK drive for general use / testing.
- if HDD has lots of bad sectors (over 50-200) and also more than 10k hours on the clock, it's either getting "tired" or marginal, and may only be OK to use for absolutely non-critical tasks.
- if HDD has very few hours on the clock (less than 1k) but with lots of bad/pending sectors (over 100), it's probably not going to last or will fail soon.

[Curious.George]

Re: best cheap/free scores 1.1

See, that's the thing: you're trying to make an exact science out of this, whereas I think of it more of an art of sorts ... or rather a balance between various factors, but no exact criteria.

Why?

Because hardware is cheap; time is expensive!

The computer (and its components) is just a tool. I expect my tools to "just work". If I encounter a screwdriver that some fool has over-torqued, I toss it out. If I find a screw with a mangled head, I toss it out. If a disk throws an error (even if it corrects it), I toss it out. I don't want to risk it NOT correcting the next error -- or, find out that in a routine patrol it encountered a bad sector and was not able to recover the data!

The time spent worrying about whether or not another problem -- from which I may not be able to recover! -- lurks around the corner is a needless distraction.

I don't use software ("tools") that I can't rely on for the same reason. If I lose an hour of work because a program crashes, there's no guarantee that I can remember what I did in that hour and successfully recreate that effort! What's the cost of forgetting some detail that I had successfully addressed, initially?

Well, you answered that question yourself with the Google study (part of which I also read a long time ago) - it's because SMART isn't the absolute verdict when trying to establish the reliability of an HDD. Sure, an HDD is more likely to fail if it has bad sectors, and especially if it keeps racking them up quickly with everyday normal use. But there are still also many cases... at least with older drives... where an HDD may rack up some bad sectors, then stop or go very slowly at it... at which point, one could consider that HDD "stable enough" for general use.

Google's complaint with SMART was that there were too many failures that SMART did not predict. I.e., cases where there was no error reported prior to the failure. OTOH, they were able to glean reasonably predictive data about drives that actually DID fail once they had reported SMART errors (particularly scan and sector faults). Their conclusion was that SMART didn't (apparently) cover all of the issues that would allow for a more comprehensive prediction.

With that said, my criteria is *generally*/*roughly* that:
- if an HDD has no bad sectors at all and at least 5k hours on the clock, it's very likely to be very reliable drive, and I will trust it with my data (but still keep a backup - as I always do!)

What happens if you encounter a failure between backups (which will always be the case -- think about it)? Presumably, those backups don't happen "continuously". Can you recall every change you've made to your disks' contents since the most recent backup?

E.g., if you backup every day, then, on average, there are 4 hours of "work" that you will lose for a system/disk failure. Can you remember what you did in those 4 hours so you can recreate those efforts? Will you, instead, have to invest those 4 hours, again? Are you sure you will stumble across the same things that you did, originally?

[My work is largely creative; if I don't recall an idea that I had earlier, it's lost! What cost, that?]

- if there are only a few bad sectors, but the HDD has 5-10k hours on the clock and neither the BSC (Bad Sector Count) nor the CPS (Current Pending Sector) are increasing much over time (less than 5 sectors per year), it's an OK drive for general use / testing.
- if HDD has lots of bad sectors (over 50-200) and also more than 10k hours on the clock, it's either getting "tired" or marginal, and may only be OK to use for absolutely non-critical tasks.
- if HDD has very few hours on the clock (less than 1k) but with lots of bad/pending sectors (over 100), it's probably not going to last or will fail soon.

My criteria is simpler; I don't worry about PoH's at all. I have drives with 40K hours that still haven't thrown a bad sector. When/if they do, I'll pull all of the data off of them and toss them out.

My fear of loss means I want to make sure I can pull all of the data in a reasonably short amount of time (hence no RAID as rebuilding a large array leaves you vulnerable for far too long). For that reason, I prefer smaller spindles -- ideally, less than 2T.

I've also been discarding "external" drives for similar reasons (at least, those that have USB2 i/f's) -- they just take too damn long to scrape all their data off if they start "acting up"!