DEAL OVER! Delta DPS-300BB 300W ATX PSU, OEM ***Now back to full price...$29.99***

[Floyd]

> Well, you learned something yourself and got some hands on experience.

No, I didn't learn anything new. Simply verified that this is one of a number of power supplies that requires a minimum load on the 3.3V rail. Given that these PSU's were likely manufactured for a specific OEM application which would always load the 3.3V rail sufficiently is a perfectly valid engineering decision. However, it certainly would have been nice if Directron had provided this information, assuming they had access to it.

> 1. These PSUs are discontinued - check Delta web site. It is just very likely you are getting salvaged hardware.

This contention has been acknowledged a couple of times. I don't especially care if they are new system pulls. On the two I received, there isn't the slightest evidence they've ever been used. No fuzz on the fan, no scratches on the frame.

> 2. They are not something exceptional as you might think after so much trouble. You'll easily find better PSU for just a little more money.

No argument with this logic, but it doesn't fit the modus operandi of those who frequent the Hot Deals forum hunting a deal. Part of the enjoyment is in purchasing something for less than its usual market value, and learning to use it effectively is another benefit derived from the hobby. If one is purchasing for a business environment, true enough, you probably wouldn't go to a liquidator. The same applies to overclocking a processor...often the time invested in obtaining a solid overclock is unjustified in monetary terms. But that's not what it's about. The individual usually comes away from the experience more knowlegable and better equipped to deal with the day-to-day technical issues that arise in operating and mantaining a computer.

Best regards,
Floyd

 

[Floyd]

> how close are these measurements to readings from your bios...i don't have a meter to measure so can only go by bios. i have the same board as you.

3.33V measured - 3.50V BIOS
4.99V measured - 4.97V BIOS
12.2V measured - 11.94 BIOS

The 3.3V discrepancy is due to the KG7's default I/O voltage of 3.50V...the board is providing local voltage regulation.

Best regards,
Floyd

 

[mindless1]

Originally posted by: stevejst
Well, you learned something yourself and got some hands on experience.
But I need to point out two things that seems to be lost in all of this.
1. These PSUs are discontinued - check Delta web site. It is just very likely you are getting salvaged hardware.
2. They are not something exceptional as you might think after so much trouble. You'll easily find better PSU for just a little more money.

I doubt that these were ever listed on Delta's website. Salvaged (as a term that WE are assuming) can mean just about anything when it comes to 'puter hardware... They aren't defective, and not refurbs or repaired either, that would be obvious... Maybe they were screwed into a new case, sat in a warehouse or two, then unscrewed again before being bubble-wrapped, so they have a fingerprint or two on them if the workers weren't gloved. This is all irrelevant... the only real issue seems to be the additional voltage monitoring, whether the end-user's motherboard puts a load on the 3V. If they were brand-spankin' new, hot off the assembly line, there'd be the same issue.

Maybe this deal is not so hot for anyone without a soldering iron or a little patience, but knowing the issues, knowing about the quality, I'd buy more of them without hesitation (if i needed some).

I've been trying to think of an "easy" solution for someone without a soldering iron, or a shoebox full of power resistors...
Radio Shack has these 12V bi-pin bulbs for $1.99 plus S/H, but in the past my local Radio Shack stores have been stocking these too. I happen to have one here, and it fits OK but not great in the AUX connector. At 3.3V it's going to put out roughly ~5W of heat, not significant to the system unless it happens to be touching something that's heat-sensitive, but I don't know if it's going to have a negative effect on that plastic AUX connector in the long run (it wasn't melting or anything, but hotter than I'd like for a plastic connector). At 3.3V it gives off a soft orange light, not really "cool" looking, like a dim flashlight. Depending on the system it might also work (or might not) plugged into 3V - 5V (orange and red), barely glowing, too hot to hold in your fingers but cool enough for the plastic AUX connector. On the system I tested, the 3V/5V remaind stable with the bulb across those rails. I would consider this the "ultra ghetto" solution, the preferred (makeshift) solution would be finding a power resistor or two (seems Radio Shack is a little light in this department but a couple of these (two in a pack) sinked to the PSU or chassis, wired in parallel, would give 5 Ohm total, probably enough. Add some heat shrink tubing if ordering online (not available in Radio Shack stores AFAIK).

