The closed loop phase response can be disturbed in many power supplies with an external capacitive load. Even some linear regulators get upset when a low ESR cap is placed across their output.
Hanging one of those "Farad Bomb" caps that the car audio guys use across the 12 volt output of a typical PC ATX supply will generally make it very angry and sometimes blow it up. Note, I discovered this about 20 years ago. Not sure if it's still true.
The effect is somewhat similar to placing a cap across the speaker leads of an audio amp that has lots of GNFB. Often the square wave response will start showing overshoot before more capacitance just rounds the corners. That overshoot will trigger the OVP in the SMPS.
Sometimes a small resistance or choke between the power supply and the capacitive load can help. I have seen a series diode help in a few cases.
Another trick I have used is a common mode choke between the SMPS and load. This requires a supply with an isolated output which this unit claims to have.
A 50/60 Hz power transformer with a dual 120 volt primary can often be used as a low frequency common mode choke. Just leave the secondary open and put a small cap across the output.
Two such chokes can be used to effectively isolate individual channels from each other in a stereo amp without huge caps. Test the phasing in all 4 wiring combinations for lowest crosstalk and ripple.
Hanging one of those "Farad Bomb" caps that the car audio guys use across the 12 volt output of a typical PC ATX supply will generally make it very angry and sometimes blow it up. Note, I discovered this about 20 years ago. Not sure if it's still true.
The effect is somewhat similar to placing a cap across the speaker leads of an audio amp that has lots of GNFB. Often the square wave response will start showing overshoot before more capacitance just rounds the corners. That overshoot will trigger the OVP in the SMPS.
Sometimes a small resistance or choke between the power supply and the capacitive load can help. I have seen a series diode help in a few cases.
Another trick I have used is a common mode choke between the SMPS and load. This requires a supply with an isolated output which this unit claims to have.
A 50/60 Hz power transformer with a dual 120 volt primary can often be used as a low frequency common mode choke. Just leave the secondary open and put a small cap across the output.
Two such chokes can be used to effectively isolate individual channels from each other in a stereo amp without huge caps. Test the phasing in all 4 wiring combinations for lowest crosstalk and ripple.
The real annoyance is the fact is won't start into cold tubes but it's sold as a power supply for a tube amp.
And yes, the HV hiccups and the timing process starts over again.
Eliminating the 30 second B+ delay will allow it to retry starting enough to latch eventually, but that's not what I'm after... I will swap out this rubbish for a 12V 13A SMPS that will run the whole system. I might add a linear B+ supply for the phono stage. DC-DC boost is usually clean enough for me though. We'll see what today brings.
I just wish it would blow up instead of using these overzealous protections.
And yes, the HV hiccups and the timing process starts over again.
Eliminating the 30 second B+ delay will allow it to retry starting enough to latch eventually, but that's not what I'm after... I will swap out this rubbish for a 12V 13A SMPS that will run the whole system. I might add a linear B+ supply for the phono stage. DC-DC boost is usually clean enough for me though. We'll see what today brings.
I just wish it would blow up instead of using these overzealous protections.
Tube amplifiers typically have reservoir capacitors being class AB. Why wouldn't a PSU designed for use in a tube amplifier support that concept? A simple PWM ramping of the output would do it easily.
For the record, I am using a 12V 250W waterproof LED supply with DC-DC boost converters to do exactly the same thing without issues. Actually, that one has much more capacitance since it includes a phono stage, too.
It's simply a poor design IMHO. It might be good enough for a simple 5W guitar amp though.
For the record, I am using a 12V 250W waterproof LED supply with DC-DC boost converters to do exactly the same thing without issues. Actually, that one has much more capacitance since it includes a phono stage, too.
It's simply a poor design IMHO. It might be good enough for a simple 5W guitar amp though.
kodabmx, you purchased that smps as a test. It doesn't work for your particular application. It seems like you do not know enough about smps to manage that particular smps, and so just fall back to calling it 'rubbish', and stating it is a 'poor design' when you have not looked at its design at all - only the sales blurb.
If it is being sold as a power supply for tube amps, but indeed will not turn on into 8 preamp tube heaters, then I will not buy one even as a test.
How often have we seen vacuum tube designs in magazine articles, internet sites, circuit boards, kits, and even a few products where it's obvious that the designer (or copycat artist) created something with no, or minimal knowledge of vacuum tubes.
