smps heater supplies

ed · Post by ed » Fri Jul 05, 2013 8:58 pm

I've started a new thread to stop polluting the others...

still very much intrigued by these smps for heater supplies, although I'm a bit mystified by some results.

Paul loaned me a cisco 48v unit sometime back, which I was going to try in a mic pre-amp that I was building. I lost interest in this application as I got the following results when I measured the unit.

open ended(no load) it presented 2 wave forms:
a) 2k at 5mv p-p (very pretty sawtooth)
b) 10mhz at 24mv RMS

with 4k7 load(to mimic 10ma phantom power):
a) 2k at 7mv p-p
b) 16mhz at 64mv RMS

I've resurrected interest after Andrews GK70 demo, and I'm still searching for a 6v unit to try with the SV572..but so far I'm thinking they are not very clean and require some clean-up...more measurements I think

here are the shots:

ed · Post by ed » Fri Jul 05, 2013 9:03 pm

and here is 4...

pre65 · Post by **pre65** » Fri Jul 05, 2013 10:12 pm

I've not scoped the 833a heater SMPS, but I suspect they could be a bit quieter (less noisy), so I wondered if something like these would be worth a go ?

http://uk.rs-online.com/web/p/leaded-inductors/1889056/

Paul Barker · Post by **Paul Barker** » Fri Jul 05, 2013 10:12 pm

Since this is used as a B+ supply in my pp amp do I take from this that I need a 2k (2 kilohertz?) notch filter on the B+?

Scary bad design! 2 khz

JamesD · Post by **JamesD** » Sat Jul 06, 2013 9:16 am

WHilst this is certainly not desirable - has anyone looked at the output from conventional heater dc solutions for a comparision?

I don't mean Andrew's new regulator or the Colmon ones but say the 317 or the 1085 ones?

I seem to recall that the 317 ones have more noise on them than this - more like 60-70mA and the 1085 ones are comparable but I don't have nice pictures of it...

SO whilst I think the use of a common mode and differential choke will greatly improve these supplies, I think they are actually decent 'out of the box' - not great but decent and usable in an economical implementation...

The nice thing is that the frequencies we want to filter here are relatively high and so the chokes are smaller and more economical than what I was trying in the GK-71 thread - I must admit that was a bit ambitious!

J

Mike H · Post by **Mike H** » Sat Jul 06, 2013 10:11 am

The advantage of SMPS is efficiency and minimal heat dissipation. But it's swings and roundabouts, also comes with noise which must be filtered out if it's an issue. Yes I would use a LC filter, this is also suggested on a buck inverter datasheet I've got. So it's obviously a well recognised issue 'in the trade'.

BTW I recently got a couple of Maplin netbook supplies to play with.

JamesD · Post by **JamesD** » Sat Jul 06, 2013 10:40 am

Excellent Mike! The more experience and measurements we have the better we can make use of them

J

Mike H · Post by **Mike H** » Sat Jul 06, 2013 11:29 am

I might be a while getting around to it

ed · Post by ed » Sat Jul 06, 2013 11:45 am

continuing saga...

I've just tested a couple of ITX 450w ex desktop supplies I've got which I use for battery recyclers.....They are much much much cleaner than the cisco unit...inasmuch as they do measure 5ish ma RMS at 31 and 62 khz..but nothing at lower frequencies....

I've got 4 laptop supplies to measure..I willl post scope images if anybody is interested....(i've found a memory stick so I can now post images instead of photos)..

JamesD · Post by **JamesD** » Sat Jul 06, 2013 11:58 am

Hi Ed.

very much interested!!!

The ITX supplies at 5mA - what voltage line was that? I would expect them to be cleaner than Laptop supplies as they are generally more expensive and have higher current capacities but the higher switching frequencies wasn't necessarily a given and is really useful data - much easier to filter!

J

ed · Post by ed » Sat Jul 06, 2013 12:02 pm

Hi James

The one mentioned was a 12v line, with a highish load on the 5v line to force activation.

Both the supplies mentioned have duff 3.3v lines, which is why I use them for other duties...Not sure whether that might be relevant.

JamesD · Post by **JamesD** » Sat Jul 06, 2013 1:09 pm

Interesting, thanks ED!

I need to check if you can stack the supplies for 17V... does anybody know?

J

ed · Post by ed » Sat Jul 06, 2013 1:32 pm

lots of things to consider...
I particularly liked the scenario where the two(or more) supplies don't start at the same moment and the slower starter is faced with voltages on its output......

can't life get complicated!

http://electronics.stackexchange.com/qu ... and-24v-dc

Post by **Nick** » Sat Jul 06, 2013 1:46 pm

Slightly bogus, I think some of those answers are confusing series and parallel. The main reason you cant connect them in series is

"And since the ATX output GND is connected to the AC input GND you would be shorting the output(s) of one supply to GND"

If the outputs were floating, then if they were in series, it would not matter what came up first, they would not see each others voltage. Thats what Paul did with the cisco supplies.

ed · Post by ed » Sat Jul 06, 2013 2:01 pm

I'm not sure if that wouldn't apply to the cisco units as well. Somebody(maybe Paul) has cut the earth wire at source on the unit that I have.....

interesting point..........

all ATX units I have seen have a 3 wire mains connection with earth connected.

All the netbook, laptop supplies I have seen have 2 wire connections.

Not sure about the mesh box type as I havn't seen one yet.

diet: maybe I'm thinking too much but if you use a 3 wire version for heater and you cathode ground the positive side of your heater then the same scenario would apply, shirley? maybe needs checking

audio-talk

smps heater supplies

#1 smps heater supplies

#2

#3

#4

#5

#6

#7

#8

#9

#10

#11

#12

#13

#14

#15