Benchtop Power Supply Question

[momaka]

Originally Posted by jiroy

Ah , okay ..The DC third wire of Hp and Compaq mostly ..Well , if you remember Momaka , some years ago , I talked about my loss (app. 5000$) when I bought some laptops affected with the Nvidia problem . I talked also about how they refused to fund money back , and instead , to grow my frustration , they offered cheap compaq something with no Nvidia inside .

Ah yes.
Well, TBH, they were probably doing you a favor by not having anything nVidia inside.
nVidia = noVideo

... not that AMD/ATI is that much better off (though certain AMD stuff can be.)

I have an old HP DV6000 laptop with Intel onboard graphics. Many of these laptops were sold with "better" nVidia GPUs and/or chipsets. And ironically, all of those are either dead or almost dead now (have a Presario V6000 in that boat.) Meanwhile, my integrated Intel is still chugging along (slowly) just fine.

Originally Posted by jiroy

Eventually , I stopped buying ANY HPs ! . Same recommendations to friends and customers ..

I don't blame you for doing that. I myself usually try to avoid anything HP as well - at least anything made by them in the last 10 years. Every once in a while, they do have some decent offerings / designs, but rarely. And modern Dells have become just another equivalent to HP, IMO - i.e. not that great.

Originally Posted by jiroy

... one Toshiba 8GB DDR2 (Yes , believe it or not , unique ,and needs more posts later), so basically i'm in no need to a third wire .

Some (many?) old Toshiba laptops were really good (OK, excluding the one that had the failing Nec/TOKIN caps, but that's understandable.)

Originally Posted by jiroy

By the way , i laughed at (Sam x3) , lol ..

Don't know where/who I saw giving him that short name first, but it was some years back. And since then, it stuck in my mind. I hope Sam doesn't mind me using it.

Originally Posted by shovenose

Thing is, if for any reason it can't detect the wattage, it'll totally nerf your laptop. Either it won't charge the battery, or your CPU will be stuck at some stupid low clock speed and the computer will be unusable.

Yup.
In case of my uncle's E6400, it did both - no battery charging and CPU (high-end C2D) stuck at 600 MHz. LOL! The 9-cell battery was already quite worn when I got him that laptop, but he didn't use it with the battery much. Anyways, at some point the ID wire on the adapter failed some years back, making the battery not charge. With the battery dead and my uncle regularly forgetting to plug-in the laptop in the wall, eventually the CMOS battery died too. This then lead to all kinds of weird behavior. The final straw was that the HDD SATA settings in CMOS (RAID/ACHI/IDE mode) got messed up, sending the laptop in a boot loop. He brought it to one of those local "big" / known laptop repair shops, and they told him he needed a new motherboard for the equivalent of about $150 without labor. I said, no way! Got the laptop, and after some head-butting, found the issues. Replaced (jury-rigged) the CMOS battery and patched the broken wire in the adapter (ghettomod / worthless repair thread material coming some day soon ). Surprise, surprise, the laptop sprung back to life again and fully working like it should be. Gave him a spare charger to keep, in case the one I fixed goes bad again (probably will at some point, given how my nephews sometimes tangle/play with the cord.) And told him to forget about that shop

[jiroy]

Originally Posted by momaka

Ah yes.
Well, TBH, they were probably doing you a favor by not having anything nVidia inside.
nVidia = noVideo

... not that AMD/ATI is that much better off (though certain AMD stuff can be.)

I have an old HP DV6000 laptop with Intel onboard graphics. Many of these laptops were sold with "better" nVidia GPUs and/or chipsets. And ironically, all of those are either dead or almost dead now (have a Presario V6000 in that boat.) Meanwhile, my integrated Intel is still chugging along (slowly) just fine.


I don't blame you for doing that. I myself usually try to avoid anything HP as well - at least anything made by them in the last 10 years. Every once in a while, they do have some decent offerings / designs, but rarely. And modern Dells have become just another equivalent to HP, IMO - i.e. not that great.


Some (many?) old Toshiba laptops were really good (OK, excluding the one that had the failing Nec/TOKIN caps, but that's understandable.)



Don't know where/who I saw giving him that short name first, but it was some years back. And since then, it stuck in my mind. I hope Sam doesn't mind me using it.

