TheMGPremium
join:2007-09-04
Canada
kudos:1 Reviews:
 ·TekSavvy DSL
1 edit | Indestructible class AB audio amplifier design? There seems to be an awful lot of different variations on class AB amplifier design, so I thought I'd probe the minds of this forum to see if anyone has something to recommend.
Basically, I'm looking at putting together an amplifier capable of delivering up to 500W RMS into a 4 ohm load, 10Hz to 20kHz.
Audio quality should be fairly good, but doesn't need to be audiophile-worthy, the goal is not to get amazing sound quality, but it should still sound respectable.
It should also be extremely rugged. The amplifier should be expected to survive indefinitely to low-impedance and short circuit conditions without shutting down. Over-driving the amplifier into severe clipping should also not harm the amplifier (blowing up the speakers is ok on the other hand).
Why? Speakers aren't the only thing I'm going to be driving with this amplifier.
I'm not looking for a complete pre-done schematic, just a base design and ideas that I can build on and modify to suit my needs. |
|
Reviews:
·Frontier Communi..
| said by TheMG:It should also be extremely rugged. The amplifier should be expected to survive indefinitely to low-impedance and short circuit conditions without shutting down.
You mean that there should be no protection circuit, even for dead shorts?
A few (4 - 12) sets of output transistors in parallel with small serial resistors for equalization and short circuit protection.
Of course there are some other options like a magnetic or a tube amplifier, but I have feeling that's not what you're thinking about. 
--
Wacky Races 2012! |
|
TheMGPremium
join:2007-09-04
Canada
kudos:1 Reviews:
·TekSavvy DSL
| I was thinking more along the lines of current limiting built into the circuit, such that under an overload or short-circuit condition the maximum current the amplifier can deliver is limited.
I would like to do so with negligible effect on audio quality and efficiency during normal operation.
It is a very common form of protection for many linear and switching power supplies, so I thought the same might be achievable for an audio amplifier?
From the many things I've designed and built over the years, a discrete class AB audio amplifier is not one of them, so I've definitely got a lack of experience on the subject. |
|
|
|
Reviews:
·Frontier Communi..
| said by TheMG:I was thinking more along the lines of current limiting built into the circuit, such that under an overload or short-circuit condition the maximum current the amplifier can deliver is limited.
Current can be limited in several places in different ways. The standard way to electronically protect against shorts is described here: »users.ece.gatech.edu/~mleach/low···rot.html
but it actually cuts off the output transistors when overcurrent is sensed and that leads to a significant amount of distortion, which is bad if you listen to it, but at the same time it's good because you know that there is a problem somewhere.
I know there are 'soft-clipping' circuits and probably others that electronically limit current without actually cutting off the output transistors, but they'll probably be more complicated than what I suggested, or the circuit above.
--
Wacky Races 2012! |
|
public
join:2002-01-19
Santa Clara, CA | reply to TheMG
said by TheMG:There seems to be an awful lot of different variations on class AB amplifier design, so I thought I'd probe the minds of this forum to see if anyone has something to recommend.
Basically, I'm looking at putting together an amplifier capable of delivering up to 500W RMS into a 4 ohm load, 10Hz to 20kHz.
Class D would make more sense. Scales easily, protection is simple. |
|
TheMGPremium
join:2007-09-04
Canada
kudos:1 Reviews:
·TekSavvy DSL
| Hmmm this is getting a lot more complex than I first would have thought.
»sound.westhost.com/vi.htm
I'm still willing to tackle this challenge though. Starting to make some sense of all of this.
said by public:Class D would make more sense. Scales easily, protection is simple.
The theory is simple but the design and construction aspect, however, not so much, especially when dealing with switching frequencies 400kHz and above.
I suppose I could buy prefab modules, but those tend to be rather pricy. |
|
public
join:2002-01-19
Santa Clara, CA | said by TheMG:Hmmm this is getting a lot more complex than I first would have thought.
»sound.westhost.com/vi.htm
I'm still willing to tackle this challenge though. Starting to make some sense of all of this.
said by public:Class D would make more sense. Scales easily, protection is simple.
The theory is simple but the design and construction aspect, however, not so much, especially when dealing with switching frequencies 400kHz and above. I suppose I could buy prefab modules, but those tend to be rather pricy. Build 8 blocks, 100 kHz, 100W each, and phase shift the clock for total 800W 800kHz amplifier. |
|
| reply to TheMG
I could recommend something like the Philips EL6471, 1000 watt tube audio amplifier, shown here:
»www.youtube.com/watch?v=btSg2SX3lUs
Unless you have issues with projectiles in the area. But at least it's EMP proof. :_) |
|
 dandemanPremium,MVM
join:2001-12-05
Chapel Hill, NC
·AT&T Southeast
4 edits | reply to TheMG
Find a good used Crown DC300A or DC300A Series II (manual linked below) ... indestructable.. the final test at the factory includes operation into a dead short...
