Please, advise on cheap and good variable power supply

[mariush]

So, I'm seriously considering purchasing a variable power supply.  I'm leaning towards a linear power supply because, while I'm a relatively beginner at electronics, I plan on designing my own variable power supply from a regular ATX psu in the future.

I'm reluctant to use eBay or various online websites to purchase due to the weight of such items, which would make shipping to where I live expensive (Romania).

Therefore, at the moment I'm planning to purchase one from Farnell, because the shipping only costs about 5 $/euro  and I can buy other parts from them.

The problem I have is that both power supplies that were attractive to me are quasi unknown, I can't find any review, any teardown, don't have a clue about them
 
I wanted one that's as cheap as possible so at first I considered buying this one :

CALTEK - PSD30/3B - PSU
 
http://uk.farnell.com/caltek/psd30-3b/psu-laboratory/dp/1270468

It's 30v - 3A single power supply, with digital display... at 86 pounds it looks like a good deal.

The problem with this one is that I could only find this page about other Caltek power supplies: http://www.rcgroups.com/forums/showthread.php?t=1030827

The guy  basically bought 2 power supplies broken from the factory and the faults were quality control related: a diode installed the other way on the pcb, dry smd joints, the switches causing shorts or sending 60v through the output...

Not sure if it helps, but apparently these guys at Caltek  are the OEMs for some Extech power supplies... this is one of their :

So on one hand price is very attractive but on the other hand I'm not really willing to roll the dice and try my luck with one.

----------------

Several days later, I was checking Farnell again and noticed this power supply:

TENMA - 72-8690 - POWER SUPPLY, BENCH, 2 X 32V, 3A

http://uk.farnell.com/tenma/72-8690/power-supply-bench-2-x-32v-3a/dp/1836056



Naturally, this is Farnell's brand... but I did some research and discovered it's a rebranded Uni-T UTP3702 : http://www.uni-trend.com/UTP3702.html 

There's another Tenma about 5-10 pounds more expensive that is the UTP3703 - same psu with fixed 5v/2A added. I don't care about that - if I buy this one i might just mod it to add the 5v/2a in the empty space.

What I like about it is that it's 2 independent channels for just a bit more price, about close to the maximum I'd pay for a variable power supply (~ 125 pounds).
Also, a major plus I feel is that one channel can track the other and so it would be great as a dual power supply, for amps and all that.
The manual also says that since they're floating outputs, it's possible to chain them and have a 64v 3A output, which may matter or not in what I'll do in the future..

Now my issue or concern with this is that again, there's no reviews, no information about Uni-T power supplies, I can't find any of these on eBay, can't find inside pictures etc . 

It's a Uni-T: I have the UT61E multimeter and I'm impressed with its quality but I'm aware other categories of devices they produce have varying quality so I'm reluctant to try my luck without any nformation about it.

.... Long text, sorry about it but thanks for reading so far.

I'd appreciate it if you guys can say your thoughts about this Tenma/Uni-T psu... if you have any info about it or other Uni-T power supplies in the series I'd love to hear it.  The Caltek is out... it doesn't seem worth it to even consider that Caltek, 2 x 32v for 30% more money seems a much better deal for me.

Also, please let me know if there's other brands you know or trust at around this price but do keep in mind the shipping costs.. those kill lots of deals for me. 

tl:dr: Anyone bought it, opened it, imported it, heard anything about these Uni-T variable power supplies marketed by Farnell under their Tenma brand?

ps. I've watched Mike's review of the Tenma psu but that's a switching power supply and it seemed ok. It's always a backup option.

Thank you for any help any of you can give me.

[Monkeh]

I've got no experience of those PSUs, but I will say this: If you buy one and want the fixed 5V output (believe me, you'll end up using it!), just spend the extra money and get it. You will not be able to hack a proper, 5V 2A circuit into it at any real saving.

[Fluxed Matter]

How about this one. It looks like your in the UK . I just ordered the US version. It is the cheapest I can find but I think it will work for most applications. Let me know what you think.

http://www.circuitspecialists.eu/bench-power-supply-060v-03a-or-030v-06a-5v-fixed-ouput-p-256.html

[mariush]

Thank you for the suggestion.

No, I'm in Romania, so at 12 Kg weight, the Circuit Specialists people will probably charge me 30-60 pounds just for shipping that here. IF they ship here... their eBay store won't offer it. That's why I love Farnell's flat rate... it's 20 RON (1$ = 3.7 RON approximately) no matter what I order.

