Recommendations for a good 24v SMPS?

[MiXiN]

As per the subject heading, I'm looking for a reliable 24V Switch Mode Power Supply at a minimum of 6A.


Is Meanwell any good, and if so, can anyone provide details of a legit' reseller as I'm led to believe they're widely copied.

[teckniqs]

To be honest I'd just go for the 15A 360w 24v for under £12.00 on eBay. I know it's not Meanwell but it will do exactly the same thing yet it's three times cheaper but 3 times more power so it will run stone cold for what you need. And it's the same in size.

http://www.ebay.co.uk/itm/361513735340

[Albert H]

If you use that Ebay job, you'll probably need a big smoothing capacitor - about 10000µF @ 35V should do the trick nicely. LED lighting supplies usually don't have very good smoothing!

Incidentally, a good "rule of thumb" for rig smoothing capacitors: about 1000µF per Amp drawn at 1.5 times the circuit Voltage.

[teckniqs]

I've used them before and they actually do work OK with no need for extra smoothing.
...You shouldn't automatically think just because it's so cheap it couldn't possibly ever work like a Mean-Well one which is basically all of the same parts inside. (well they seemed to work fine for me anyway and it won't be very expensive to find out if it does for you lol)

For a 6amp 24v supply that 15a power supply should be able to nicely do the job for just a very small fraction of the price, otherwise here is the closest I could find to the Mean-Well that you need from RS Components at just £155.00:
http://uk.rs-online.com/web/p/embedded- ... s/7704027/

[shuffy]

I've used them before and they actually do work OK with no need for extra smoothing.

I can concur with Teck, I've used a 15A one from ebay (about £16) driving one of those cheap Chinese MRF186 boards. The results were encouraging, it was quiet enough without a smoothing capacitor on it. The rig I built was just to test the concept, so it was a lash-up. The exciter board I commonly use produces just over a Watt out. This ws powered by a decent PSU but to test the MRF/cheap PSU combo, I lashed up an RD06 stage to bump the drive up by a couple of Watts. Soak tested for a day or two, but not actually used on air - this is a different matter!

I do think I'd add extra smoothing on a "production" version, but it depends how noticeable any artefacts are on the output.

[MiXiN]

Nice one for the input.

I decided to order the one suggested but at a little extra cost from a UK based seller.

I'll get back to you when I've given it a power up and report my findings.

[Albert H]

I found a 10A "LED" supply in my workshop - bought for some LED striplights, but never put to use. I tried it - turned up to 27.5V - into an old rig with an MRF317 final. At 27.5V, it does exactly 100W on 98MHz. The exciter derives its 15V from a well filtered 7815 regulator IC from the main supply.

Findings: A tiny bit of hum on the carrier, but very low level. Looking on the spectrum analyser, there were spurs 120kHz either side of the carrier (the switching frequency) which were only -42dBC. There were also some 100Hz sidebands, (a tiny bit of AM on the final). Adding a smoothing capacitor (10000µF @ 63V) eliminated the 100Hz completely and reduced the 120kHz spurs to -51dBC.

I added a 120kHz trap to the input of the 7815, which gave a small improvement to the spurs, but the "fix" was a trap in the driver stage (a 2SC1945). I used a Panasonic ELC12D181E 180µH 1.5A choke in parallel with an 8n2 and a 1n5 which reduced the 120kHz spurs to virtually nothing. I was going to try another similar trap on the final stage, but I didn't have an inductor of that kind of value with enough current-handling.

The rig is now clean enough with its £15 power supply, but a word of warning - if you're going to try this approach for cheap power supplies, you'll need a spectrum analyser and be prepared to try a few things out to get rid of the carrier noise and the switching frequency sidebands that will be generated.

Incidentally - in the old days, I used to buy cheap car battery chargers from Halfords. These were not great as chargers, but they had a huge transformer inside, with a secondary current capability of about 15 Amps! The unloaded voltage out of these brutes was about 15V, dropping to about 14.5V on load. I used to use two of these transformers with the primaries in parallel and the secondaries in series. I used to take a few turns off the secondaries to reduce the voltage a bit (which gave me some nice, fat enamelled copper wire for PA stages!). Rectified and smoothed, my target voltage was about 32V (on load). This power supply would happily power five 587BLY transistors (one for the driver, and four for the output) to deliver a healthy 300W!

