Inexpensive Chinese bench supply kit

[Mephitus]

The kit in question(link)



I have seen this kit referenced once or twice on the board. While going with something like this deprives the need for a newb, like myself, to wade through the process of designing one. The low cost for such a fundamental tool is too good to pass up. (All my "fun" money I cant help but spend on my son, hehe) A quick glance at the kit makes it obvious to me that it is begging to be modified/upgraded. Hooking in a few panel meters, banana plugs for output, proper grounding connection, and jumper connections on the pots is a no brainer. As is adding a fuze and replacing the single turn pots with multiturn for precision control.

I am not so confident in my other observations. (Or maybe I am just overthinking a bit ) So any guidance is greatly appreciated. For example: I may be reading the board wrong, but is it missing output protection diodes? Wouldn't it be a good idea to replace the 1/4w resistors with higher wattage equivalents for better heat tolerance and stability? Adding an additional filter cap in parallel (and replacing it with a real brand)? Adding a thermistor for overheating protection? Separate switch to turn output on/off?

Even with the proposed modifications, it seems to me, would make a very nice and compact bench supply if put in an enclosure. Anyone think maybe there are other mods/upgrades that would be appropriate that I overlooked?

[MapleLeaf]

I bought this kit but haven't put it together yet. Going through the components, they all looked good until i got to the electrolytic capacitors. They are an off brand. I measured their ESR and according to my chart for good / bad electrolytic capacitors, they were bad. So throw them away and get some good ones would be my suggestion. Blowing up electrolytic is messy.

[GreyWoolfe]

Wilfred, I have to ask, have you ever had the pleasure of an electrolytic cap blow up on you.  That was a fairly common occurrence during my ITT Technical Institute days.   On the rare occasion it happens here at home, SWMBO gets seriously pissed at me because the house seems to smell for a full day before it dissipates.  The worst was testing a couple of plugins for an old Tek tube scope someone gave me.  One of the spare plugins erupted in fire (small flame) and the house smelled for 3 days.  I was lucky she didn't make me sleep in the shed.  She did make me put the scope and plugins on the curb for trash pickup.

I have learned not to use anything but brand name capacitors.  The company I work for was caught up in the Dell/Capxon debacle years ago.  Besides Dell paying us to replace 2500 motherboards and extending the extended warranty an extra year,  I fixed a couple of hundred 15" LCD monitors that were out of warranty before the equipment refresh because of blown Capxon caps on the power supply board.  I ordered in a bunch of Panasonic replacements and there wasn't any issues after that.  Up until recently, I still had a couple of them as spares.

[Mephitus]

I would also want to replace the caps with good brand name ones since, from what i always see veterans post online, stability is the #1 most important feature of a bench supply. Besides, if a little extra effort now gives me a tool that is reliable for many years, it saves me from having to replace it again in the future.

Sent from my SM-G920P using Tapatalk

[Shock]

You could always sub out those bad caps for resistors with long legs, same thing

[KuchateK]

This is kit version of very popular design from electronics-labs. Has some problems, but generally not bad. Especially for the price.

http://www.electronics-lab.com/project/0-30-vdc-stabilized-power-supply-with-current-control-0-002-3-a/

[Mephitus]

This is kit version of very popular design from electronics-labs. Has some problems, but generally not bad. Especially for the price.

http://www.electronics-lab.com/project/0-30-vdc-stabilized-power-supply-with-current-control-0-002-3-a/

Thanks for that find, makes understanding its operation much clearer. Guess im going to buy it when I next empty the spare change jar. Do you think any of the changes I proposed are good or bad?

[Cliff Matthews]

The most needed upgrade seems to be fine adjustment. I'd parallel-wire the wipers to another 20k pot with 47k resistors linking the outside contacts. AFAIK, this should add about 4% (+-) fine adjustment without PCB changes. The kit can be bought for less than $10 on ebay too: http://www.ebay.com/sch/i.html?_nkw=Continuously+Adjustable+DC+Regulated+Power+Supply+DIY+Kit&_sop=15

Also missing is metering: http://www.ebay.com/itm/DC-100V-10A-Voltmeter-Ammeter-Blue-Red-LED-Dual-Digital-Volt-Amp-Meter-Gauge-/131577787046?hash=item1ea2a5b2a6:g:lFcAAOSwMmBVzF8~

[Shock]

Thanks for that find, makes understanding its operation much clearer. Guess im going to buy it when I next empty the spare change jar. Do you think any of the changes I proposed are good or bad?

$10 for the kit, $20 for 2 decent 10 turn pots, $5-10 for digital display.

