Power-One power supply mod

[bushtech]

 

[bushtech]

I'm now rather worried. Is there any way that I can check the specs?

 

[Audioguru]

I'm now rather worried. Is there any way that I can check the specs?

If you cannot phone the manufacturer to ask about the 120VA or 240VA rating then find a torroid transformer in a catalog online with the 240VA rating and compare its weight with the one you got.

 

[bushtech]

lol, and there I was thinking that was urban legend, but I had just weighed it at 1.35kg/2.97lb

 

[bushtech]

Unfortunately manufacturer sleeping currently

 

[bushtech]

Here's one that's pretty close: http://uk.rs-online.com/web/p/toroidal-transformers/6719139/

But it's not clear to me if the 3.33A is per channel or total for the 2 channels

 

[kellys_eye]

That's 3.33A per winding (secondary) - but only 18V. If they were wound for 28V then the current would fall to 2.1A

A typical 240VA transformer will weigh around 3kg.

Your transformer has a THIRD winding....???

 

[bushtech]

kellys_eye: No third winding, I think it's just a standardised label

 

[Audioguru]

Yes, the transformer you got is too puny.

 

[hevans1944]

I asked the manufacturer to give me 2 secondaries at 28V/120VA each as I'm planning to eventually end up with a power supply with + and - rails. Is your impression that this is not the case? Then I need to check with the manufacturer

The markings on the transformer indicate this is not the case. The total power is 120 VA shared by the two secondary windings. At 28 VAC for each winding, that works out to about 120 / 28 = 4.286 A shared across two windings, or about 2.15 A per winding. That's a pretty far cry from the 3 A per winding you were shooting for. Or are you still hoping to push this design to 5 A per output?

Why not just buy another transformer and (essentially) build two power supplies? If you need 5A at 30 V DC, that's 150 watts, so the transformer you purchased is a bit on the wimpy side for that load. OTOH, 3 A at 30 V DC is only 90 watts, so your 120 VA transformer should be adequate for that... unless you want both positive and negative power supply outputs with respect to a common, in which case you need two of those transformers.

I went back through this thread looking for an actual power supply design but failed to find one. I did find lots of stuff on the Internet with regard to the so-called Hiland power supply design, including numerous "hacks" of the Banggood product offering. One of the more interesting is described in a series of blog posts by a Norwegian hobbyist who goes by the name paulv. You can follow him starting here: http://www.paulvdiyblogs.net/2015/05/tuning-030v-dc-with-03a-psu-diy-kit.html

But what I am wondering is: are you really learning anything from this thread? And are you any closer to constructing a "laboratory bench power supply" for use in your hobby? I do look forward to you posting pictures of your progress, but truth be known: power supply design is not exactly rocket science until you begin pushing the limits of adjustability, maximum voltage, and maximum current simultaneously. Adding "bells and whistles" in the form of metering, transient protection of the output when power fails (or surges!) and automatic fan control of the heatsink temperature are all interesting and, perhaps, valuable touches to incorporate into the final design.

However, I think most hobbyists would be satisfied with a pair of three-terminal voltage regulators (for positive and negative outputs with respect to a common terminal), a step-down center-tapped secondary winding providing 18-0-18 VAC (plus or minus a few volts) at a few amperes, a generic bridge rectifier capable of a few amperes with a few dozen (or a few hundred) peak inverse voltage rating, and a largeish electrolytic "smoothing" capacitor of ten thousand microfarads or so. This is all you need to play around with op-amp and other analog circuits. If you want to add some digital capability and resurrect some of those TTL parts gathering dust in your junque box, then add another low-voltage secondary winding, bridge rectifier, smoothing capacitor, and a 5V three-terminal regulator just for that. Or purchase some wall-wart power supplies.

If this were my project, I would be thinking of other ways I might accomplish my goal. Linear regulation may not be "dead" yet, but it is definitely on the way out. The old complaints about switching regulators introducing too much noise on the power supply rails, are pretty much addressed by proper circuit design and higher, more easily filtered and suppressed, switching frequencies. And there is a humongous reservoir of tried-and-tested circuit designs available on the Internet, along with application-specific integrated circuits to do the actual heavy lifting. Just add a MOSFET switcher and an inductor to finish the design.

