Toaster
12-22-2001, 01:42 PM
Hello folks,
I have a habit of buying up multiple "failed" parts and pieces from shops.
This includes most anything harware related. When a failure "rash" occurs, I start looking closely to detiermine cause and effect. Parts include HDDs, power supplies, system boards and the like.
Several shops started using a power supply manufactured under the "Allied" name brand. Many of the shops have had a rather high percentage of these fail in the field. While this isn't something of high concern, it did however warrent investigation. In most of the failures, the supply simply failed without damage to connected components. There was a small number of instances where connected components were damaged. The number was quite small however.
I was given 10 such power supplies, all of the same make and model.
The advertised, (stickered) specs and model information follow:
Name implied: Allied
Model/part number: AL300-ATX (each were identical but part# suffix varied)
(the AL300-ATX has suffixes of xxxx-ATX-P4-TB etc.)
Form factor: ATX w/P4 support
Claimed wattage: 300watts (stated as maximum)
Fan/fan type: Simple fan (4") no thermal control.
Classification: (claimed) AMD and Intel certified/approved. ATX 2.02 compliant.
Claimes compliancy with AMD/Intel/P4 (yes, yes, yes)
Certification labs (claimed): CSA, DVE, CE, RU, CB
There was no mention of UL certification in any form.
Input voltage selection. Claimed as "auto detected" (no switch)
This power supply suffered from input rectification/filtering failures and was the primary cause of failure. All but 2 of the supplies failed for the above.
The remaining 2 failed due to overheating resulting from fan failure.
The failure stemmed from a bridge rectifier failure and/or filter capacitor failure due to under rated parts. The filter capacitors were rated at 400uf/100volts DC. The rectifier assembly (bridge rectifier) was rated at 5amps/200 volts DC. If the filter capacitors failed first, these shorted, and represented a short to the rectifier bridge and thus damaging the rectifier and blowing the A/C input fuse. The input A/C fuse was rated at 6 amps.
One of the supplies also failed when the input A/C filtering was under-rated which resulted in overheating and failure. (MOV failure) (used for surge protection and spike suppression)
The primary failure was A/C to D/C conversion (filter caps and rectification).
The rectifier should have had a minimum 400volt rating and the caps should have been rated for 400 volts minimum (for universal A/C input which was claimed)
I repaired 3 of the supplies that failed for the above reason for testing purposes with properly rated parts. I had 5 amp fuses but there was no availability of 6amp fuses. Therefore, A/C fusing was 5amps.
Power supply wattage/amperage: (claimed as maximums)
3,3volts-18amps
5volts(+)-25amps
12volts(+)-9amps
Standby (5v standby)-1.50 amps.
-5/-12 volts- 350ma (each)
No mention of maximum wattage of combined 5volt/3,3 volt.
PCB thickness-1.6mil
PCB support- 3 points via screws, one point via a nylon standoff.
When the repaired supply was bench tested at 25%, 50%,75%, 100%, 120% of rated output, the outputs remained within 4% of nominal except at 120%. At 120%, the line fuse failed. (remember the fuse was below specified capability)
The outputs remained quite clean from 25% upto and exceeding 75%.
The outputs were still clean beyond 75% but "switching noise" increased somewhat. (this is normal really, it just means the supply was working quite hard)
The supply provided clean outputs and started properly when the line voltage was between 92volts and 125 volts. (average)
This supply was intolerant of minimal loads. The supply failed to "start" below 15% of rated power. At 20% of rated power, the outputs were 7% of nominal ratings and noise/ripple increased markedly. The power supply was in its element at loads of and exceeding 25%. The supply ran cool above 25% but ran noticably warmer at 20%. The supply was set to 100% load and left to run for 6 hours.
At the close of 6 hours of operation, the supply was running reasonably cool considering the output developed. The power supply tolerated full power from a cold start. (this is good)
The power supply output cabling was too light IMHO but was adaquate at 20guage.
This supply would have faired better were the output cable size were 18g or better.
The supply cooling was adaquate but easily improved. Various wiring interfeared with ventilation. When this wiring was moved and secured, the noise level of the supply dropped and the fan became more efficient.
Bottom line and recommendations:
This supply "could" have been a desent if the maker made a few changes. The changes should't have had an impact of production costs. I surmise the supply may have cost and additional 0.25$ with the correctly rated components.
These components are: Rectifier bridge assembly, 2 filter caps and a fan.
When repaired, the supply easily met its claimed wattage.
Users may wish to choose a different supply until these anomolies are corrected.
The supply never "smoked" and the line fuse was properly sized.
When the components failed, they most often shorted blowing the mains fuse.
I do NOT recommend this supply, especially on 220volt mains where the internals would be grossly under-rated.