Wouldn't it be nice if Directron offered buyers of these PSU a free (or darn near free) AT motherboard connector with a 4 Ohm power resistor attached and ready to plug in (hint). I think BGMicro (among other places) offers better deals on bulk power resistors than Radio Shack, but I'm unsure of the best place to get male AT power connectors. If Directron was willing to give an estimate of the number needed and ship me the parts I might even assemble these as a service to those in need.

 

[wasamicron]

decided to get going on my Abit KG7. Put everything together using a CompUsa 250w power supply, and everything worked fine. I installed the Delta, and fan started up for a sec., and that's it. Seems like i'll have to do what others have done (as soon as i study the fix a lil' more) to get it to work. I'm still not sure why the Delta power supply would work in my pII system. Anyone know?? Did it provide the load on the 3.3v line like others said was required?? Why doesn't this problem affect all motherboards?? Thanks for everyone's help. I appreciate it.
p.s., I'm admittedly ignorant/stupid about power supplies, but my 250w CompUSA power supply had higher +5, +3.3, and 12v values than Codegan 300w, and a generic 350w. Does this make it a better power supply??

 

[stevejst]

No, the key is stability.
Some PSUs, say Sparkle, have dials inside to adjust the voltage but that is not a good advice to do if you are not confident in what you can do.
Codegen belongs to a low end of AMD approved PSUs, it is not as good as Delta, and certainly nothing to compare with Enermax, Antec, or PC Power and Cooling.

Codegen 300W 12V line stability.

That would be rather poor performance as compared to top quality PSUs. If you are very exclusive about your hardware you would certainly want to have better than Codegen.

By the way, mindless, did you notice CWT-420 at newegg refurbished? I just got two.

 

[kenja]

Dammit, quit making me buy things I don't need!

Listed under Chieftec, there are 43 Channel Wells left.

 

[mindless1]

my 250w CompUSA power supply had higher +5, +3.3, and 12v values than Codegan 300w, and a generic 350w. Does this make it a better power supply??

The short answer is, no, higher values don't in themselves mean anything, or at least it's not that simple. The PSU should output very near it's specified values, and age gracefully, changing little over time. It's up to the motherboard how well this voltage is preserved, and the motherboard's voltage report can just as easily reflect a problem with the motherboard as with a power supply, making it necessary to put a meter to the PSU to tell what the voltage output really is.

More significant is whether the voltages drop when a load is applied, all the way up to the specified capacity will the voltage stay at the specified value at the output connector (ignoring tons of issues like input voltage tolerance and reserve, noise, ripple, recovery, MTBF, etc), this is one of the areas where you can often tell a cheap PSU from an expensive one... I've seen some good 235W-250W PSU that held their voltage under additional load better than some 300W cheapies. I don't know what CompUSA is selling as their 250W.

As sad is it may be, the most important feature of a PSU might be the fan... if it's failure isn't noticed, any power supply can be damaged, though the better ones often have thermal shutdown. <Rant> I would always avoid PSU with Yate Loon Sleeve Bearing Fans, they're the most common cause of PSU failure I've ever seen (though Yate Loon does have ball-bearing models that aren't such an issue). </Rant>

------------------------------------------------

stevejst, thanks for pointing those out, but I just realized something that gives me a little pause... In another AT PSU thread I mentioned that I had an Enlight 340W PSU that "seemed" to be doing fine, but when I pulled it to replace with one of my new Deltas, I discovered a design problem. The 340W has a load resistor resting against a capacitor (actually cemented to it), which overheated it, caused it to vent. I had no trouble repairing that PSU, but obviously that's not the point... upon examing Chieftec's website listing for those PSU, I see that Enlight 340W (actually made by Sirtec) listed along with the 420, and now wonder if the 420 is going to have the same issue. I guess it's not important though, now that I know about that potential I could be sure the resistor isn't resting against the cap.

