Power Toroid Wdc vs rated VA and acceptable temperature rise

[Tubetec]

I remember seeing an interview with Rupert Neve, initially he was very suspicious of Smpsu's ,he didnt see the logic of essentially using a noisemaking circuit to create his supply rails ,Ive been hanging on the sceptical side of the fence in this regard also .
I guess if your switcher circuits are nicely sheilded in aluminium ,spurious rf radiation is minimised and no need for exotic mu metal sheilds.
Longevity is the other question ,we do know that a conservatively rated old style supply can live 50 years or more ,I guess de-rating your smpsu's will help alot in this regard ,but still, long term stability is a bit of an unknown .
I see in some of the better high end regulated supplies ,as in my studer C37, theres always low ohm series resistance in the heater circuits ,presumably this limits inrush current when powered on from cold , cold heaters are only a fraction of the resistance of warmed ones .
A little higher voltage and dropping resistors might play out better in the long term for smpsu's for heater supply ,both for the valves and the supply itself.

 

[kambo]

for switching power supply on heaters,
do u still need to  rate its power x 2_ish too, or with good vent it will be fine?

 

[Tubetec]

I havent used Smpsu's in any audio ,but overspecing /derateing sounds like a good plan for long term reliabillity ,that goes for any electronics really .

 

[Tubetec]

when it comes to commercial electronics this affects the bottom line ,but for home made stuff why not over spec ?

 

[ruffrecords]

for switching power supply on heaters,
do u still need to  rate its power x 2_ish too, or with good vent it will be fine?

The thing about tube heaters is that their resistance when cold is a lot lower than when hot so when first powered up the current they would like to draw is several times the nominal value. The problem with SMPSUs when I first tried them a few years back is that most of them used foldback current limiting. This means if the load tries to draw more than the rated current, the current is reduced, often to about 25% of the nominal amount. For tubes you need enough current at switch on to ensure the heaters do start to heat up so their resistance rises enough that the current drawn drops and the power supply comes out of current limit. I found that to ensure this you had to rate the SMPSU for at least twice the nominal current and often more if it has foldback current limiting.

As I mentioned previously,  when looking for a SMPSU for the lunch box, I came across a new style of over current limit called hiccup mode. This apparently attempts to maintain a constant output power by turning off and restarting the supply so you get bursts of current until the load no longer draws more than the rated current. I had no idea how well this would work so for my first test I chose a SMPSU rated at twice the nominal heater current. Luckily it works fine. Two times over rating may not be necessary for hiccup mode SMPSUs so I might later try one rated just a little over the nominal heater current and see if that works. These supplies are so small and cheap that over rating has very little financial or space penalty.

Cheers

Ian

 

[sahib]

I remember seeing an interview with Rupert Neve, initially he was very suspicious of Smpsu's ,he didnt see the logic of essentially using a noisemaking circuit to create his supply rails ,Ive been hanging on the sceptical side of the fence in this regard also .
...................................

He must have had that view before the Amek 9098 and Purepath series rack mount stuff which carry his signature.  They have switching for step down and  linear as the final regulator stage.

 

[Tubetec]

Well I guess Rupert  was speaking retrospectively in relation to smpsu's ,Its getting hard to find stuff nowadays without switchers .
Im coming across more and more issues with class D amplification recently ,the actual output stages are holding up ok ,but its the driver Ic's that seem to be the weak link .

By the way hello Sahib,I think I bought some spares for my Amek BC3 from you a few years ago ,saved me a pile of trouble wiring up edacs.If you happen to come across anymore of the jack back panels for the BC3 do let me know. I see your line of pre's and eq's has gone from strenght to strenght ,always been something a bit special about Graeme Langley eq sound.

 

[sahib]

Yes sir. I think in 2013. I'll need to check the archive e-mails.  I'll certainly drop a mail if another BC3 comes up.

Also thank you for your comments. We have more designs coming out.

 

[CJ]

note that with a toroid, all the heat generated by the core has to go through the coil, which is another heat generator.

brain food>

 

[CJ]

thermal schematics>

 

[CJ]

more thermal brain food...isn't this exciting?

 

[abbey road d enfer]

I havent used Smpsu's in any audio

  I have, and I must say that noise performance is quite variable between brands. It seems to be more consistent now than it was a few years ago, though.

but overspecing /derateing sounds like a good plan for long term reliabillity ,that goes for any electronics really .

Up to a point. SMPS should not be used with a low load; under about about 10% load, they dissipate more heat than at nominal load.

 

[ruffrecords]

  I have, and I must say that noise performance is quite variable between brands. It seems to be more consistent now than it was a few years ago, though.

Which brands have you found to be the best?

Cheers

Ian

 

[abbey road d enfer]

Which brands have you found to be the best?

Cheers

Ian

I've found the XP Power to be very quiet and consistent.

 

[Tubetec]

Hi
I have a few questions ,interested to hear opinions

Do you typically connect direct to the outputs ,add extra lc filters , or use a linear reg after the switcher
and what kind of grounding arrangement on the heaters  do you consider best 

Would ac lead dress and centre taped resistors to ground be appropriate here?
maybe common mode chokes and caps forming LC filters to drop off a volt or two after the buzz generator ,
just to kill off any supersonics ,
a couple of ohms of copper in series with the load might  help reduce cold  start current also

Another question that comes to mind with switchers is mains isolation ,traditionally the mains transformer galvanically isolates the mains voltages from the user  , whats the situation with switchmodes in this regard

cheers,

 

[Tubetec]

switching frequency , does it matter ?
seems reasonable to use a frequency well beyond the capabillity of the amplifier your powering ,

 

[AusTex64]

Neve Portico stuff all uses SMPS. The newer Manley stuff is using SMPS for all supply voltages, designed by Bruno Putzeys. I am very curious to check one of these units out. Tubetec's question about isolation provided by a mains transformer is interesting. I know a fellow that designs SMPS, will ask about that.

 

[ruffrecords]

I've found the XP Power to be very quiet and consistent.

That is good because they also have hiccup mode.  Thanks Abbey.

Cheers

Ian

 

[ruffrecords]

Another question that comes to mind with switchers is mains isolation ,traditionally the mains transformer galvanically isolates the mains voltages from the user  , whats the situation with switchmodes in this regard

cheers,

They are mains isolated. Most start with a filter followed by a bridge rectifier and a reservoir cap. The feeds the switch which is connected to a transformer. The secondary of the transformer is the output and feed back is achieved by means of an opto-isolator. So yes, they are galvanically isolated.

Cheers

Ian

 

[kambo]

would these work on heaters

https://www.amazon.com/12-Volt-Power-Supply-Standard/dp/B01BDFQ7NO/ref=sr_1_4?ie=UTF8&qid=1503037680&sr=8-4&keywords=12v+5+amp+power+supply