Some were obviously never taken beyond the simulator stage before publishing, posting, or worse, selling kits. The most memorable was an article in an EE trade journal, either EE times, or EDN where two authors created "boosted triode" or "super triode" mode to get big power out of a 6L6GC by running the screen grid several hundred volts higher than the plate. It works in LT spice, try it on a real tube and see what happens. This article was published over 10 years ago, but I can't find it now.
How often have we seen vacuum tube designs in magazine articles, internet sites, circuit boards, kits, and even a few products where it's obvious that the designer (or copycat artist) created something with no, or minimal knowledge of vacuum tubes.
Some were obviously never taken beyond the simulator stage before publishing, posting, or worse, selling kits. The most memorable was an article in an EE trade journal, either EE times, or EDN where two authors created "boosted triode" or "super triode" mode to get big power out of a 6L6GC by running the screen grid several hundred volts higher than the plate. It works in LT spice, try it on a real tube and see what happens. This article was published over 10 years ago, but I can't find it now.
A simple SMPS PFC or boost only uses a low side mosfet and thus don't control current inrush. A large cap simply appears as a short.. thus the inductor and diode see a considerable inrush. The only thing a low side only mosfet topology can do is move the short forward which doesn't really help, the result is a spike from the inductor. In the end there's a loop that goes from the live through the inductor to the diode to the cap... that's the problem. (ignoring the isolation transformers which end up dying with large inrush)
A topology that has the option to chop the inrush current is a better bet assuming the design is ok for that level of switching. Active rectification for example could put that way ahead of the PFC/PSU.
Put 4700uF 450V for a OTL and you'll start finding all the issues with simple SMPS topologies impact tube amps as well as solid state :/
Last edited:
Post 30 indicated is starts the heaters... But only 8 preamp tubes, and it had to restart several times for that, too. It does not work as advertised.
The specs say it should power a stereo 6L6 amp based on the power ratings (with almost 400mA to spare) but it doesn't even power a phono preamp.
This "newb" uses reservoir capacitors like everyone else. This is sold as a replacement for a passive PSU, and it doesn't work as advertised. Big caps aren't just for smoothing but you know that
In any case, I've replaced it with a different SMPS and that starts into the load as expected.
The specs say it should power a stereo 6L6 amp based on the power ratings (with almost 400mA to spare) but it doesn't even power a phono preamp.
This "newb" uses reservoir capacitors like everyone else. This is sold as a replacement for a passive PSU, and it doesn't work as advertised. Big caps aren't just for smoothing but you know that
In any case, I've replaced it with a different SMPS and that starts into the load as expected.
If it starts the heaters, then it starts the heaters - job done.
Amps with smps B+ don't need a large first filter cap - job done.
Adding extra criteria like cannot hiccup at the start, or must use large first filter cap because that is what classic schematics show, is just audiophoolery imho.
Amps with smps B+ don't need a large first filter cap - job done.
Adding extra criteria like cannot hiccup at the start, or must use large first filter cap because that is what classic schematics show, is just audiophoolery imho.
And in my opinion if it doesn't turn on the first time, it's not working properly.
It's like buying a car that has a rated 0-100 in 9 seconds, but when you get the car home you find out it will only do that in freefall from a cliff. There would be lawsuits.
In this case, I simply bought it to try. It sucks. I'm writing about it here so others won't buy it and expect it to work as expected. You shouldn't have to augment a PSU to make it work properly IHMO.
If the thing is rated to run tube heaters, and it's rated for 4A, I should be able to light up 4A worth of cold tubes with it - It's for a bloody tube amp after all.
It's like buying a car that has a rated 0-100 in 9 seconds, but when you get the car home you find out it will only do that in freefall from a cliff. There would be lawsuits.
In this case, I simply bought it to try. It sucks. I'm writing about it here so others won't buy it and expect it to work as expected. You shouldn't have to augment a PSU to make it work properly IHMO.
If the thing is rated to run tube heaters, and it's rated for 4A, I should be able to light up 4A worth of cold tubes with it - It's for a bloody tube amp after all.
Perhaps you didn't appreciate that valve heaters rated at 4A continuous require about 16-20A of start-up if powering from a constant voltage supply. The smps ratings say DC12.6V@4A, and DC 6.3V@4A. If you want to operate the smps at over 4A, even for a short time, then expect it to do what it did. It is your expectations that need adjustment, and what you should comment on imho, rather than voicing disdain at the smps' design or not meeting its specs.