Same thinking here , not much to add Momaka ; just an update on the project . Till now , I'm in favor of old school and i think it's more reliable than switching powers when comes high amperage . I mean ,I'll wind a transformer , the usual 4 diodes and big capacitors . It will be around 25 A. since the asus alone is 9.23 A .. I'll repost more when i start .

[eccerr0r]

Problem with linear PSUs is regulation and weight/expense...
regulation makes it inefficient, and weight/expense is always an issue.

Do you have all the parts for a line frequency 25A transformer already perhaps? That's a lot of thick wire to wind, not to mention a huge core.

[JC173]

Originally Posted by momaka

I'd say if you're just starting with learning electronics, 0-30V and up to 5-6 Amps should cover plenty of ground. If you plan on doing experiments with many "heavy duty" DC motors (i.e. anything size 540 and up), consider dropping the voltage requirement a little (down to 24V, perhaps) to get more current: 10-15A for such motors. Even then, that might not be too cheap, considering 10 Amps at 24V is already a 240 Watt PSU, and that can be expensive.

IMO, best way is to get several different PSUs, and not necessarily bench type. In fact, I *still* don't have a (working) bench PSU, despite experimenting with projects quite often. Majority of small breadboard projects I power with small plug-in power adapters (5V, 9V, 12V, 15V, etc.) And when I need more heavy current with low voltage, that's where my ATX PSUs come in. For higher voltage, I've used various contraptions - from non-isolated resistive mains droppers (light bulbs, soldering irons, countertop ovens, and etc.) to AC dimmer-controlled transformers.

It's certainly more cumbersome to do it my way, but it's also been cheaper (and I find it more rewarding, since I sometimes have to *actively* think about how I want to power something, rather than just dial in a number on a bench PSU and hope it all works fine.)

That said, there have been times where I did wish I bought or made an adjustable PSU. Most of these were when I needed low voltages (3.3V or less) at medium to high currents (more than 2-3 Amps) and good regulation. But again, I think 0-24V or 0-30V @ up to 3-5 Amps should work well enough for you to get into electronics. That along with a computer ATX PSU and a 60-90 Watt 19-20V laptop adapter should cover a good deal of projects.

This is a fantastic idea. I think I can scrounge an ATX supply. It'll give me a place to start and I should learn something in the process.

Any recommendations on a tutorial for the process?

[jiroy]

Originally Posted by eccerr0r

Problem with linear PSUs is regulation and weight/expense...
regulation makes it inefficient, and weight/expense is always an issue.

Do you have all the parts for a line frequency 25A transformer already perhaps? That's a lot of thick wire to wind, not to mention a huge core.

Frankly , eccerr0r; it's the opposite . just imagine 5 or 6 adapters plus cables and inlets on the same table running hamoc between laptops ; Actually , the sixth is now under the table , lol , so with only one power unit , it will gloriously return . Why ? , because other then the adapters and cables , it's critical in my mind to gain a one point direct view at most of them , while controlling the farthest with wireless mouses .
About the parts , it's the last of my worries . I have by kilos all kind of transistors , diodes , ceramic and millar capacitors . I made a robot 35 years ago , so I really doubt if I would need any in the end .. My parts arsenal is really huge .
I'll put an examplary picture or two tomorrow ..

[eccerr0r]

Have you fully specced out how to make that transformer for 25A? It will be HUGE. And that's without the regulation, how are you planning to regulate it, big bank of TO-3's becoming its own griddle?

I'd rather the switchmode PSUs and swap tips off the ends as one fail or is borrowed/machine moved elsewhere for other testing, at least I have 4 left...

[jiroy]

Originally Posted by eccerr0r

Have you fully specced out how to make that transformer for 25A? It will be HUGE. And that's without the regulation, how are you planning to regulate it, big bank of TO-3's becoming its own griddle?

I'd rather the switchmode PSUs and swap tips off the ends as one fail or is borrowed/machine moved elsewhere for other testing, at least I have 4 left...

It's still under study project , thousands of probable options . So far , I chose a housing , and a special Legrand switch anti-electrolyzing . Next , for the core ,
I have many 110v to 220v. big transformers , not used anymore since ten of years , as Lebanon's now is using 220v. 50Hz , so basically , the core is no problem , it just need to be modified accordingly to produce 20volts . The old school is already gaining here , and parts , money and patience are no problems .
Next steps are tens of options , still under study now .