Unless you're making several channels in one enclosure by the time you finish it you might as well brought a Gopher 32V 5A supply for $60 delivered on Ebay (or one of the various beige boxes for $50). But if you want to DIY or upgrade an existing supply it could work out cheap.

Even at the cost of a Gopher you're starting to get into decent secondhand supply territory (if you're patient and hunt around).

[Cliff Matthews]

Voltlog's source link is old - this one works: http://www.electronics-lab.com/project/0-30v-laboratory-power-supply/
Edit: I don't know if the kits correct this, but comments say running TL081's and expecting final output above 29v seems to present an issue.. TL081 max supply is +(-) 18v, so a 22v transformer seems a tad safer (or bodge-in a few jelly-bean rectifiers just after the bridge + output for a 24VAC transformer). See the electronics-lab comments.. this all seems familiar.

Edit2: Yeah, there was a thread on this circuit back in April: https://www.eevblog.com/forum/beginners/diy-bench-power-supply/msg652480/ So although the sale info on this kit says 0-30v, you should feed it less and expect 28v max. The changed circuit (still using TL08x's) is here: http://diyfan.blogspot.no/2013/03/adjustable-lab-power-supply-take-two.html

[jdraughn]

I bought one of those, put it together - it worked great. So I put it on old KVM enclosure with a beefy transformer - and I accidentally wired the transformer backwards and put almost 700 volts into it. So I bought a second one that I could use to fix the first one - I figured I could add a second output to my little power supply But the second one didn't work. I troubleshooted it and figured it was the transistor on the very right, right under the output. I used the transistor from my first one to repair the second one. Now I need a replacement transistor for the first one, plus whatever else fried when I hooked up the transformer wrong.

[Kleinstein]

The output transistor is likly not powerful enough (SOA limit) to safely allow for the full 3 A at 30 V (or close). So if the transformer is more than about 15 V AC (thus about 16 V max output) one should reduce the maximum current a little.
The filter cap is also not really large enough for 3 A.

Also using a 24 V transformer max result in too much voltage for the OPs (TL081). So 20-22 V should be about maximum. Also the voltage rating of the filter cap is not high enough for a 24 V transformer.

So it's a resonable regulator up to about 25 V 2 A.

[rdl]

...
So it's a reasonable regulator up to about 25 V 2 A.

Exactly. 20 volts 1 amp would be even better.

BTW, this project has been around for well over 10 years and I think the op amp originally specified was OPA445, very expensive but rated for +/- 45 v input.

[Kleinstein]

There were several other higher supply OPs used in this type of circuit. The TL81 is rather limited, but cheap. The OP sets the upper limit for the voltage. One might gain a little, by limiting the voltage at the OP (e.g. resistor and zener) and possibly using separate diodes and a capacitor to reduce ripple, so one could get a slightly higher output voltage from a given transformer.

The current is limited by the SOA: in case of short circuit the transistors sees the full raw voltage (e.g. 33 V) and maximum current. Depending on the cooling one additional has to go down from the typical 25 C curve found in datasheets.

[Mephitus]

$10 for the kit, $20 for 2 decent 10 turn pots, $5-10 for digital display.

Unless you're making several channels in one enclosure by the time you finish it you might as well brought a Gopher 32V 5A supply for $60 delivered on Ebay (or one of the various beige boxes for $50). But if you want to DIY or upgrade an existing supply it could work out cheap.

Even at the cost of a Gopher you're starting to get into decent secondhand supply territory (if you're patient and hunt around).

I had some el-cheapo dmm's I was going to hack apart for the displays. Though I guess you are right for the comparable cost, and value for my time, I may be able to find a used equivalent. I had figured this would just be a fun, and useful, project. I also have some diagrams on hand for a pic controlled supply I might want to build eventually. But thats a ways off.

[rdl]

I would go ahead and build the kit. The other stuff doesn't have to be all that expensive. I bought 3 used Bourns 10 turn pots for $5.95 a couple of years ago and some cheap LED meters for 99 cents that read within 3 mV of my Fluke 87-III

[Gurpreet]

What rating of transformer should I buy to power up this PSU?

how about this transformer? http://www.digikey.ca/product-detail/en/VPS24-3300/237-1276-ND/666162

[Cliff Matthews]

What rating of transformer should I buy to power up this PSU?

how about this transformer? http://www.digikey.ca/product-detail/en/VPS24-3300/237-1276-ND/666162

The design has been discussed already. 20-22vac would be safer, at 24 (~33vdc across filter caps) it may toast the Chinese TL081's. So it's really a 26vdc supply.. but for this kit, who would complain?