Here is a pic of the 340W version I have, the area of potential problem (as best as I can remember) is circled (not visible, under the wires), just in case those from Newegg have this same design problem... how does a 420W PSU end up refurbed anyway?

 

[kenja]

Interesting comment, mindless 1. Falls in line with the criticism in this review, which I was thinking was more of an appearance issue.

 

[LumMoose]

I ordered 5 of these Power Supplies because of the great price...tested them out on my Andara Barebones 1 Gig Duron system (came with a 240W Delta originally), and all 5 work just great....no issues at all. Haven't yet tested them on my K7S5A...and will have a MSI K7N415 Pro MoBo in a few days to try it on too.

About a month ago I ordered a couple Enhance PS's from NewEgg Refurbished, but of course one did not work, so one went back, and one is sitting on a shelf.

So..Right now I'm using 2 Antec (350W and 300W), 1 Enermax (300W), and 1 Delta PS (240W).
All I can say to these Deltas is that...it's the best deal I've ever gotten on power supplies. And they're NEW, and GOOD QUALITY for $13!

Can't beat that price.

LumMoose

 

[stevejst]

I would be very cautious about Mikhailtech reviews.
I don't see those pictures too revealing, I do not have CWT-420 yet, it is coming now. I have CWT-350 which doesn't overheat. This is a copy of Antec PP403 or so I hear, and that was a recommendation enough for me.
Make sure you know that will not have extra 12V connector.

 

[Replay]

Grabbed a pair of those $15 shipped CHANNEL WELL CWT-420's from Newegg earlier this evening. The mikhailtech review claims "the CWT-420 is the Antec PP412X! Or, to put things in a more logical context, the Antec power supply is a renamed CWT-420! This is a top notch unit and is respected by many even more so than the Enermax line.". Maybe so...

Great timing! Bought 3 cheap cases today at a local show & had the vendors pull out the crap power supplies and just sell me the cases. All of the cheap case power supplies had bad airflow (no bottom vents) and were feather-weight junk.

Hope these Channel Wells work better than the DOA Newegg GeForce2-GTS-V refurb from a prior order.

 

[stevejst]

PP412X has 4-pin 12V connector, I don't see this one has it, I might be wrong.
I think this is PP403X. Just don't expect too much. There are some differences.
I've seen some strange claims from Michaeltech before, don't buy everything he writes.

PP4XX is one great line and they were the main retail competitor to Enermax 465 line but they are considerably louder and half the MTBF hours in specifications.
I think their new True Power line which has MTBF 30,000 more hours and is considerably quiet plus some other things is probably the best retail line of PSUs at this point. And that is from a fan of Enermax.

 

[stevejst]

I am trying to speculate why they are so cheap so suddenly.
I guess they were sold with Chieftech server cases and because of the lack of Intel ATX12 support makes them invalid for Intel server boards.
So some of us AMD users might be just lucky to get a nice PSU. Well, if not I'll loose $30, not a big deal.

 

[wasamicron]

Success for my Abit KG7! Many thanks to u guys for your super troubleshooting skills (how u figured it out is beyond me). I looked around and found a 40mm bare lead Aavid fan that was 5v & .20a. Not knowing anything about ohms or anything, I thought I'd give it a try. I hooked the fan up using the orange & black wires from aux. connector. Works perfect. My values according to bios are:
Vcor 1.83
3.3v 3.57
5v 5.25
12v 12.54

What do you all think..is the fan a fix, or should i get a resistor from radio shack?? My compgeeks case has a perfect spot to mount the fan at the back towards the bottom.

 

[Floyd]

That's good news, wasamicron. You said the fan is rated at 5V, so putting it on the 3.3V line will make it spin a little slower than its rated speed, but draw more current than its rated 200mA at 5V. I don't see a problem with that at all. Instead of dissipating the power as heat, you're turning it into something useful. Find somewhere to mount the fan permanently and use it to guide airflow through the case or to cover a chassis hot-spot (somewhere air does not normally circulate, but which could probably use it...maybe blowing across your DIMM's).

Best regards,
Floyd

 

[stevejst]

Vcor 1.83
3.3v 3.57
5v 5.25
12v 12.54

If the readings are correct your core voltage is barely safe whichever Athlon you are using. A little more - over 1.85V - and you are in risk of blowing your CPU. If you are not overclocking intentionally with higher voltage you should be concerned what is defective.
3.3V line is way over the safe limit, it is a full decavolt of what Antec considers a safe range and therefore there is a significant risk to damage the motherboard sooner or later, if the readings are correct.

 

[sleefer]

stevejst- I would not base what you stated above on bios readings alone.
Wasamicron- most likely your readings are off, I'd use a mutimeter.
Note: I received a CWT-420 that's mentioned above with a Cheiftech case from Newegg. Wound up sending it back (doa) and they replaced it with an Antec PP-412X. That was several months ago but, back then Newegg was sending some of them set for 230V instead of 115V. You may get lucky and only need to flip it to 115V. On the other hand I would be very cautious of a unit that some customer may have tampered with and then returned to Newegg for whatever reason. I wouldn't just pop one of these RMA units in and hit the switch on it. I believe the CWT-420 did have seals on it if I remember right, so you should be able to tell if someone did open it up. I doubt Newegg does any testing at all on these, if they are returns. Then again as stevejst mentioned the may just be pulls. My experience with Newegg refurbs is that you can get either pulls or returns from them.

Back to the Delta again, I tested one with the KK266-R and 8KHA+. Both needed a load on the 3V rail in order to power up. I think that pretty much confirms what's being reported here.

 

[mindless1]

wasamicron, if that 3.3V reading is of voltage regulated down from the 5V line on the motherboard (which it likely is), you may be fine, can ignore the rest of this post or take a wait-and-see attitude.

Oh the other hand, if the 3.3V reading (at 3.57V) really is the 3.3V output from the PSU, then on THAT particular system, the fan you used is barely enough to keep the PSU turned on, it isn't loading the 3V rail enough to stabilize the output and keep it at it's intended voltage. If this is the situation, it may not be a problem in itself if the system isn't using the 3V output (or else you wouldn't need the load to start with), but then again it isn't ideal either. You could leave it alone, wait and see if problems develop, which they might if you haven't provided a high-quality contact between the AUX connector and the fan. The better long-term solution might still be a power resistor, and your test with that fan suggests that a higher value resistor could be used, like 6 or 8 Ohm, would generate less heat than the initial test value I originally proposed of 2-3 Ohm.

wasamicron, the following isn't directed at you, but it's certainly worth mentioning, especially since it's likely that the less-electrically-adept owners of these PSU may now be considering doing the mod:

-------------------------------------------------------------------------

Anyone doing this mod is doing so at their own risk. Regardless of whether you're happy or upset with Directron's deal, it's not their fault (and I also deny responsiblity) if you damage your system. Having an ADEQUATE, STABLE, ELECTRICALLY AND THERMALLY ISOLATED load on the 3.3V rail is not in itself a potential for damage.

 

[stevejst]

stevejst- I would not base what you stated above on bios readings alone.

I know, the warning is conditional. I'd be careful about that anyway.

Sleefer, I am aware of the risks when buying newegg refurbished. $15 is acceptable risk with newegg, in my view. The best pricewatch deal for CWT-420 I could find was $46. They sold about 50 or so CWT-420 yesterday which a large number, usually they have few PSUs only as refurbished. For example there have only 4 Enermax 465P now. I wrote about my reasoning above, it seems this is about the same type of supply as this Delta but the deal is sweeter, $15 FOR 400W versus $20 for 300W, considering shipping.
I did need 400W PSU, I didn't really need 300W, I have two lying around.