The model does look nice, heavy due to the beefy transformer... but I'm wondering who the OEM is ... they don't do these themselves...  Yeah, a google search later reveals this is an ATTEN APS3005S-3D 30V 5Amp or a slight variation of it: http://www.atten.eu/index.php/power-supply/atten-aps3005s-3d30v-5amp.html  (picture is for the dual but description is for the single output one)
(later edit:) And here's the exact model in the Australian store TrioSmartCal : http://www.triosmartcal.com.au/dc-bench-supplies/2432-atten-aps3003s-3d-dual-power-supply-0-30v-3a.html


The teardowns of simpler models don't give me much confidence:

https://www.eevblog.com/forum/product-reviews-photos-and-discussion/atten-aps3005s-teardown/
https://www.eevblog.com/forum/product-reviews-photos-and-discussion/atten-aps3003s3005s-variable-power-supply-review-and-teardown/

Stuff like power supply outputting 30v for less than a second when pushing the On button and other quirks that show up randomly during a review is what I'm hoping to avoid.

I'll keep this model under consideration.

Anyone else, please feel free to suggest other models or provide feedback regarding the two power supplies in the first post, it's really appreciated

[Spikee]

The power supply (kit) or fully built from elv.de seems a good power supply and resnably priced + shippin is only 7 euro's IIRC in europe.
The ELV PPS 5330 3A 30V liniair power supply at 99 euro for the kit (smd stuff already on it) or like 119 already built.

I just got it in yesterday and it feels like a high quality power supply. The enclosure is from thick aluminum and the frontplate is really nice. I will post a review and a timelaps of building it tomorrow. And some measurements of stability etc and noise using my oscilliscope and dmm.

The customer service speaks good english and the have an repair service if you built the kit wrong or have broken it. The building instructions are in german but you can probably find the pdf online and put it in google translate. There are plenty of pictures so very basic idea of what you are dooing is only needed.

Link of building instructions:
http://www.elv-downloads.de/service/manuals/PPS5330/PPS5330_KM_G_080303.pdf

The kit:
http://www.elv.de/prozessor-netzteil-pps-5330-komplettbausatz.html

Fully built:
http://www.elv.de/prozessor-power-supply-pps-5330.html

[T4P]

1) the UNI-T is also a mastech clone, the tenma is a rebranded UNI-T
therefore it's also a clone!

[Achilles]

Well, the ELV PPS5330 is on my wish-list, too.
I have the SPS5630 (0-30V, 0.3-6A) as kit. It was partly hard to solder as I needed to clean up some soldering pads (didn't want to stick on it but could be cleaned with a glass fiber pen). Some parts were sent wrong but ELV answered very quick and sent me the missing/wrong parts.
The SPS5630 is quite nice and more or less the bigger brother of the PPS5330, but for electronics I would surely recommend the PPS5330 as it really goes down to 0V and the current can be fixed below 300mA. The SPS5630 has also a lot of ripple (measured about 170mVpp).
So if you are going with the ELV, be sure to take the smaller fellow and maybe already built as it is not much difference there...

[Fluxed Matter]

mariush thank you for the information. When the supply arrives I will try to do a "scratch n sniff" and get some pictures. I hate waiting, and UPS also take so long, but their shipping it to my work so maybe it will get here sooner.

[mariush]

Spikee:

The ELV is nice, but the problem I have with them is the use of proprietary parts like the LCD display and the fact that they're hiding some chips under their own name. 
I've seen it in other of their devices... basically it's no good that they're providing the datasheets if, for example, the microcontroller is ELV07670 and the lcd elements driver is ELV07671. 

They're hiding the name of the actual parts in an attempt to force you to buy the chips with them with the firmware burnt by them before shipping it to you, at a much higher price than the market price. Also, they used SMD versions precisely so that it would be harder for the average person to replace the ICs by themselves... there's really no need to go for such high pin ICs in a power supply.

I can't tell from the pictures that the lcd driver or the microcontroller is Samsung one, so if the mcu is Samsung that means it's probably an ARM core, which means I'd have to learn another platform if I have to debug or repair it.... if I'd be able to extract the firmware or other crap like that.
Why bother with all this when those other linear psus with led displays are straight forward, with simple logic and easy to fix?