This amp was actually labeled as a "laboratory power amplifier" in that it was all DC coupled and had to adjustment capability to make sure it was zeroed when the two sides were operated together in "bridged mono mode" to be used as a servo amp.
When strapped in mono mode.. it can deliver 600+ watts into a 4 ohm load..
Bought one new in 1980, still going strong...
»www.crownaudio.com/pdf/legacy/dc300a2.pdf |
|
TheMGPremium
join:2007-09-04
Canada
kudos:1 Reviews:
·TekSavvy DSL
| reply to TheMG
Main reason I was thinking about building instead of buying is that this will be a fun project and I have most of the parts I need, aside for the final output transistors. I have an assortment of transformers, heat sinks, bridge rectifiers, small-signal transistors (mostly TO-92 style) of a couple dozen different kinds, and of course resistors, caps, etc. |
|
| said by TheMG:Main reason I was thinking about building instead of buying is that this will be a fun project and I have most of the parts I need, aside for the final output transistors. I have an assortment of transformers, heat sinks, bridge rectifiers, small-signal transistors (mostly TO-92 style) of a couple dozen different kinds, and of course resistors, caps, etc.
you're talking around +and - 120V and 1000W for the power supply, since it's AB. i haven't seen any schematics for that kind of wattage(500W) AB power amp around the net. a tube amp would be simpler and more rugged though. |
|
| reply to TheMG
i found the schematic for the EL6471 1000 WATT amp. »dl.dropbox.com/u/25758490/EL6471.pdf and main page with other amps. »members.chello.nl/m.janssen36/Am···lves.htm |
|
| Where can someone get an output transformer for a 1000W amplifier?
--
Wacky Races 2012! |
|
TheMGPremium
join:2007-09-04
Canada
kudos:1 Reviews:
·TekSavvy DSL
| reply to iknow
said by iknow :you're talking around +and - 120V and 1000W for the power supply, since it's AB.
+/- 63VDC, plus some extra to compensate for ripple and losses.
I have a 1000VA 100V CT transformer kicking around that should do the trick. Should provide around +/- 69VDC. Can bump that up to about +/- 75VDC if needed by tapping it differently.
Also got some nice 40,000 MFD 100V capacitors (5 of them).
Power supply is not an issue. |
|
| said by TheMG:said by iknow :you're talking around +and - 120V and 1000W for the power supply, since it's AB.
+/- 63VDC, plus some extra to compensate for ripple and losses. I have a 1000VA 100V CT transformer kicking around that should do the trick. Should provide around +/- 69VDC. Can bump that up to about +/- 75VDC if needed by tapping it differently. Also got some nice 40,000 MFD 100V capacitors (5 of them). Power supply is not an issue. that's not the way it works though. you have to take into account losses through the transistors.. |
|
sk1939
join:2010-10-23
france
kudos:6 Reviews:
 ·T-Mobile US
 ·Verizon FiOS
| reply to aurgathor
said by aurgathor:Where can someone get an output transformer for a 1000W amplifier?
Probably by having it purpose built. |
|
| reply to TheMG
here's some examples of the voltages needed. »www.diyaudio.com/forums/solid-st···mps.html |
|
sk1939
join:2010-10-23
france
kudos:6 | reply to disconnected
I would be terrified of turning that thing on, much less using it. Think about it's power draw too, if it's 1000W output, at 60% efficiency, that's almost 1700W of power draw just for the output, not to mention the tube heaters. |
|
TheMGPremium
join:2007-09-04
Canada
kudos:1 Reviews:
·TekSavvy DSL
| reply to iknow
said by iknow :that's not the way it works though. you have to take into account losses through the transistors..
Thus why I said I'm able to achieve up to +/- 75VDC.
That would theoretically allow for up to 12V of losses at the peaks.
Anyways, it is not absolutely critical that it can achieve 500W output. That's pretty much just a ballpark figure. If it performs a little lower I'm ok with that.
I'm not willing to put any serious money into this, so I will have to make do with available parts, especially for such an expensive part as the transformer.
Anyways, VI limiting seems to be the way to go as far as protection is concerned, so I think this is what I'm going to do. Will start off with a basic class AB schematic with VI limiting and work my way up from there. |
|