So you got ATX 12 certified PP412X, you are lucky. That is a great PSU according to everybody that has it.
CWT-420 and PP4XX supplies are a little on a noisy side from what I hear and know about CWT-350 but I use Enermax in my 24/7 computer so these guys are only for my computers that are used during the daytime. I can take it, they cannot be worse than Enhance which is OK if you are not using it 24/7.

 

[wasamicron]

Thanks for all the info. I'm going to go at it slow, and try to check everything out. This is an excellant learning opportunity for me. I would like to quote Floyd about his readings on the same board that I have-an Abit KG7.
> how close are these measurements to readings from your bios...i don't have a meter to measure so can only go by bios. i have the same board as you.

3.33V measured - 3.50V BIOS
4.99V measured - 4.97V BIOS
12.2V measured - 11.94 BIOS

The 3.3V discrepancy is due to the KG7's default I/O voltage of 3.50V...the board is providing local voltage regulation.

Best regards,
Floyd

Doesn't this sound like what my board may be doing?

 

[stevejst]

Doesn't this sound like what my board may be doing?

I don't have Abit KG7 so I cannot make comparison but your readings are ALL on the upper side, and 12V, 5V are barely safe as well.
Out of 13 computers I have, I have never seen core voltage 1.83V without me increasing it actually.
Perhaps the readings are off on the high side but it is equally possible that something gets blown in your PC, in my view. Since we are all guessing, you will have to make the judgement.
A good idea to pursue this further would be to take another PSU and see what the readings are with a different PSU. That would give you a little more information to juggle with.

 

[mindless1]

wasamicron, yes, that 3.3V does sound like a locally regulated value, I hadn't noticed your mobo was the same as Floyd's.

As for the 1.83 Vcore stevejst mentioned, the PSU has nothing to do with that, is a motherboard issue.

 

[Floyd]

wasamicron, just to confirm what you are thinking, the KG7 allows one to select between either 3.5V or 3.65V for the I/O Voltage. The reading in the BIOS tracks this change pretty closely (within 3 hundredths of a volt). It's peculiar that the writers of Abit's BIOS would retain the label of "3.3V" in the PC Health section when the board regulates to 3.5V nominal (default). Nonetheless, regardless of whether I choose 3.5V or 3.65V in SoftMenu, the Delta is providing a solid 3.33V on its AUX connector. In other words, your motherboard's voltage appears to be in the range Abit's engineers intended.

Best regards,
Floyd

 

[stevejst]

As for the 1.83 Vcore stevejst mentioned, the PSU has nothing to do with that, is a motherboard issue

Again I have no KG7 but none of my motherboards shows anything like that when at the default voltage. I have one other 761 motherboard by Giga-byte and the core is 1.75V dead even.

For an Abit motherboard where the importance of core voltage readings directly impacts the ability to overclock such an inaccuracy would seem surprising to me. Because these boards are made for overclockers.

As far as 3.3 versus 3.5 in BIOS, download the newest BIOS update, it appears they corrected that and the BIOS reads 3.5V.

Check here.

 

[sleefer]

Original by by stevejstSleefer, I am aware of the risks when buying newegg refurbished. $15 is acceptable risk with newegg, in my view. The best pricewatch deal for CWT-420 I could find was $46. They sold about 50 or so CWT-420 yesterday which a large number, usually they have few PSUs only as refurbished. For example there have only 4 Enermax 465P now. I wrote about my reasoning above, it seems this is about the same type of supply as this Delta but the deal is sweeter, $15 FOR 400W versus $20 for 300W, considering shipping.
I did need 400W PSU, I didn't really need 300W, I have two lying around.

stevejst-just because you are aware of the risks is no indication that everyone here is. My posting on the CWT-420 was not directed at you, but for everyone's information. You saw fit to post it in this thread. I feel that it's only fair to see to it that everyone else be made aware of the risks. My own experience with Newegg refurbs is the basis for the warning I gave. I understand your reasoning and it could prove to be true, but that hasn't been verified by anyone yet either. If your right about it, it's a great deal. But what if your wrong? Some people may not be as happy to lose $15 as you are.