When you consider that the car battery chargers cost just £7.50 each, the smoothing capacitors were "out of spec." ones we got from a computer manufacturer for nothing, the high voltage output filter and matching capacitors were home made, and the 587BLY only cost £3 each (unless you did a deal for 200 of them like I did, and got them for £375!), I could build a 300W PLL rig in a nice 4U 19" case with a big heatsink and fans on the back for about £125 and two evening's work!

There are still car battery chargers around - it's frequently cheaper to buy one than just to buy the transformer inside it! I also used those transformers in Medium Wave boxes, both for power and for modulation. I remember going into a branch of Halfords and explaining that I was buying all the chargers that they had in stock because I was giving them as Christmas presents!

[thewisepranker]

^ They're even nicer when you have a trade card.

[Maximus]

All these SMPS's look the same. To be fair I have a stack of broken Meanwells here. The only good thing I can see is that they have OK smoothing caps and an ample amount of heatsink compound.

Used the cheap 48v power supplies in the past and they look the same on the inside when compared to the Meanwells. Never had a problem with noise.

Will post some pics of the insides when I get a chance.


Sent using Tapatalk

[Albert H]

I was interested in the differences between various classes of SMPSU, so I've got the lids off a "Meanwell", an "Advance", a "Coutant", a "Babcock" and a cheap Ebay Chinese LED supply.

The Coutant and Advance PSUs have switching frequency traps in series with the DC output, and a fairly large smoothing capacitor across the DC as well - 10000µF in the Coutant and 3900µF in the Advance. The outputs from these supplies - even under full load - is clean as the proverbial whistle.

The "Babcock" effort is pretty horrible, and is evidently designed to be connected across a big lead-acid battery as a charger. This might be quite nice - with the battery connected - as a non-interruptible supply - for rigs in places with part-time mains (like the idiot who wired a rig up to the timed output of a street light switch panel!)

The "Meanwell" is quite clean, but compared to the higher priced "Advance" and "Coutant" it's a crude circuit. It's a bit better than the "LED supply", but not much!

You really do "get what you pay for". In all instances, I'd prefer to use a linear supply for a rig. I know that this can be an expensive option, and that the transformer can often be heavier than the rest of the rig, but there's a damned good reason that the best transmitter manufacturers around the world avoid the use of SMPSUs!

A switched-mode PSU introduces another variable into your design. It's already difficult enough to keep the output of a rig properly clean, so adding high frequency ripple to the DC rail really isn't sensible. It's a great way to get spurs that can't easily be removed. It must also be realised that any RF products that aren't on the frequency that you're trying to generate are just waste energy!

With some effort (as detailed in an earlier post) I made an "LED" supply clean enough for a 80 - 100 Watt rig. Considering the time, components and additional effort that was required, a linear power supply would have been cheaper!

[MiXiN]

I ended up buying this for a slightly higher price than one from HK, as this buyer was UK based and I didn't want to wait too long for delivery - http://www.ebay.com/itm/24V-15A-360W-Sw ... Sw0HVWDbHh

I must say, it runs really cool at 6A, there's no hum present on the carrier, and the build and soldering quality is really good - albeit the large Capacitors are cheap as to be exoected. It's also variable from 19.5 to 30.5V.

If anyone wants a SMPS in the UK, these really are pretty good, and I received this next day delivery well packed.

[teckniqs]

That's exactly the same voltage range the one I suggested gave me when tested. Probably the same thing. Just shows there's no real need to go out and pay silly money for something which is the same thing.

[MiXiN]

That's exactly the same voltage range the one I suggested gave me when tested. Probably the same thing. Just shows there's no real need to go out and pay silly money for something which is the same thing.

+1 on that mate.

Now time to search for a 12V 30A one to replace my old big heavy Palstar (Manson) 25A Linear PSU that's the same weight as an old CRT TV.