I think mine would be easier to do as it acts more like a VARIAC more than anything else. all those circuits are interesting, but what one thing that caught my eye is the actual data on HUF75344's they are only rated at 50V with a nice 75A potential still a bit "weak" as far as voltage, the IXTK 120N25P is rated at 250V and 120A and think it would be a wiser choice if it was to be used on a class G or H for controlling rail voltages that are near 120V or more amplifier friendly for those of us with rail voltages in excess of 75V.
Also theres a couple of versions of the SSR's that only require a 5V input and a potentiometer to set its output i believe they are in the $75 range and anyone can install them into their amp and basically may even elminate the need for a soft start as your limiting the voltage going to the toroid.
Most of you remember the old radios used to have the power switch right on the volume knob, and the same thing can be done here !
Also theres a couple of versions of the SSR's that only require a 5V input and a potentiometer to set its output i believe they are in the $75 range and anyone can install them into their amp and basically may even elminate the need for a soft start as your limiting the voltage going to the toroid.
Most of you remember the old radios used to have the power switch right on the volume knob, and the same thing can be done here !
wg_ski said:
Don't split ur drivers from the outputs - asking for trouble.
Not fast recovery you want soft recovery.
Ur still gonna waste money and space on large kva toroids instead of going paralleled? Tsk tsk.
_-_-bear
Adrculda said:
Those Fairchild hexfets are the lowest ON resistance in a single package I can get my hands on at the 55V level. They only need to handle the difference between adjacent rails - the QSCs switch a full 55V (not 47V) with a single IRFZ44. Biggest problem with that is the PCB traces for TO-220's are puny. They make 100V versions that have higher ON resistance. Those huge over-volatge IXYS devices will have much higher resistance per pF of input capacitance. The lower the VDS rating you can stand, the better that ratio.
If you're still bent on varying the primary side, study Carver's magentic field power supply. It ises a Triac controlled chopper to effectively increase the line frequency to the trafo - so it can be made smaller. You can also do regulation that way (which is what you're thinking of) - many of the old HP lab supplies do exactly that. Bear in mind you can't get the voltage to change *quickly* doing this. It will work fine for the time-average, similar to Crown's supply on the VZ series, but it can't track the waveform like a true class G. Carver's amps still use class G or H (and on some, a combination) rail switching in addition to the funky power trafo control.
I don't want to start a flame war or rain on anyone's parade, but I felt compelled to say this. In today's era of global warming, energy geopolitics and energy costs, don't you think it is irresponsible to be building monster class A amps? How much carbon will a large class A amp add to the atmosphere compared to a 40 watt AB or even a class-D? I myself would prefer to drive a large powerful car but I have retired my monster SUV and now use a small 2 liter sedan for my daily driving. It's a matter of responsibility I think. We should all do what we can to use energy as wisely as we can where possible. Some of the higher powered class D's sound quite good. I use a Class D myself now and I am an audio nut that is very satisfied with the sound quality. I think that we, as a community, must do our part and give up the old ways too. I'm sorry if this will be taken by many as being out of line, but I had to say it.Ecological responsibility must extend to ALL of our activities as an unthinking daily habit if we are to leave our descendants a livable planet.
bear said:
Can split the drivers from outputs if you track their VBEs separately. Same as separating the PNP from PNP. What most newbies try to do is just use one Vbe multiplier transistor and hope it's in the hottest spot. I love that Leach-style Vbe multiplier with four diodes on the base leg. One directly mounted to the NPN driver, one to an NPN output, one to the PNP driver, and one to an output. I've fixed a lot of amps with that.
Neither fast nor soft recovery are needed in the mains rectifier. One of the above is needed in a rail switch, but they don't have to charge 68,000 uf caps through them. You *could* use schottkys in the mains rectifier, but 170V units are as high as I've seen. That may not be enough. +/-85V rail is as high as you would dare go.
One big large toroid is more a waste of money than space. Personally, I run toroids in series because lower volt higher amp units are a lot easier to find surplus and already you have taps for voltage tiers. A bigger toroid is needed if you're counting on voltage regulation to prop up the rails. Big ones have lower inductance. If your goal is 2K at 2R (that is 4kW bridged at 4), you need +/-95V under load (89 for peak swing plus 6 for transistor drops). Now if you look at the RMX5050's current draw specs, 2kVA will do for an average output of 1/8 power at 2 ohms per channel. But if you start with only a 70-0-70 toroid, you won't keep +/-100V under that load. You'll end up with a smaller amp. With a 6kVA unit, you'd have a chance because the regulation might be enough. To use only 2KVA worth of trafo, we raise the rail voltage to +/-140 and let the sonofagun drop. It will fully flywheel (inductively filter, with zero ripple) at +/-89V, but that would be around 20kVA worth of load. I would expect it to be in the range of +/-95 to +/-105 at a "full load" of 28.5A for 2x2kW/4R. I load tested at 35A and it dropped to +/-95V. Beacause it was darn near flywheeling, the ripple was less than 6V p-p.
audiosteve said:
Nobody around here is considering class A at these power levels. It was a "newbie mistake" when the thread started. But even class B at these levels is probably ridiculous. I use class H in all my new designs over 200W because it seriously cuts down the heat and AC draw. Class D is just too expensive at kW levels and is on the hairy edge of being non-doable for DIY.
