Last weekend, I posted a skimpy CWT unit here:
https://www.badcaps.net/forum/showthread.php?t=39222
… and it wasn't a very well built PSU, to say the least. Today, I have another Channel Well PSU from their ISO line. I actually found this unit a while back (maybe 2008 or 2009), but never got myself to post it here. I don't think it ever saw any use as there was absolutely no dust inside it or outside. Manufacturing date on the warranty sticker suggests this PSU was manufactured April 2005.
As usual, I'll start with the case pictures and label:
https://www.badcaps.net/forum/attach...1&d=1407035562
https://www.badcaps.net/forum/attach...1&d=1407035562
Familiar sight, isn't it? It’s made by… wait, L&C?? (a.k.a. Deer, a.k.a. Solytech?) That's what the UL number (E164554) points to, at least (or did a few years ago – now you get no good info from the UL database). The model number, LC-A420ATX, also looks like something L&C would use. Hmm
. Let's see some inside pictures.
Top removed:
https://www.badcaps.net/forum/attach...1&d=1407035562
It seems pretty roomy in there, thanks to the lack of guts. Three different transformers, so we have us another half-bridge design here. Main transformer is typical size 33 with center tap – something that almost all cheap PSU use. BJT drive traffo is size 19 (supposedly, but not really), and the 2-transistor 5VSB uses a 19 as well (this one is tall and proper size for a 19). There is also a toroid next to the middle transformer, so that means this unit has a mag-amp regulated 3.3V rail – quite an improvement from what we saw in the last CWT PSU above.
Another notable cost-cutting measure is the wires on the input: everything is rated for 300V and 20 AWG, just like the other CWT unit. Good power supplies use 18 AWG, 600V rated wire
Primary side (input):
https://www.badcaps.net/forum/attach...1&d=1407035562
https://www.badcaps.net/forum/attach...1&d=1407035562
A-ha, so it is a Channel Well (CWT) power supply, after all! How we know: look where the missing input filter coil is (and don’t get shocked that it is missing
– it’s a cheap PSU). Now look at some of the text next to it.
“T6.3AH/250V FOR ISO-350,400,450…”
Whenever you see that text, you are dealing with Channel Well’s ISO line of power supplies. They can make good PSUs or they can make pretty gutless units. Here, it looks like the latter, and perhaps CWT tried to cover it up by disguising themselves as L&C. Nice try, Channel Well.
Aside from the complete lack of input filter, CWT cut many other corners on this unit.
Bridge rectifier consists of four discrete RL205, 2 Amp-rated diodes.
Primary caps are MK 330 uF, 200V, 85C, 22x30 mm. I forgot to measure their capacity when I had an ESR meter borrowed, but I’m sure it is much lower, just like it was on the other CWT unit.
Primary switchers for the main PS are a pair of TO-220 cased TT2146 NPN BJTs (possibly by Sanyo), which according to chipcatalog.com are very similar to the 13007’s we normally see in most cheap PSUs (so that tells me, output should be limited to 250W or less continuous).
The BJT drive circuit for these transistors has a pair of 10 uF, 50V, Fuhjyyu TN or TNR capacitors.
5VSB has a 2SC5353 NPN BJT in an isolated TO-220 case.
Moving on…
Secondary side (output):
https://www.badcaps.net/forum/attach...1&d=1407035562
If the ISO marks on the board, QC sticker on the case, heatsinks, Rulian Science fan, and the case itself weren’t a giveaway that this is an ISO PSU, then the physical layout of the PWM controller (TL494) and supervisory IC (TPS3510P) on the secondary should be. The TPS3510P takes care of all of the standard protections like SCP, OVP, and UVP. As far as I know, there are no over-power (OPP) or over-current (OCP) protections, though.
3.3V rail has 1x 10V, 1000 uF, 10x16 mm OST RLP cap before a PI coil and 1x 6.3V, 2200 uF, 10x20 mm Fuhjyyu TNR after it. Rectifier is rated for 15A and 45V (an STPS1545CT). Minimum load resistor on this rail is 15 Ohms, which is a bit low (I like to see less than 0.5W dissipation, and here we have close to 0.8W).
5V rail also has 2x 10V, 470 uF, 8x12 mm OST RLG caps and PI coil bypassed with jumpers. Rectifier is rated for 15A and 45V (an STPS1545CT). Load resistor on this rail is 150 Ohms, which is appropriately picked.
12V rail has a single 16V, 680 uF, 8x16 mm Fuhjyyu TMR cap, though there is an empty spot on the board for a second 8mm cap. PI coil is bypassed again. Rectifier is rated for 10A and 200V (an STPR1020CT). Load resistor on this rail is 470 Ohms.
Both the -12V and -5V rails have only a single 16V, 220 uF, cap each (Fuhjyyu TN for the -12V and Teapo SEK for the -5V). And both are rectified by a pair of 1.5A (FR153) diodes. Load resistors are 470 Ohms and 150 Ohms for -12V and -5V rails, respectively.