[sam_sam_sam]

I am following this post to see what you come up with

[momaka]

Originally Posted by JC173

This is a fantastic idea. I think I can scrounge an ATX supply. It'll give me a place to start and I should learn something in the process.

Any recommendations on a tutorial for the process?

Back when I was getting into electronics, I've looked up a dozen number of tutorials on how to convert these to a bench PSU. But I never built one, because I just kept hooking/unhooking things every time.

That said, I haven't looked at what tutorials or guides currently exist for these PSUs. The ones I saw before all pushed on the idea of installing Banana plugs for the various rails. Some even added a cheap LCD voltage display for each rail. I think that's unnecessary, though, unless you already have Banana cables and all other standard bench equipment like that.

The simplest idea I've had... and I've thought about building this many times but just kept getting busy with other projects... is to just buy a cheap ATX power supply tester, preferably with an LCD readout like this:
https://www.ebay.com/itm/262966850006

Then, just open the ATX power supply tester and solder some heavy duty wires to the 3.3V, 5V, and 12V rails. Connect these wires to a breakout board or terminal posts. Like I said, I'm not a fan of banana jacks, simply because I just don't have any on my bench. However, I do have a lot of alligator clip wires... so my preference would be to just solder the heavy duty wires to a bunch of nails and punch those into a wooden board - cheap, ugly, but very effective. Of course, if you do want to make yours neater, you could certainly do that and get connectors or breakout boards that fit your needs. Heck, there's even this if you don't want to hack open an ATX PSU tester:
https://www.ebay.com/itm/234769449960

Note however, that the above breakout board costs the same as an ATX PSU tester, and the ATX PSU tester will have an LCD readout for the voltages. The readouts may not be too precise or accurate... but at least you will have some reference. Otherwise, I do like the idea of the breakout board above. It even has built-in fuses, in case you short something on your project. The bigger advantage is the On-Off switch, though. Older versions of the LCD PSU testers used to have one too, but looks like that's been removed now. Either way, if you do hack open an ATX PSU tester, it shouldn't be too hard to modify it so that you can add an external On-Off switch.

In any case, the reason I don't suggest directly converting an ATX PSU into a bench PSU is because not all ATX PSUs are suitable. For example, if you use a more modern high-power PSU (500 Watts or more) with a single 12V rail design (rated upwards of 30 Amps), you could get into a very risky situation if you happen to accidentally short something on your breadboard or project(s), especially if you're using thinner wire to provide these with power. Before you know it, that wire and/or your project/breadboard would be up in a cloud of smoke, possibly setting something on fire too. An ATX PSU with more modest outputs and good protections should shut down pretty quickly if you short-circuit something by accident.

Another reason is, if your ATX PSU goes bad, you can easily replace it with another and continue to have a working bench PSU. But if you build the bench PSU into the case of the ATX PSU, now you're locked into using that PSU only as a bench PSU.

This is why I suggest the ATX PSU tester route, because it can allow you to connect any PSU you desire. And it still doubles as an ATX PSU "tester" too (well, maybe if you provide a little load on the PSU and attach an oscilloscope to the outputs ). For starters, you might want to go with a more modest and/or older PSU - something in the range of 150-300 Watts and ideally something that has good short-circuit protections in case you have an accident with your projects. My recommendation for such is usually OEM PSUs - i.e. something that came with a Dell, HP, or similar. These are typically made by Delta, HiPro / Chicony, LiteOn, and Bestec, all of which are pretty good PSUs. If you do grab an older unit from these, you might find that it needs a recap... and that's OK. Get it recapped and use it as either a bench PSU or even to test PCs. They are pretty reliable after a recap.

Right now, these are dime-a-dozen online and they have pretty well-balanced current distribution as well as good protections:
https://www.ebay.com/itm/225301555762

And here is the slightly older models of the above PSU (these lack APFC, so they may be even more reliable as a bench PSU):
https://www.ebay.com/itm/325299804670

^ This one will definitely need a recap at some point, though. Otherwise, it's a really great PSU both for testing PCs or as a bench PSU. I've documented the build quality and recap on several of these on BCN. See below links.

