For sheer "ooomph!", put up your best MOSFETs against the venerable old BD 249/250 C, and let's rock.
The only thing TI botched up with that magnificent series is that the C version will handle only 100V. They promised D for 120V and E for 140V, but to the best of my knowledge, never delivered. Ditto for the TIP series.
Scusi signori, but a transistor capable of delivering 25A continuously, and 40A in peaks (50A if it's by SGS-Thomson) is a power amp designers's wet dream. That notwithstanding, they also have very uncommonly short Ton, Tstore and Toff times as you would never expect from such devices - not on the MOSFET level, true, but approximately 3 times faster than any other BJT device of their day. Only 125W rating, but cheap, so one can series/parallel almost at will.
The only thing TI botched up with that magnificent series is that the C version will handle only 100V. They promised D for 120V and E for 140V, but to the best of my knowledge, never delivered. Ditto for the TIP series.
Scusi signori, but a transistor capable of delivering 25A continuously, and 40A in peaks (50A if it's by SGS-Thomson) is a power amp designers's wet dream. That notwithstanding, they also have very uncommonly short Ton, Tstore and Toff times as you would never expect from such devices - not on the MOSFET level, true, but approximately 3 times faster than any other BJT device of their day. Only 125W rating, but cheap, so one can series/parallel almost at will.
My old (18 years) Kit 150wpc mosfet amp will still drive a 1 ohm load easily (single Bose 901 drivers to see what they can handle.. bucket loads BTW)
"Easily"?
How "easily" and to what power levels (i.e. output voltage)?
We are talking tens of Amps here, so please forgive me for being doubtful about the word "easily". No amp drives 1 Ohm loads easily, unless it has been specifically designed for it.
Many years ago, Europe's "Elektor" magazine did a design for just that, a very high quality low impedance power amp, however, it was designed straight off the drawing board for it and had appropriate power supply lines for it. I'll dig up my archives, see if I kept it.
"easy" as in putting the much maligned bose 901 driver (1 ohm)on a small baffle to see what it sounds like as a mid driver.. ear splitting volume 300 - 4k, hardly turned up on the amp, no "audible" clipping distortion at normal listening level, after about 5 minutes testing with music ,with the driver inside a closed sealed box(I value my hearing) an infrared thermometer pointed on the output fets showed up hot ( but not like when I drive them hard with sine waves testing PA sub drivers)... i.e. the BOSE drivers don't need to be driven hard because they are very sensitive🙂 four wired in series has about 99dB for 2.83 Volts (4 ohm, 300Hz- 4K)
Not good enough, I'm afraid.
300 Hz is about the borderline (though I think it's over the bordeline) between bass and mid, where the required energy from the amp is already way lower than it is at say 40 or 50 Hz, not to veture down to 20 Hz. Also, as I understand it, you drove just one channel - not realistic, is it, you'd have to drive both at once.
This is why a few active loudspeakers using driver fuse protection often have 20A fuses for bass, 14A fuses or less for mid and 10A fuses or less for treble.
I'd like to know how the amp fares when you drive it relatively hard, say 20 Vrms (28.3 peak), into 1 Ohm loads. That's equivalent to 50/100/200/400 Watts into 8/4/2/1 Ohm.
Last edited:
At least your ears still work, if nothing else .... 😛
This is not really a good test , subject both channels and listen to it , soa shows up in 10-15 mins ...
Yep, what were they using for outputs , only 3 pr , 36V psu ...
No, Jacco, that's their mid power amp, rated at 50W/8 Ohms nominally. I know it well, it's very popular locally. Uses 90W devices (BD911/912) at the output, for low impedance loads you'd have to use at least 200W devices.
This was a far more serious project, costing a lot of money to make, and it came out (if memory serves) about mid 90ies, along with a different version of the same amp, but more appropriate for us who do not own short circuit speakers like Wayne, with nominal impedances between 8 and 4 Ohms.
Years later, they published another High End design they called The Gigant, but that was a completely new project from the ground up (app. 250 Wrms into 8 Ohms).
"Not good enough, I'm afraid'.. Is this some sort of a competition ? like I said it was easy, my amp didn't overheat, didn't turn it self off, didn't blow a fuse, didn't crap itself I did it about 12 years ago while I was building a speaker system for someone, no doubt If I was going to run the amp full range both channels driven with low sensitivity 1 ohm speakers it would have quickly complained, but I didn't do that did I? struth.. I wont even mention the time I tested my DIY ribbon tweeter (< 1 ohm) on the same amp
Just an example, and not exactly a good amplifier.
Total Pd of the output stage is about 9 times the continuous power in 8 ohm. (e.g. the Gigant design does less than 5)
Beyond 30-40Vdc rails, life becomes much more difficult for bipolar output stage devices (beta and SOA)
(btw, best garantee to blow up a Perreaux lateral MOSFET PMF model from the '80s is by enlarging the capacitors in the power supply. Frank is demonstrating his vast knowledge bank again)
Of course it's no comeptiton, my point was that the test you described is at best a quarte of a test.
Under normal circumstances, a power amp would drive TWO speaker boxes, thus doubling the load on its PSU. Furthermore, for the test to be telling at all, one would have to drive it real world music signals, where peaks would be somewhere between say 14 and 25 Vpeak, depending on the type of music, speaker efficiency, room volume and desired loudness. NONE of these factors was in your test. Let lone mentioning what were the distortion figures.
In other words, many amps will drive a 1 Ohm load at 2.83 Vpeak, that's just 2.83 Amps. Many will do double that, but that's it, they will give up afterwards, at least most of them.
Let's talk about 20 Vrms across 1 Ohm, that's 20 Amps. Let's see your amp do that in steady state regime, for say 5 minutes. I assume you will agree that this is the minimum requirement for a decent home amp.
Do remember that IHF standards assume you run the amp at 33% of nominal for 30 minutes before you measure maximum steady state power ratings. If we set 20 Vrms as our nominal power into 1 Ohm, 1/3 of that voltage is 7 Vrms, try that for size, and see whether your amp's overheat protection fires or not, bacuse it WILL become very hot ater just 5 minutes, let alone 30.
Last edited:
Exactly! Quite so, Jacco.
Low impedance loads demand lower PSU voltage rails. It is of course possible to build an amp which will do 100/200/400/800 W into 8/4/2/1 Ohms, but it will costs a hell of a lot because one will have to use many output devices of significant power, exceptionally stiff power supplies, vast trasformers with a per channel rating of 1,000 VA each at least, mammoth electrolytics, etc. I hate to even think of the problem of cooling that thing down.
The basic question never was can it be done, of course it can, the question is which bank will give you that kind of loan.
For around €30,000, Milan Karan of Karan Acoustics will build you an 86 kg of an amp which will deliver around 2,400 Watts into 1 Ohms. Made to order only. Most are happy with dinky little toys delivering 250 and 400 Watts into 8 Ohms.
Last edited:
As brought forward by dutchie , good to look at Beta droop when deciding number of outputs, if memory serves me correctly 5amp on the avg per driver was typical for best hFe , making our 7 pr output choice pretty feasable and easy for 25 amp capabilty and good SOA ...
I would love to build as PS audio did with their amp with seperate stages ...
http://www.psaudio.com/wp-content/uploads/2011/11/C250-manual.pdf
Dual supply , high voltage vas (60-75v) and 40v output stage ...
I would love to build as PS audio did with their amp with seperate stages ...
http://www.psaudio.com/wp-content/uploads/2011/11/C250-manual.pdf
Dual supply , high voltage vas (60-75v) and 40v output stage ...
Now isn't that strange ... the Perreaux is still well and truly alive, in fact it sits in a box very close to where I'm sitting now, but hasn't been used for a while - playing with other ideas has surmounted interacting with that fellow. The insides are an amazing shamozzle of caps, and bits and pieces, a technician would have a nervous breakdown - it used to kept running constantly, I shudder to think what an odometer reading would be. Plus, it was regularly hammered hard in listening sessions - but I was very attentive to the temperatures it was enduring, had a thermocouple on hand and regularly monitored the temperatures of heatsink, and within. One of worst heat generators was the transformer, a monster EI unit, this got stinking hot just doing normal duties ...
There is a little bit more to it than "enlarging the capacitors", it does help to understand what's going as well ...
With power supplies there's the sledgehammer approach, and then there are all sorts of intelligent ways to provide clean power, just what's required when the music demands it. If you want something to heat the lounge in winter, just buy a bar radiator.
The real solutions to audio sound are totally engineered, and balanced, approaches - not "gotta get the best type of such and such" or "anything less that this expensive is pure junk" thinking ...
I hadn't listened to music for bit, for a variety of reasons ... and just fired up the PC setup yesterday, to do a comparison test. I was reminded, as I contantly am, of what gives me the kick in this game - because this simple unit injected that sudden jolt of surprise when listening, a punch of sound comes through that's tremendously uplifting - the complete opposite of the dreariness, almost tediousness of much hifi sound ..
AAH, from president/owner of PS audio and the Threshold corporation to canning beer.
Good ol' Randy, where would we be without marketing moguls.
(or was that mongols ?
)- Status
- Not open for further replies.