5VSB has 2x 10V, 680 uF, 8x16 mm Fuhjyyu TMR caps. This time, there is a PI coil installed between them. Output rectifier is a 2A diode (SB2x0). The minimum load resistor is a 100 Ohm (thus only wasting 0.25W).
With these capacitors and rectifiers, it is obvious that the label is a complete lie. 3 Amps for the 5VSB with a 2A rectifier? And come on, 40A on the 5V rail?? No way! Of course, there is no combined rating for the 3.3 and 5V rails, and no combined rating for the overall power either. Please DO NOT expect 420W from this unit - not without fireworks anyways.
Finally, here’s a shot of the solder side:
https://www.badcaps.net/forum/attach...1&d=1407035562
Overall soldering quality is decent. It only gets blobby near the output wires. Speaking of output wires, they are all 20 AWG, short, and there aren’t many output connectors – just standard 20-pin ATX, P4 12V connector, four peripheral, and two floppy.
5VSB section:
Like with the other CWT, I decided to take some time to draw the 5VSB circuit (also helps me get my mind back in circuits, which I will need when I start classes in the Fall again).
The circuit itself is similar 2-transistor design, though not exactly the same as the other CWT PSU above. Looks like Channel Well had more components in this one. Not sure if that helps with anything or not. I’ve loaded the 5VSB with 1.2A before for a fairly long amount of time, and nothing overheated. The 5VSB was pretty stable too.
As stated, the 5VSB main transistor is a 2SC5353 NPN BJT, driven by a H945 (also a NPN BJT).
There IS a “critical” capacitor in this 5VSB circuit, though it’s only a 3.3 uF cap. Probably too little capacitance to cause major over-voltage if it fails, but I still recommend replacing it if you have one of these “better-built” CWT units that you’d like to save.
Here’s the circuit:
https://www.badcaps.net/forum/attach...1&d=1407035562
I didn’t do as thorough of an analysis as I did with the other CWT. But with the same incandescent bulb as a load and also without any load, the PSU appeared to have a similar efficiency to the other CWT.
--------------------
All in all, this PSU appears to be well-designed, but just lacking in components. I have used it to power many low-voltage projects on my breadboard. The voltage output has always been stable and close to the stated values. When I get time on my hands (after posting a few more of my crappy and no-so-crappy PSUs), I might end up upgrading this power supply and actually use it in a low-power computer. It does have some potential. A low-power 12V-based system will likely be no problem at all. The 2-transistor 5VSB should be okay with good caps.
https://www.badcaps.net/forum/showthread.php?t=39222
… and it wasn't a very well built PSU, to say the least. Today, I have another Channel Well PSU from their ISO line. I actually found this unit a while back (maybe 2008 or 2009), but never got myself to post it here. I don't think it ever saw any use as there was absolutely no dust inside it or outside. Manufacturing date on the warranty sticker suggests this PSU was manufactured April 2005.
As usual, I'll start with the case pictures and label:
https://www.badcaps.net/forum/attach...1&d=1407035562
https://www.badcaps.net/forum/attach...1&d=1407035562
Familiar sight, isn't it? It’s made by… wait, L&C?? (a.k.a. Deer, a.k.a. Solytech?) That's what the UL number (E164554) points to, at least (or did a few years ago – now you get no good info from the UL database). The model number, LC-A420ATX, also looks like something L&C would use. Hmm
. Let's see some inside pictures.Top removed:
https://www.badcaps.net/forum/attach...1&d=1407035562
It seems pretty roomy in there, thanks to the lack of guts. Three different transformers, so we have us another half-bridge design here. Main transformer is typical size 33 with center tap – something that almost all cheap PSU use. BJT drive traffo is size 19 (supposedly, but not really), and the 2-transistor 5VSB uses a 19 as well (this one is tall and proper size for a 19). There is also a toroid next to the middle transformer, so that means this unit has a mag-amp regulated 3.3V rail – quite an improvement from what we saw in the last CWT PSU above.
Another notable cost-cutting measure is the wires on the input: everything is rated for 300V and 20 AWG, just like the other CWT unit. Good power supplies use 18 AWG, 600V rated wire
Primary side (input):
https://www.badcaps.net/forum/attach...1&d=1407035562
https://www.badcaps.net/forum/attach...1&d=1407035562
A-ha, so it is a Channel Well (CWT) power supply, after all! How we know: look where the missing input filter coil is (and don’t get shocked that it is missing
– it’s a cheap PSU). Now look at some of the text next to it.“T6.3AH/250V FOR ISO-350,400,450…”
Whenever you see that text, you are dealing with Channel Well’s ISO line of power supplies. They can make good PSUs or they can make pretty gutless units. Here, it looks like the latter, and perhaps CWT tried to cover it up by disguising themselves as L&C. Nice try, Channel Well.
Aside from the complete lack of input filter, CWT cut many other corners on this unit.
Bridge rectifier consists of four discrete RL205, 2 Amp-rated diodes.