HIPRO P3017F3 build quality:
https://www.badcaps.net/forum/showpo...postcount=3061

HIPRO P3017F3 recap:
https://www.badcaps.net/forum/showthread.php?t=81905

Of course, you're free to use whatever other ATX PSUs you like/find/scavenge. However, keep in mind that some of the really cheap ATX PSUs may not have very good short-circuit protections. So once you find an ATX PSU to use for this project, first turn it on and test the short-circuit protections. Start by bridging the 3.3V rail to ground and seeing if the PSU shuts down. If it does, remove the short on the 3.3V and do the same for the 5V and 12V rails. If the PSU shuts down for all 3 rails, you should be good to use it. If you get smoky, jelly, glowing wires from the PSU, then it's good you didn't use it as a bench PSU.

Lastly, if you find that you really do want to have an adjustable output (approximately 1.2V to about 9.5V @ max 2 amps), you can also easily add an LM317 module board kit to your bench PSU. These are about $5 on eBay (and probably elsewhere online.) Here is one such example:
https://www.ebay.com/itm/225321298871

^ If you do get one like that, ditch the included heatsink for the LM317 IC and instead mount a much larger heatsink from a CPU. This should allow you to get the full or almost full current output of the LM317 IC (i.e. 2 Amps) without worrying that you will overheat it (unless these cheap LM317 ICs have smaller dies - then you won't be able to get that much current out of them for sure.)

And that's how pretty much I would do it (and how I've been thinking about doing it) if I had the time to sit down and build one of these.

[momaka]

Originally Posted by eccerr0r

Problem with linear PSUs is regulation and weight/expense...
regulation makes it inefficient, and weight/expense is always an issue.

Agreed.

While I do love the idea of using a big bulky oldschool line transformer, those alone will probably use 5-10 Watts with the core sitting "idle" with no load on the output... which in many cases will be more than 5 separate SMPS adapters. Only when loaded more moderately you will see a good return on the efficiency (possibly up to 90% or more just for the transformer... though that will depend a lot on the transformer, of course.)

And then like you said, regulation will probably eat another 5-30 Watts or more if it's of the linear type.

Originally Posted by eccerr0r

Do you have all the parts for a line frequency 25A transformer already perhaps? That's a lot of thick wire to wind, not to mention a huge core.

Not that big.

Any good ol' 300-500 Watt UPS transformer should get you there.
Input: 120 or 230/240V AC.
Output: 12 to 14V AC typically... which translates to about 16 to 19V DC after rectification. So this may even be usable as-is without a regulator, provided the unloaded voltage doesn't go too high. I'd say anything above 23-25V is dangerous for most laptops, as some may only have 25V-rated parts on the input if the manufacturer cheaped out somewhere instead of using 35V or higher -rated parts.

Originally Posted by jiroy

Frankly , eccerr0r; it's the opposite . just imagine 5 or 6 adapters plus cables and inlets on the same table running hamoc between laptops ;

Hey, 5 or 6 is not bad at all!
Back when I used to work as a refurbisher/tester at Micro Center, sometimes I'd have laptops over laptops on my workbench, all doing something. Certain days, I'd have 10-20 machines - laptops and desktops mixed. Mess of wires? You don't say.
At home, the situation is more tame - I only have 6 desktops stacked on top of each other next to my desk, 2 in my foot area, and one the other side of the room with wire running back to my desk a-la Cyberpunk style.

[jiroy]

Originally Posted by sam_sam_sam

I am following this post to see what you come up with

Very good .. Just be patient , since I'm an ultra tasking man , with always tens of projects at hands , and not necessarily Electronics only ..
I'm right now building an over voltage timer , AC side of the trans , since Electricity in Lebanon is intermittent and we relay too much on local Electric motors providers . Things always get ugly if you don't take many safety precautions .

[jiroy]

Originally Posted by momaka

Agreed.

While I do love the idea of using a big bulky oldschool line transformer, those alone will probably use 5-10 Watts with the core sitting "idle" with no load on the output... which in many cases will be more than 5 separate SMPS adapters. Only when loaded more moderately you will see a good return on the efficiency (possibly up to 90% or more just for the transformer... though that will depend a lot on the transformer, of course.)

