SMPS from Connexelectric?

[Boscoe]

I have been interested in building a very high power amplifier for a while now but the materials costs are very high for class AB so I began looking at class T and D and so found Connexelectric. They sell which look like fairly high quality amplifiers and SMPS for audio. After reading the thread on the absence of SMPS in this forum I read the reasons why, they began to worry me about buying a SMPS for an audio amp.

THIS is the one I have in mind. What do you think of it powering two class T amps capable of 1000Wrms into 4ohm?

Thanks
Boscoe.

[kipman725]

well the +/-31V rail voltages aren't enough for 1000W into 4ohm for a start unless you were using it as a single 62V rail and each amplifier was operating as a full bridge whereupon you still won't get 1kW, only 960W in the ideal case. In addition to this the regulation on a quasi resonant supply is not good so your rails will sag and you will get even less. Looks like a fine PSU though.

[eem2am]

i dont see any NTC's or inrush protection.

Also, is there primary current limiting.?

-if not , then it will blow up if started up into a short circuit.

[eem2am]

This SMPS is obviously a quality product, made by experts, though from a marketing veiwpoint, to encourage peopl such as the threadstarter to buy it...i feel that more information would be needed.......


Also, look at those tiny little red capacitors near the transformer.

-Do those tiny things, even in parallel, look like they can handle the ripple current that would be prevalent with a 2KW load.?

I am very doubtful.

If i was you, id ask them to give you the datasheet on those caps, ...then ask them to tell you what is the ripple current in those caps..........when its on 2KW load, Min Vin....also ask for the switching frequency......


those caps just dont look up to 2KW to me.

if you really want to be sure about this power supply, then request that they tell you the following information.............

Primary FET rms current at 2KW loading (also request the FET datasheet)
Primary 100Hz ripple current in electrolytics at 2KW loading. (also request the cap datasheet)
Mains input rms current at 2KW load and min vin.
Output diode current at 2KW load and Max vin (also request the diode datasheet)

FET case temperature at Min Vin, 2KW load, 30 degC ambient, no forced cooling.
Diode case temperature at Max Vin, 2KW load, 30deg C ambient, no forced cooling.



What assurances can they give you that this is a bona fide product.?

or is it that one needs to already be a design capable smps engineer in order to be able to buy this product , and assure that it is safe to put into the application?

[!]

I doubt there's anything wrong with the red film caps, there is a fair amount of bulk electrolytics on both sides of the transformer. More significant to me would be seeing the output on a scope at rated power and at expected amp loading range during use.

As for inrush current, it looks like there might be a sandstone resistor right next to the AC input.

[eem2am]

Sandstaone resistor seen...thank you for that...looks liek a relay next to it to switch it out(?)

Anyway, the caps are 200V, Cornell Dubilier.........the unit is for 230VAC, so these caps must be in series. (2 pairs in series)

I do not know which one of cornells caps this is...but we can take a look at their "high ripple" current range.................

http://www.cde.com/catalogs/SLP.pdf

....the 1500uF, 200V variety can only handle 3 Amps of 100Hz ripple

...we have two pairs in series , that means we can handle 6 Amps of ripple from the mains current.

There is no PFC here...so mains input current spikes will be enormous at 2KW load....................and there is no way that 6 Amps of ripple current rating will be enough to handle the 2KW loading.

So those caps have been under -rated.

its up to you whether or not you think your load is going to be high for short-time enough for those input electrolytic caps to survive.

......in short, what we have here is an smps that cannot handle the maximum load if it was utterly continuous..............
-however, what audio load is utterly continuous.....none..........

---so why would you want to pay for an smps that can handle max load continuously for ages....................

you can save money and buy the connexelectric smps.

-but shouldnt you do extensive testing prior to integrating it in your product?...just in case.

-and if this smps goes obselete.......you have to pick another one , and then do all that testing all over again.

or suppose connexelectric change a component...perhaps due to obselesence.......then does that mean you have to repeat all your smps testing on the updated smps.?..............very possibly.

-specially if they change those input caps..or worse still , those film caps.

[eem2am]

Hi,

I think i have found the datasheet for those red film capacitors in the picture of the smps.

Film cap datasheet:
http://industrial.panasonic.com/www-...BD0000CE40.pdf

I think its something like the ECWF series.......

This series has one of the highest ripple current abilities on the market, so this will be a representative type, even if its not the right one.

...but if you check on page 6, you can see that the dissipation factor starts rising worryingly as frequency increases.

The 400V, 1uF ECWF105 has an esr of around 1.3 Ohms at 85KHz.

That is a lot.

and when you think that power goes as I^2.R...you can easily see how hot this capacitor is going to get with just a little current in it.

eg...just 3 Amps of ripple current would make it dissipate 13.6W.....and do you think a little capacitor like that can dissipate 13W?....Im afraid i do not.

-how much current would flow in this cap when the load was 2KW........a lot more than 3 Amps.

There is worse to come for this cap, because as page 7 shows, the current rating degrades seriously with increase of temperature.

So i am very doubtful that this smps could be able to handle max load continuously for any length of time.................but then again...maybe customers dont mind that....maybe they just want something light that can provide for the odd, short pulse of power.

[kipman725]

Ah I see there is a higher voltage version. The power supply needs to have a boost converter followed by a full bridge or half bridge driving the transformer primary at this power level. This is because the boost converter can correct the power factor, without it running at maximum power level is a bad idea. Also soft start is mandatory which from the big resistor near the input this supply has, would be interesting to see if that resistor has the thermal capacity to withstand a start into a short circuit.

[eem2am]

I am surprised they didnt put a PFC stage in front of SMPS2000R.

It would have given a nice high Vin and you avoid sky high rippl ein the input 'lytics.

I think SMPS2000R is a Series resonant half bridge converter.

At any given frequency , it has a fixed gain, and theyve just chosen this for the Vout needed.

the fet duty cycle will just be 50% each...with a bit of deadtime........the primary current just sorts itself out...depending on the load...but i hope they have primary side current limiting.

[!]

eem2am, this type of capacitor is used in SMPS all the time without failing.

While this SMPS may not have the cleanest output (which is why I'd want scope readings), I am fairly confident that they aren't selling PSU that they've never even tested.

I agree it may not be capable of max continuous 2000W output but it is doubtful the capacitors would fail, except if the rest of the PSU had inadequate ventilation or they were otherwise defective discrete components.