Perhaps you don't appreciate If it's MARKETED as an SMPS for lighting up tubes, it should be able to deal with the inrush. End of.
And yes, Tony. I usually derate to 40%. It all depends on the design though. In this case, the total loading would have been about 75W -- I didn't expect the thing to power an amp full of 6P45S, but a preamp?
I don't expect to derate a waterproof potted SMPS that says it will output full power @ 50°C, but the convectively cooled ones I always derate. The only other SMPS I don't derate is a computer ATX but I don't buy the cheap ones and I usually just buy the price point even if it's more power than I need. My current computer is running on an 850W EVGA but the maximum power used my the machine is under 300W. At least I can add a second video card if I want
Happy Saturday
And yes, Tony. I usually derate to 40%. It all depends on the design though. In this case, the total loading would have been about 75W -- I didn't expect the thing to power an amp full of 6P45S, but a preamp?
I don't expect to derate a waterproof potted SMPS that says it will output full power @ 50°C, but the convectively cooled ones I always derate. The only other SMPS I don't derate is a computer ATX but I don't buy the cheap ones and I usually just buy the price point even if it's more power than I need. My current computer is running on an 850W EVGA but the maximum power used my the machine is under 300W. At least I can add a second video card if I want
Happy Saturday
I find it's a crap shoot. This supply is well made and cost over 60$ but doesn't work as expected. The cheap 12$ 12V 13A supplies I use for heaters in my power amps work perfectly every time.
When I find a PSU that works as expected, I tend to stick with them.
I have a 12V 10A supply that can start into cold tubes without retrying for example. It sounds like a degauss coil when you turn it on but it works LOL It was a very old SMPS design though. It had a 120V/240V switch on it instead of being universal.
When I find a PSU that works as expected, I tend to stick with them.
I have a 12V 10A supply that can start into cold tubes without retrying for example. It sounds like a degauss coil when you turn it on but it works LOL It was a very old SMPS design though. It had a 120V/240V switch on it instead of being universal.
I see, a filament can be 1 ohm cold and 40 ohms hot!
A cheap 1A switcher wont complain if you tack on a little current source board. With this you don't even need to buy a grossly oversized amperage switcher, just one with enough voltage to overcome the drop out. The voltage trimmer on the switcher could be handy getting the drop out margin optimized, not sure. This arrangement shouldn't be used with Filaments wired in parallel though, because if you remove a tube or a filament blows, the remaining tubes will be given a gross overvoltage as the CCS raises the voltage to maintain the original current set point, then you burn out all the remaining Filaments. A very bad failure mode, but for series Filaments every failure mode I can think of seems pretty safe.
Last edited:
Windcrest, up to 5:1 resistance ratio hot:cold, not 40:1.
Also note that the smps starts the heater supply up fine, and regulates to 6.3Vdc. There is no 'need' to add an extra current limiter.
There is no functional benefit in adding 220uF or whatever as an extra first filter B+ cap - the smps includes a B+ cap that keeps the B+ stable and regulated.
Also note that the smps starts the heater supply up fine, and regulates to 6.3Vdc. There is no 'need' to add an extra current limiter.
There is no functional benefit in adding 220uF or whatever as an extra first filter B+ cap - the smps includes a B+ cap that keeps the B+ stable and regulated.
Last edited:
It doesn't start it up "fine" and nobody ever said it did except trobbins.. It must retry several times - that is NOT fine. You are suggesting to remove all PSU caps from the amplifier circuitry?
Removing the 220uF caps from this circuit results in a crosstalk from L to R channels, and generally poor performance. Combine this with a line stage, tone controls, and buffers? All should have zero PSU decoupling caps? I don't think so.
It also might be defective since I have no others to compare it to. I don't know why you're defending such a garbage PSU, but If you want it, I'll give it away free + postage. ~25$CAD to send to Melbourne as surface mail, 45$CAD for air mail.
Removing the 220uF caps from this circuit results in a crosstalk from L to R channels, and generally poor performance. Combine this with a line stage, tone controls, and buffers? All should have zero PSU decoupling caps? I don't think so.
It also might be defective since I have no others to compare it to. I don't know why you're defending such a garbage PSU, but If you want it, I'll give it away free + postage. ~25$CAD to send to Melbourne as surface mail, 45$CAD for air mail.
Attachments
Last edited: