In the effort to take some of the weight out -- considering use of SMPS to supply tube amp. One way is 48V adapter feeding half bridge DC-DC chopper.
This could efficiently supply low and high voltage requirements.
48V adapter also satisfies safety requirements
This could efficiently supply low and high voltage requirements.
48V adapter also satisfies safety requirements
Mackie use a SMPSU in their 'newer' powered speakers. They produce 320volts DC, sufficient for most valve amplifiers. So a dual output SMPSU is what I would design, 6.3v DC for heaters and 400v for HT. All out from the same power supply.
Don't forget to ensure the SMPSU has Power Factor Correction otherwise it will be expensive to run! 0.5 as a PF reading will cost twice as mush to run as a PF of 1.0 .
They tend to be a little less reliable than old fashioned linear supplies, as you will find out.
Don't forget to ensure the SMPSU has Power Factor Correction otherwise it will be expensive to run! 0.5 as a PF reading will cost twice as mush to run as a PF of 1.0 .
They tend to be a little less reliable than old fashioned linear supplies, as you will find out.
Is your goal to "do a design" as a SMPS engineer? Unclear what the achievement is for; one off personal use or product?
Then I'd wonder how a power converter responds driving another power converter, which present that effectively negative resistance at its input. Under transient conditions which some music is going to have. At least the heaters will give Mr 48 a good non-zero static load to start from and recover to.
I bet it depends on which one. I thought I just saw one, showing all sides, at least one insert for the input line switch and I'm thinking, where's the certs? That little tag with all the regulatory symbols all over it. Doesnt have one? Let me guess; that goes on the product the power supply goes in. Hmmmm....
We used to kid the designers, who'd come up with this big monster feeding distributed power to the next set of converters; "Let's see 0.8 X 0.8, oh 64% efficient is it now?" They'd get miffed.
Then I'd wonder how a power converter responds driving another power converter, which present that effectively negative resistance at its input. Under transient conditions which some music is going to have. At least the heaters will give Mr 48 a good non-zero static load to start from and recover to.
One off design
The 48V source will feed a ZVS half bridge chopper with diode bridge rectified output. Clean switching, no spikes. No negative impedance -- does not regulate
Simulate today
The 48V source will feed a ZVS half bridge chopper with diode bridge rectified output. Clean switching, no spikes. No negative impedance -- does not regulate
Simulate today
Seems like it could work. There is some sag -- the three plots are 3, 25, and 50 watts. Series effective leakage 5uH -- that might be a challenge. Higher leakage
creates more sag. Starts up very smooth and ZVS at normal load range
OK, it looks like the gig is up, and I just bought 10 more of them, so here is the deal. A surplus house called Marlin P Jones has high quality open frame 48 volt 3 amp switching supplies on close out for $10 each. Buy enough of them and you get $7 flat rate shipping. So what does a 48 volt supply have to do with tube amps? Simple, buy a few and connect them all in series. They are rated for series operation but there is a limit due to a poly cap from the negative output terminal to chassis ground. Wire 13 of these supplies in series and those caps will have a short unhappy life. Take them out and all is well. I bought 13 of them several years ago when they were $13 I think, and they had a few hundred of them. There were 65 left when I just ordered 10 more.
I wired all 13 of them in series after removing the cap. I did successfully run a one channel test amp to over 500 watts of power output on them but did not draw enough current to trigger a current limit or shutdown. The long term plan is to run two channels at 500 WPC. It is not clear how these will act when one current limits before the rest, so put a big fat diode across the output of each one.
I have no affilliation with MPJA, except for seeing them at the various south Florida hamfests going back 40+ years.
https://www.mpja.com/48-Volt-Power-Supply-3A-Cosel-LDA150W-48/productinfo/34713+PS/
I wired all 13 of them in series after removing the cap. I did successfully run a one channel test amp to over 500 watts of power output on them but did not draw enough current to trigger a current limit or shutdown. The long term plan is to run two channels at 500 WPC. It is not clear how these will act when one current limits before the rest, so put a big fat diode across the output of each one.
I have no affilliation with MPJA, except for seeing them at the various south Florida hamfests going back 40+ years.
https://www.mpja.com/48-Volt-Power-Supply-3A-Cosel-LDA150W-48/productinfo/34713+PS/
Myself and at least one other forum member has been talking about building a 1KW tube amp for a long time. I have seen two different circuit topologies plus Pete Milletts Engineers Amp run at over 500 watts for several minutes in a single channel test but have not actually put all of the necessary stuff onto a chassis and into a cabinet yet. There is yet another "magic bullet" that could seriously improve the efficiency of the amp that I haven't totally worked out yet. I want to investigate it before building anything. Now, do I need, or even want a 1 KW tube amp. Not really, except for those winter days when it's 40 something degrees F in the basement. If Built it will never see max power in real life, and there will be some means to operate it at reduced power levels, by shutting down pairs of output tubes and reducing the supply voltage.
Never see max power? Not even once? You’re getting soft in your old age.
When I get mine going, I intend to do at least one rave with them. And you know how DJ’s treat the equipment.
When I get mine going, I intend to do at least one rave with them. And you know how DJ’s treat the equipment.
Doing the "Pheonix" low power bench tests before flying the big craft. 50 watts max for now.