I'm also just a tad bothered by the custom heatsink (round extruded one) with the puny noisy fan.

T4P: I'm not bothered if it's a Mastech clone... as long as it has the characteristics of some of the good Mastech units. There's good Mastech units and bad ones and a lot of clones... seen lots of reviews.
But there's no reviews out there for these Uni-T units, nothing...

[T4P]

Nobody's going to take the risk with the UNI-T sets ...

[poodyp]

This might be out of your price range or less powerful than what you want, but check ebay for HP/Agilent E3600 series power supplies. I got a 3610 for $75 because of some damage to the heatsink, but it still works perfectly.

[T4P]

Agilent is having a promo in some areas around the world for when you buy any E36* PSU you get a U1272A for free

[poorchava]

IMO unless you want to spend $1k on brand name like Agilent, you'll be best off building one on your own.

There are countless schematics on the net. There are some quite sophisticated out there. And of course when doing your own device, you can be sure of built quality and don't have to wonder "if chinese guys saved on isolating pads or not".

Actually I think a decent power supply is something that is still economic to build on your own.

[metalphreak]

I got a massive Zhaoxin dual output 30V 10A (per output) with a fixed 5V supply off ebay for a decent price (~$300? I think). Has been perfectly reliable so far. Out of all the chinese clones that look similar, these one's are never rock bottom prices, so I would assume they are some of the better ones (and they have a massive range of them too).

http://www.goodluckbuy.com/index.php?target=products&mode=search&subcats=Y&type=extended&avail=Y&pshort=Y&pfull=Y&pname=Y&pkeywords=Y&cid=0&q=zhaoxin&x=0&y=0

http://www.goodluckbuy.com/zhaoxin-rxn-305d-30v-5a-linear-adjustable-dc-power-supply-2ch-output.html This is the 5A version of mine.

Just another one to consider on your list. GLB have been reliable in the past for me. I supposed someone like Element14/Farnell would be easier to deal with for warranty though.

[mariush]

Thank you for the tip, metalfreak.

Well, I must say the Agilent offer was tempting, because I could potentially sell the multimeter for $200-250 and get a good brand power supply for close to my budget. The cheapest supplies are 1670 RON + vat ... so about 580$ or 360 pounds. However, I contacted the sales department of Farnell and they no longer take part in this promotion.

The search continues...  any suggestions of other power supplies and/or feedback about that Tenma (Uni-T rebrand) in the first post are still more than welcome (the price makes it very attractive to me).

[DarkPrince]

Please keep helping him! I am watching this thread closely... I had created a thread a while ago (months) with the same reason... still without a power supply because I am not happy with any of them. It is amazing at the number of choices and how every single one has at least one disadvantage. I know it is common but still.

Maybe buy a cheap one, rip the guts, and redesign the power management side of the system... interesting thought.

[FenderBender]

I found this power supply schematic a few months ago. I'm planning on building it when I get the chance.

http://www.extremecircuits.net/2010/06/fully-adjustable-power-supply_17.html

Unlike most Lm317 supplies, this actually has a properly implemented adjustable current limit system. I have not built it up yet, but it does look like a pretty good design. I'm sure it's not perfect but looks like it would work a treat.

As others have said...you might not have the same fit and finish as you would have with a prebuilt supply...However, you can guarantee good quality since YOU will be making it. You can use prime-spec parts and do your best job soldering. You wouldn't even need a premade PCB for this. Some good jumperwire and some protoboard would do well.

[david77]

I've said it before and I'll say it again: A home made lab PSU isn't that hard to build and it might teach you quite a bit. I wouldn't touch a LM317 based circuit, though, it offers no advantage over a discrete design, in fact there are only drawbacks. The LM317 was not intended as a lab PSU regulator and it's limitations are quite severe in this application.
The minimum specs for a universally usable PSU seem to be 0-30V at 0-3A, you'll have a hard time achieving that with the 317.
You won't be able to match the chinese supplies if you have to buy expensive things like transformer, case and heatsinks. A bit of imagination can save you a lot though, all I say is: old audio power amps. Get one cheap and reuse transformer and heatsinks.

[PA4TIM]

I needed a symetrical psu so i bought the calltech from the startpost a few months ago. I tested it on my dynamic load and it had rather good performance. It has tracking and is usable for what i want.
I plan to calibrate it and put 10 turn potentiometers in it.