And then like you said, regulation will probably eat another 5-30 Watts or more if it's of the linear type.


Not that big.

Any good ol' 300-500 Watt UPS transformer should get you there.
Input: 120 or 230/240V AC.
Output: 12 to 14V AC typically... which translates to about 16 to 19V DC after rectification. So this may even be usable as-is without a regulator, provided the unloaded voltage doesn't go too high. I'd say anything above 23-25V is dangerous for most laptops, as some may only have 25V-rated parts on the input if the manufacturer cheaped out somewhere instead of using 35V or higher -rated parts.


Hey, 5 or 6 is not bad at all!
Back when I used to work as a refurbisher/tester at Micro Center, sometimes I'd have laptops over laptops on my workbench, all doing something. Certain days, I'd have 10-20 machines - laptops and desktops mixed. Mess of wires? You don't say.
At home, the situation is more tame - I only have 6 desktops stacked on top of each other next to my desk, 2 in my foot area, and one the other side of the room with wire running back to my desk a-la Cyberpunk style.

5 or 6 adapters , even on standby , will equal any dissipation power of the core , while on the long run , nothing can destroy a good built transformer when comparing to the low level design current adapters , and the same thing can be said when loading .. In the end , the Electricity bill will be high anyway , but when comes to damages and repairs , the old school is better and longtime reliable .

About desktops , they're in an another room , and another huge table . They're now mostly for Data storage and repairing tests purposes . One of them , in particular , contain app. 8000 movies (all legal) , plus some Half-Laptops !! , yes , That's what I call a Laptop which its repairs exceed the benefits of selling . Sometimes , a Laptop with a bad keyboard , defect battery , deteriorating screen , etc , becomes a burden to repair and sell .

I have another 40-50 quarter Laptops , lol , for parts donner purposes . Yep , lot of mess and I have to start reorganizing somehow ..

[jiroy]

Like I promised , some examples of my backup parts

[sam_sam_sam]

Originally Posted by jiroy

Like I promised , some examples of my backup parts

WOW How often do you used these reclaimed parts in your projects

[JC173]

I'm always pleased to learn way more than the simple answer to a question. These comments are great.

[jiroy]

Originally Posted by sam_sam_sam

WOW How often do you used these reclaimed parts in your projects

Those are never used parts from new not used devices , most of them given as bonuses on quantities sales or left overs defaults , or stocks . Those are surely the 0,001% of my arsenal , the ones that I didn't have time to arrange in their accordingly drawers and left on purpose at home for quick repairs .

If you were a projects designer , you would know that those are for repairs mainly , and more important , a precaution measure against an obsolete part . When I want to do a project , I prefer ordering a complete kit , like this one attached , mainly for time saving more than money saving . This kit was delivered 2 days ago .

[jiroy]

Originally Posted by JC173

I'm always pleased to learn way more than the simple answer to a question. These comments are great.

Thanks dude ; I was worried if you feel your thread was hijacked ..

[sam_sam_sam]

Originally Posted by jiroy

Those are never used parts from new not used devices , most of them given as bonuses on quantities sales or left overs defaults , or stocks . Those are surely the 0,001% of my arsenal , the ones that I didn't have time to arrange in their accordingly drawers and left on purpose at home for quick repairs .

If you were a projects designer , you would know that those are for repairs mainly , and more important , a precaution measure against an obsolete part . When I want to do a project , I prefer ordering a complete kit , like this one attached , mainly for time saving more than money saving . This kit was delivered 2 days ago .

Interesting

[JC173]

Originally Posted by jiroy

Thanks dude ; I was worried if you feel your thread was hijacked ..

Nope, not at all. It's all great info I'll use somewhere

[jiroy]

So far , I've finished the over-voltage timer , and looking to start designing the main board .
Till now I've decided on one winding DC part (low AC) with one tapped center , despite that two windings are better for noise and easier repair . Anyway , I won't wind it until I finished designing the whole circuit , including protection ones . It's the preferred method anyway to work it in reverse , since it will be decisive in core's calculation .
Right now , I'm studying the different options of diodes or rectifiers ..Attached some of the primary ideas , but nothing decisive yet ..