For high power it might work to use PFC front end and LLC DC-DC or DC-DC like the one I posted earlier. That one can be regulated by varying frequency
For high power it might work to use PFC front end and LLC DC-DC or DC-DC like the one I posted earlier. That one can be regulated by varying frequency
I have no speakers capable of eating 500 watts. I had built a pair of cabinets with 4 X 12 inch "Pro Audio" speakers from Parts Express that would eat my Carver Cube at full tilt, but they remained in Florida when I left. Oddly the Carver claimed 201 WPC, but Pete's Engieers Amp with my 125 WPC mods was much louder. I think the Carver might have been capable of "201 watt" peaks but running it that loud for more than a few seconds would make it mad and it would give you the silent treatment while blinking its LED's at you. If I really want to get loud, I could build another set of speakers, but 1KW of sound in a concrete walled basement could be deadly.
Sounds like you need to build some new speakers. Say, a 2x15, with Kappa LF’s. Unfortunately, the drivers are now $195 a pop. Plus mids and tweeters.
You would need to bring them up for air during a full power test.
That Carver getting smashed by the Pete Millet amp - no surprise at all. We were just discussing why in the instrument/amp forum. Even the Phase Linears didn’t quite have the goods, but the BGWs (and their faithful clones) sure did. But if you’re having trouble moving stuff around these days, better get some amp racks with built-in dollies, and get some for your speakers.
How much do 13 of those 48 volt supplies weigh, together? Bet it’s more than one single Labgruppen FP14k.
You would need to bring them up for air during a full power test.
That Carver getting smashed by the Pete Millet amp - no surprise at all. We were just discussing why in the instrument/amp forum. Even the Phase Linears didn’t quite have the goods, but the BGWs (and their faithful clones) sure did. But if you’re having trouble moving stuff around these days, better get some amp racks with built-in dollies, and get some for your speakers.
How much do 13 of those 48 volt supplies weigh, together? Bet it’s more than one single Labgruppen FP14k.
You could always use a "power soak" and measure the 1 KW, just for bragging rights. Who says you have to put yourself in the position of "dissapative element" when some resistors would do? Save on speaker purchases too; you can still listen to it albeit at much lower volume, to hear how it sounds when exerting that much grunt. ;')
That’s how all my big ones are tested. War volume into a dummy load, 1k in series with a speaker. One can hear exactly how it sounds driven 6, 12 dB into clip. Then go to a sine wave and see how much the voltage at the wall socket drops, as well as what the amp is putting out.
Each 48 volt power supply weighs 458 grams, so 13 of them is ia about half of a FP14K. Far less than the pair of 120 / 240 to 120 / 240 600 VA isolation transformers (about 15 pounds each) and the associated box full of electrolytics that I first planned to use for a nearly 2 KW supply that could feed a 1KW amp. That supply put out about 650 volts unloaded, 635 when loaded with one channel at idle, and 615 volts with one channel at the edge of red glow (500 watts out). The SMPS string made 625 volts under all conditions without touching any of the voltage trimpots.
I see that the FP14K sells for about $7000. I'm looking to build my amp for under $700 counting the OPT's and power supplies.
Power Soak, I got your power soak right here. Rated for "8 ohms, 500 Watts," I bought two of these many years ago after causing an OPT to spontaneously combust burning my workbench while pushing a pair of 6LW6 tubes on 750 volts well beyond the "80 VA" power capability of the budget OPT's that I have lots of. My previous trusty "audio dummy load" was made from a series parallel combination of Radio Shack "non inductive audio test resistors" that were mounted on heat sinks. Some of them shattered, some just died, all went into the trash.
Attachments
Single 48V to 300V using boost rather than stacking 48V units. Pretty easy to do and high efficiency -- 97% for a good design for the boost stage.
Example is 100 watts but no problem going to kilowatts. The other advantage is a single 48V supply at that power will have PFC..
Example is 100 watts but no problem going to kilowatts. The other advantage is a single 48V supply at that power will have PFC..
Hello Tubelab, wondering if you can tell us the amount of noise/ripple and voltage regulation with this SMPS module. Lots of people criticized SMPS is such an evil.
Johnny
Johnny
Simulation is all well and good - implementation of a high power SMPS isn’t as straightforward. Bet those 48 volt units didn’t work well enough to get safety certs the very first time. Probably took some tweaking on the bench to optimize the gate drive waveforms, and multiple PCB iterations to deal with all the unwanted inductances.
My first real dummy load was one of those metal cased ones. You can overload one on a heat sink with a fan, right? Nope. 600 watts into one and it spits fire, catching the carpet. The amp (one of my creations into a Bose1801 chassis with original transformer) survived just fine. Now I have 8 that look like yours. 33 ohms, from Tanner’s for 7.95 each.
The “evil” of the SMPS is those that George has are only good for 150 watts. Trying to use one for a 150W amplifier is fruitless - because of the 3X peak power draw that ALL amps exhibit. Using 13 of them to power a 500 watt amp is actually about right. At what he paid for them its a bargain. For wat they originally cost, probably a lot less so.
My first real dummy load was one of those metal cased ones. You can overload one on a heat sink with a fan, right? Nope. 600 watts into one and it spits fire, catching the carpet. The amp (one of my creations into a Bose1801 chassis with original transformer) survived just fine. Now I have 8 that look like yours. 33 ohms, from Tanner’s for 7.95 each.
The “evil” of the SMPS is those that George has are only good for 150 watts. Trying to use one for a 150W amplifier is fruitless - because of the 3X peak power draw that ALL amps exhibit. Using 13 of them to power a 500 watt amp is actually about right. At what he paid for them its a bargain. For wat they originally cost, probably a lot less so.