But it's NOT my project... so onward to bigger and (hopefully) more suitable transformers.

 

[bushtech]

Thanks hevans1944. Wise words as always. I'm only shooting for 3A.
I am learning quite a lot already. I have started putting this circuit in LTSpice so I can see if the simulation matches what I read on the pcb. Lol, and I have also learned something about speccing transformers and how difficult it is sourcing parts on the sharp end of Africa.

Well, I have mailed the manufacturer, we'll see what they say when they wake up in an hour or so.

 

[bushtech]

Well, you guys were correct. Manufacturer states the 120Va is shared by the secondary's Some serious negotiation coming up.

 

[bushtech]

It's so nice when you come across honourable people these days. Manufacturer will take back old transformer, give me a 240Va version and I pay the difference.

 

[kellys_eye]

Great to hear - well done them.

Adding to @hevans1944 comment, one of my most reliable and well-used power supplies was an LM317/LM337 pair delivering a variable 0-18V output that satisfied 99% of all my project requirements for years.

I'd still be using it if I hadn't fallen for the 'bigger-is-better' trend and moved on to CC/CV devices.

 

[Audioguru]

Hevans, for 30VDC at 3A the output power is 90W but the AC from the transformer is about 119VA because the bridge rectifier heats with its voltage drop and the output transistors are powered from the rectified peak voltage of the 28VAC which is (28V x 1.414)= 39.6V - 2V (the drop from the bridge rectifier) = 37.6VDC. Then the transformer load is 39.6V x 3A= 118.8VA.

 

[hevans1944]

Yeah, I tend to ignore the losses in the bridge rectifier and the series pass transistors, folding them into about a
+20% "fudge factor" increase in transformer VA rating above that actually required by the load. So, by that criteria, I would pick a VA rating of at least 90 + 18 = 108 VA. If you figure in the power lost in the pass transistors to regulate down from 37.6 V DC to 30 V DC at 3 A, that's about 23 watts you need there, plus another 6 watts for power lost in the forward current through the diode bridge for a total of 29 watts, so in this instance +20% isn't quite enough and 112 VA is a more reasonable estimate. I don't feel comfortable using the peak rectified voltage of 39.6 V, multiplied by the regulated current of 3 A, as a measure of the transformer VA requirement, although it does tend to yield a conservative estimate. In any event, it appears that the 120 VA rating the OP got for his "custom wound" transformer is adequate for one 30 V DC power supply providing 3 A DC current.

 

[kellys_eye]

Can't recall the last time I used more than 1A on any circuit I built/tested recently.

I repair a lot of marine electronics and TV's and rely on dedicated power supplies for the marine (12V) and the (hopefully working!) PSU in the TV's.

I'd have to check to see if my existing variable PSU (max 3A) can still 'wind' up past 1.1A!!!! And, if I think about it carefully, it probably didn't reach 1A for most of the last couple of years either.....

Might be simpler and cheaper to just roll back your expectations a little and build with what you've got. If you use the appropriately rated parts you can always change the transformer at some later date to increase the output current if you ever find it necessary.

But if the transformer manufacturer is going to replace the under rated device anyway then.... meh.

 

[bushtech]

There is some fantastic info being shared here by the gurus. But I think I must start a new thread about my pending build so at least the info fits in with the heading.

 

[kellys_eye]

The published mods (as advised at various intervals throughout this thread) are quite concise and easily made - turning the original 'defective' design into something really quite useful.

A new thread dedicated to its construction - start-to-finish - will prove popular. Give the thread the correct heading and Google will point 'thousands' of people here that have already purchased one and found it lacking in its original guise.

I look forward to reading and contributing.

 

[bushtech]

New thread started here: https://www.electronicspoint.com/th...ioguru-psu-build-down-the-rabbit-hole.286096/