But it has some disadvantages.
The voltage and current setting is very coarse and it switches over.
It does not like inductive loads
The current and voltage readings are way off ( around 10%)
The manual tells it is only short circuit proof for a limmited time ( strange if it has CC mode)
Buttons are flimsy

As a reference, i build several powersupplys myself, the latest can do over 20A but I limmited it to 10A and 0 to around 25V. It has a lot of options ( like pulse and triangle modulation for battery experiments or as power pulsegenerator)
Commercial: 2 voltcrafts, they are not bad, a Dutch Delta (30V-10A, CV, CC, OVP and sense) PM2811 ( same things as Delta but lots of other functions) the cute little firtst transistorised HP PSU)

Buiding yourself in not difficult. The problem and money is in a good cabinet, knobs, transformer ect.
Take a LM317 as Vgs, a few IRFP256 mosfets for voltage regulation. A second LM317 for Vgs for a few mosfets as shuntregulators. This is the current regulation. A few opamps to lift the voltage above the voltdrop over the shunts. ( so if there is 5V over the shunt the voltage adds 5V so 12V stays 12V ( as possible, over a short you wil not get 12V so a second opamp is a sort of OVP.

But Bob Pease has a nice schematic in analog troubleshooting. A few transistors and a opamp. It had a foldback current limmiting.

Fred

[T4P]

Stay away from the LM317 for lab supplies, nobody worth their money even said they are good for lab PSU's anyway

[shebu18]

What about the lm338?

Sent from my GT-I5800 using Tapatalk

[PA4TIM]

Stay away from the LM317 for lab supplies, nobody worth their money even said they are good for lab PSU's anyway

This is a bit unfundated. If you write things like that, please tell why you have that opinion.

You can not make a lab psu just using one LM 317 , that is right, but that does not make it not wordth the money. If used the right way it is a very good and usefull thing. And with some extra parts you can make a nice supply.

I use two in my lab supply but not the usual way, and not like you probably think, it is a rather radicaland unusual design ( see it as a current source stacked on a dynamic load with opamps that control the floating output voltage between the two)

But if you are not very good in psu design, are on a budget a psu using two LM317s ( on for voltage and one for current) can make a very usable lab psu that has CV and CC. And will not be worse than some of the cheap no-brand supplys ( on a Dutch forum dpsomeone used a cheap lab psu ( around 50 euro) to test diodes and with CC set at 300 mA and Voltage at 6V he smoked the diodes that could handle 1A / 40V and 25A pulse. It was a SMPS and it started to make noise too while smoking the diode, So talking about crap...

And if you disagree i would suggest, build one and test it like i did ( just as an experiment) the schematic is in the datasheet and with some extra parts you can make it even better.  It will do nice for most normal experiments.

[G7PSK]

I bought a Zhaoxin power supply on ebay from a supplier in the UK. It arrived with a US type plug but labeled as 240 volts on the back of the unit, I contacted the supplier before switching the unit on in case it was 110 volts with the wrong label on rather than the wrong lead. They told me I could cut the plug off and fit my own and if it blew up to contact them. I plugged it in with a travel adapter that I had and it did not work the light came on but no power out. So I contacted the supplier and as they had no more in stock and  the manufacturer would not take returns they said to bin it and gave me a full refund straight away, came through in minuets.
As I now had a non working new power supply for free I decided to have a look inside and see what's up. At first it looked like the output transistor (2n055) was not up to scratch but changing that did not do much further prodding was not going well when I happened to touch the zener diode which had tested fine and looked like it was soldered in well, the whole thing sprang to life. On investigation the joint although it looked bright and clean had not bonded to the diode's wire on one end, I removed the solder cleaned the wire end and re-soldered. End result one power supply for free.

[jorgemef]

Is there some tear down for these Chinese ones? I know they are cheap and probably very unreliable. The price pays the box.
http://www.ebay.com/itm/5A-30V-DC-Power-Supply-Adjustable-Dual-Digital-Variable-Precision-Lab-Grade/272798805512?_trkparms=aid%3D222007%26algo%3DSIM.MBE%26ao%3D1%26asc%3D20140131123730%26meid%3De5ad01ddb00248fb8202629c9b98d288%26pid%3D100167%26rk%3D5%26rkt%3D15%26sd%3D282618025088&_trksid=p5411.c100167.m2940