Primary caps are MK 330 uF, 200V, 85C, 22x30 mm. I forgot to measure their capacity when I had an ESR meter borrowed, but I’m sure it is much lower, just like it was on the other CWT unit.
Primary switchers for the main PS are a pair of TO-220 cased TT2146 NPN BJTs (possibly by Sanyo), which according to chipcatalog.com are very similar to the 13007’s we normally see in most cheap PSUs (so that tells me, output should be limited to 250W or less continuous).
The BJT drive circuit for these transistors has a pair of 10 uF, 50V, Fuhjyyu TN or TNR capacitors.
5VSB has a 2SC5353 NPN BJT in an isolated TO-220 case.
Moving on…
Secondary side (output):
https://www.badcaps.net/forum/attach...1&d=1407035562
If the ISO marks on the board, QC sticker on the case, heatsinks, Rulian Science fan, and the case itself weren’t a giveaway that this is an ISO PSU, then the physical layout of the PWM controller (TL494) and supervisory IC (TPS3510P) on the secondary should be. The TPS3510P takes care of all of the standard protections like SCP, OVP, and UVP. As far as I know, there are no over-power (OPP) or over-current (OCP) protections, though.
3.3V rail has 1x 10V, 1000 uF, 10x16 mm OST RLP cap before a PI coil and 1x 6.3V, 2200 uF, 10x20 mm Fuhjyyu TNR after it. Rectifier is rated for 15A and 45V (an STPS1545CT). Minimum load resistor on this rail is 15 Ohms, which is a bit low (I like to see less than 0.5W dissipation, and here we have close to 0.8W).
5V rail also has 2x 10V, 470 uF, 8x12 mm OST RLG caps and PI coil bypassed with jumpers. Rectifier is rated for 15A and 45V (an STPS1545CT). Load resistor on this rail is 150 Ohms, which is appropriately picked.
12V rail has a single 16V, 680 uF, 8x16 mm Fuhjyyu TMR cap, though there is an empty spot on the board for a second 8mm cap. PI coil is bypassed again. Rectifier is rated for 10A and 200V (an STPR1020CT). Load resistor on this rail is 470 Ohms.
Both the -12V and -5V rails have only a single 16V, 220 uF, cap each (Fuhjyyu TN for the -12V and Teapo SEK for the -5V). And both are rectified by a pair of 1.5A (FR153) diodes. Load resistors are 470 Ohms and 150 Ohms for -12V and -5V rails, respectively.
5VSB has 2x 10V, 680 uF, 8x16 mm Fuhjyyu TMR caps. This time, there is a PI coil installed between them. Output rectifier is a 2A diode (SB2x0). The minimum load resistor is a 100 Ohm (thus only wasting 0.25W).
With these capacitors and rectifiers, it is obvious that the label is a complete lie. 3 Amps for the 5VSB with a 2A rectifier? And come on, 40A on the 5V rail?? No way! Of course, there is no combined rating for the 3.3 and 5V rails, and no combined rating for the overall power either. Please DO NOT expect 420W from this unit - not without fireworks anyways.
Finally, here’s a shot of the solder side:
https://www.badcaps.net/forum/attach...1&d=1407035562
Overall soldering quality is decent. It only gets blobby near the output wires. Speaking of output wires, they are all 20 AWG, short, and there aren’t many output connectors – just standard 20-pin ATX, P4 12V connector, four peripheral, and two floppy.
5VSB section:
Like with the other CWT, I decided to take some time to draw the 5VSB circuit (also helps me get my mind back in circuits, which I will need when I start classes in the Fall again).
The circuit itself is similar 2-transistor design, though not exactly the same as the other CWT PSU above. Looks like Channel Well had more components in this one. Not sure if that helps with anything or not. I’ve loaded the 5VSB with 1.2A before for a fairly long amount of time, and nothing overheated. The 5VSB was pretty stable too.
As stated, the 5VSB main transistor is a 2SC5353 NPN BJT, driven by a H945 (also a NPN BJT).
There IS a “critical” capacitor in this 5VSB circuit, though it’s only a 3.3 uF cap. Probably too little capacitance to cause major over-voltage if it fails, but I still recommend replacing it if you have one of these “better-built” CWT units that you’d like to save.
Here’s the circuit:
https://www.badcaps.net/forum/attach...1&d=1407035562
I didn’t do as thorough of an analysis as I did with the other CWT. But with the same incandescent bulb as a load and also without any load, the PSU appeared to have a similar efficiency to the other CWT.
--------------------
All in all, this PSU appears to be well-designed, but just lacking in components. I have used it to power many low-voltage projects on my breadboard. The voltage output has always been stable and close to the stated values. When I get time on my hands (after posting a few more of my crappy and no-so-crappy PSUs), I might end up upgrading this power supply and actually use it in a low-power computer. It does have some potential. A low-power 12V-based system will likely be no problem at all. The 2-transistor 5VSB should be okay with good caps.
Attached Files
?)
Comment