Upcoming: (so far the SPI/sparkle looks the best)
SPI (sparkle)
Antec
Enlight
I have a habit of buying up multiple "failed" parts and pieces from shops.
This includes most anything harware related. When a failure "rash" occurs, I start looking closely to detiermine cause and effect. Parts include HDDs, power supplies, system boards and the like.
Several shops started using a power supply manufactured under the "Allied" name brand. Many of the shops have had a rather high percentage of these fail in the field. While this isn't something of high concern, it did however warrent investigation. In most of the failures, the supply simply failed without damage to connected components. There was a small number of instances where connected components were damaged. The number was quite small however.
I was given 10 such power supplies, all of the same make and model.
The advertised, (stickered) specs and model information follow:
Name implied: Allied
Model/part number: AL300-ATX (each were identical but part# suffix varied)
(the AL300-ATX has suffixes of xxxx-ATX-P4-TB etc.)
Form factor: ATX w/P4 support
Claimed wattage: 300watts (stated as maximum)
Fan/fan type: Simple fan (4") no thermal control.
Classification: (claimed) AMD and Intel certified/approved. ATX 2.02 compliant.
Claimes compliancy with AMD/Intel/P4 (yes, yes, yes)
Certification labs (claimed): CSA, DVE, CE, RU, CB
There was no mention of UL certification in any form.
Input voltage selection. Claimed as "auto detected" (no switch)
This power supply suffered from input rectification/filtering failures and was the primary cause of failure. All but 2 of the supplies failed for the above.
The remaining 2 failed due to overheating resulting from fan failure.
The failure stemmed from a bridge rectifier failure and/or filter capacitor failure due to under rated parts. The filter capacitors were rated at 400uf/100volts DC. The rectifier assembly (bridge rectifier) was rated at 5amps/200 volts DC. If the filter capacitors failed first, these shorted, and represented a short to the rectifier bridge and thus damaging the rectifier and blowing the A/C input fuse. The input A/C fuse was rated at 6 amps.
One of the supplies also failed when the input A/C filtering was under-rated which resulted in overheating and failure. (MOV failure) (used for surge protection and spike suppression)
The primary failure was A/C to D/C conversion (filter caps and rectification).
The rectifier should have had a minimum 400volt rating and the caps should have been rated for 400 volts minimum (for universal A/C input which was claimed)
I repaired 3 of the supplies that failed for the above reason for testing purposes with properly rated parts. I had 5 amp fuses but there was no availability of 6amp fuses. Therefore, A/C fusing was 5amps.
Power supply wattage/amperage: (claimed as maximums)
3,3volts-18amps
5volts(+)-25amps
12volts(+)-9amps
Standby (5v standby)-1.50 amps.
-5/-12 volts- 350ma (each)
No mention of maximum wattage of combined 5volt/3,3 volt.
PCB thickness-1.6mil
PCB support- 3 points via screws, one point via a nylon standoff.
When the repaired supply was bench tested at 25%, 50%,75%, 100%, 120% of rated output, the outputs remained within 4% of nominal except at 120%. At 120%, the line fuse failed. (remember the fuse was below specified capability)
The outputs remained quite clean from 25% upto and exceeding 75%.
The outputs were still clean beyond 75% but "switching noise" increased somewhat. (this is normal really, it just means the supply was working quite hard)
The supply provided clean outputs and started properly when the line voltage was between 92volts and 125 volts. (average)
This supply was intolerant of minimal loads. The supply failed to "start" below 15% of rated power. At 20% of rated power, the outputs were 7% of nominal ratings and noise/ripple increased markedly. The power supply was in its element at loads of and exceeding 25%. The supply ran cool above 25% but ran noticably warmer at 20%. The supply was set to 100% load and left to run for 6 hours.
At the close of 6 hours of operation, the supply was running reasonably cool considering the output developed. The power supply tolerated full power from a cold start. (this is good)
The power supply output cabling was too light IMHO but was adaquate at 20guage.
This supply would have faired better were the output cable size were 18g or better.
The supply cooling was adaquate but easily improved. Various wiring interfeared with ventilation. When this wiring was moved and secured, the noise level of the supply dropped and the fan became more efficient.
Bottom line and recommendations:
This supply "could" have been a desent if the maker made a few changes. The changes should't have had an impact of production costs. I surmise the supply may have cost and additional 0.25$ with the correctly rated components.
These components are: Rectifier bridge assembly, 2 filter caps and a fan.
When repaired, the supply easily met its claimed wattage.
Users may wish to choose a different supply until these anomolies are corrected.
The supply never "smoked" and the line fuse was properly sized.
When the components failed, they most often shorted blowing the mains fuse.
I do NOT recommend this supply, especially on 220volt mains where the internals would be grossly under-rated.
Upcoming: (so far the SPI/sparkle looks the best)
SPI (sparkle)
Antec
Enlight