Now if you want efficiency - the best place to put that is the loudspeakers. Labhorns are 60% efficient in the proper grouping (6-8). Compare with 10% for a wall of 18" woofers.
audiosteve said:
I thought we were duty and ethically bound to seek out new life and new civilizations, and to boldly go where no man has gone before. Today's kW power amp designer may be tomorrow's warp engine designer. I'm sure that's how Zeke started out - he did like his music LOUD!
bear said:
Hi, Bear
I haven't gotten to the toroids yet, but i did take your advice with running twins 😉
i'm just having the rectifiers and caps laid out so i can have the boards made. each one will have its own rectifier and cap bank and that capacitance i gave was for 1 supply board.
Currently trying to find a local distributor that can source me 2 1.5KVA or 2KVA's with 90-0-90 secondaries.
wg_ski said:
And Again wg_ski with more useful info and insights.
looking at tring to find 90-0-90 if not i guess i have to take 70-0-70 units. And someone on here had a really nice forced cooling heatsink up For Sale that i think would have worked perfect for me as i would have ran it down the middle of the amp from front to back and place each toroid with is cap bank and respective amp on the respective side therefore reducing crosstalk between channels.
will be using XLR inputs and like the securitu of the Neutrik Speak-on's ( besides i have 8 10 gauge cables already 😉 )
Plitron's got 45-0-45's two in series would do the trick, but you'd have a common supply for both channels. I don't consider that a show stopper. But at $260 each, OUCH!!!! I have a problem with that. Apex Jr. has 90VCT's at 600 VA for $30. Four or 6 in series-parallel would be the bomb. Shipping up your way might be a bit high, but when looking at the alternatives it might not be bad. That's how you build these amps and not go broke.
wg_ski said:
Yes.. agreed.
Almost tempted to wind my own, because i know its not rocket science !!
But finding the cores and the rest of the stuff here locally can be time consuming. been poking and proding at suppliers around town that they almost know me by first name !!! LOL
And my PCB guy still swears im building a TIG Welder
You'll probabaly never find *cores*. But if you find a whole trafo that's high enough VA to make it worthwhile and it has the wrong secondary, pull the old one off and wind a new one. I would not recommend rewinding a primary - I always leave the original intact. I may add a few turns to put in series with a primary if I want to drop a couple volts, but I'll use regular insulated wire (not magent wire) for that.
wg_ski said:
Noted, and would solid insulated copper work for that ??
Secondly i just got the MOST OUTRAGEOUS quote for today!!
Hammonf toroid 1500KVA
182T60 1500 120V C.T. @ 12.5A 60V @ 25A
$295
3x 24000uF @ 100VDC
$310
50A Rectifier
$35
YIKES
thats just for one side!!
do they want my kidney too ??
$295 for a 1.5 kVA doesn't sound half bad.
The regular route to get such anomaly sizes by the cheap is NOS. I've got a quartet of 5KVA toroids manufactured for yacht shore-socket powersupplies, there's a lot of salvage material in the boat business.
The regular route to get such anomaly sizes by the cheap is NOS. I've got a quartet of 5KVA toroids manufactured for yacht shore-socket powersupplies, there's a lot of salvage material in the boat business.
sounds like you are getting fleeced...
http://www.antekinc.com/AN-152115.pdf
1500VA 120V 10.2A (conservative)
$160 US
The caps are about right.. I saw $60-80 APIECE for 100 -150v
15k - 25k uf mallory /CDE caps..
As far as bridges go ,HA.. $6.40 apiece for 50A/400v
http://www.mouser.com/Search/Refine.aspx?N=4932236+4294605305
Even for derating overkill -50A/1000V bridge..only $7..
@Mouser.com.
OS
Thanks for the info ostripper...
but i think this 1500VA Toroid is all i need plus it gives me a 2 12v coils that i could use a boosted voltage on the drivers and the current is manageable too on PCB levels 😉
Waiting for the 3rd supplier to get back to me with pricing on the 100V 15000uf caps.
Now ...
what size resistor do you guys use to keep them caps in check ??
but i think this 1500VA Toroid is all i need plus it gives me a 2 12v coils that i could use a boosted voltage on the drivers and the current is manageable too on PCB levels 😉
Waiting for the 3rd supplier to get back to me with pricing on the 100V 15000uf caps.
Now ...
what size resistor do you guys use to keep them caps in check ??
- Status
- Not open for further replies.