ok ok, I will post something more productive since you dont want to talk about your amp more.
what about a diy version of the PBN Audio Montana WAS?
That would be two 18" McCauley 6174 woofers per cabinet, a Scan Speak Revelator D2904/7100 tweeter, two 18W midbass drivers (the PBN uses 9" midbass drivers but I'm not sure the model), and two 15W midranges (I believe). Thats roughly 3000 in drivers, lets say another 500 or so for the crossover, and whatever it costs you to build the cabinets. I would say these would show off the capabilities of the amplifier just fine, the power handling would be around 1000 watts rms easy, pretty high sensitivity, etc.
what about a diy version of the PBN Audio Montana WAS?
An externally hosted image should be here but it was not working when we last tested it.
That would be two 18" McCauley 6174 woofers per cabinet, a Scan Speak Revelator D2904/7100 tweeter, two 18W midbass drivers (the PBN uses 9" midbass drivers but I'm not sure the model), and two 15W midranges (I believe). Thats roughly 3000 in drivers, lets say another 500 or so for the crossover, and whatever it costs you to build the cabinets. I would say these would show off the capabilities of the amplifier just fine, the power handling would be around 1000 watts rms easy, pretty high sensitivity, etc.
BlueWizard said:
curious steve, if this is the way to demo the amp capability, what would you suspect this demo to sound like at each stage if done with a 'normal' amp??
I mean there must be a difference that is being demonstrated here right?
Hmm, good question.
Frankly, one of the biggest advantages of being able to drive a low impedence load would be the ability to get a sealed speaker to go deep (less than 40HZ) in a small box.
That is an impressive speaker, but it probably does around 90db with full baffle step compensation at 2.83 V/m, with a minimum impedence of 3 ohms or so. You wouldn't need a monster amp to make that play loud.
(that is a total guess -- based on the two 15w's wired in parallel)
JJ
I know the PBN uses all passive crossovers, so it can be done, but my concern here, and with most large multi-way speakers, is that the bass crossovers will require very large inductors. The lowest DCR you will find would be the Toroidal inductors, which you can get at Parts Express. I would personally use 2nd order there to avoid having two inductors in the signal path.
The rest of the crossover would take some work and some planning. A lot of measurements. Oy, more work than I care to think about.
The rest of the crossover would take some work and some planning. A lot of measurements. Oy, more work than I care to think about.
more cabinet building than I care to think of--I just looked up the McCauleys -- I saw one set of measurement that said for a Qtc of .707, you need a roughly 7 cubic foot box (for one
)with an F3/Fcb of only 41 hz.
We can do better.
JJ
Not so much with low impedence -- your caps are what will need to be BIG. But electrolytics would be fine in that position...
jupiterjune said:
more cabinet building than I care to think of--I just looked up the McCauleys -- I saw one set of measurement that said for a Qtc of .707, you need a roughly 7 cubic foot box (for one
with an F3/Fcb of only 41 hz.
We can do better.
JJ
Thats actually really good for a sealed box with a Qtc. of .707
How would you do better?
terry j said:
For a normal consumer non-commercial non-Pro amp -
Stage 1 - one pair of speakers. Sounds fine everything is cool.
Stage 2 - TWO pair of speaker, total load 2 ohms. If the amp doesn't shut itself down, then I would expect electrical distortion and tons of excess heat.
Stage 3 - THREE pair of speakers, total load 1.33 ohms ....snap, crackle, pop...smoke...assuming the protection circuits didn't shut it down first. Most consumer amps would consider 1.33 ohms a dead short and wouldn't even turn on. If the did turn on, heat and excess current would shut them down quickly.
The Custom Amp in question, if everything goes right -
Stage 1, 2, and 3 - amp cruising among fine with no problem, no shut down, no excess distortion. I would expect a slight rise in distortion as lower impedances, but nothing off putting.
Glen:
Some question have been answered, but other still remain a mystery, and clearing them up will certainly help the situation.
First, what is the purpose of this amp?
A nice experiment in Amp design?
An amp for commercial PA or distributed speaker application?
A super stereo Hi-Fi amp?
Is this a private build, or is it a commercial venture, or hoping to be a commercial venture?
How you envision the amp being used matters. There are many commercial Pro amps that can source to very low impedance loads. There only purpose in home stereo is for the immense stability and reserve of power.
It is only in commercial ventures where the ability to routinely drive very low loads actually matters. I simply can't see any realistic stereo Hi-Fi application where it would be necessary. Though without a doubt, having that stability and reserve would still be advantageous.
Now, I could perhaps envision a situation where a audiophile could combine two 4ohm speakers into a 2 ohms load and want to be sure they could drive it securely. There are actually a few common consumer stereo amps that are 2 ohm stable. So that is not so far fetched.
But, in my view, the advantage of this amp for stereo Hi-Fi is in its rock solid voltage stability and it's huge reserve of current. NOT in its ability to actually drive ohm and fractional ohm loads. The fractional ohm ability is a measure and indicator of how well it will drive normal loads.
So, again, beyond commercial Pro applications, there is no need or purpose in driving ohm or fractional ohm loads. The ability is important to have, but using that ability routinely, is like I said, akin to driving you car in low gear all day.
Do you see my point? Just because you can drive fractional ohm loads doesn't mean you should do it beyond demo purposes, and again, I'm speaking of non-commercial applications here.
If you want a real short cut, the Cervin-Vega CLR215's which are US$500 each, or US$1,000 per pair (2x15", 6.5" waveguide Mid, 1" waveguide soft dome Highs) are rated at 500 watts. Three sets of those would be US$3,000 and could take all the power you could throw at them. Though the Cerwin Vega website doesn't specify the impedance, I would suspect 4 ohms since it has two 15" woofers.
http://cerwin-vega.com/CLS215.php
Again, I can understand using ultra low impedances for demo purposes, but not for routine daily use outside of commercial applications.
Again, do you see the point I'm making here?
Steve/bluewizard
BlueWizard said:
Pretty much just A and C.
BlueWizard said:
Hmmmm.... OK, but how about if you had a pair of speakers with ribbon drivers or one of the number of sought after high end HiFi speakers out there with impedance curves with resonant dips down to an ohm or less?
Another thing to keep in mind is the fact that, due to the nature of resonant circuits, an impedance dip down to 1 ohm (just for example) can draw significantly greater transient current than a resistive load of 1 ohm.
My amp will deliver 800W per channel into 1 ohm. That may seem af first glance a ridiculous overkill for a home HiFi setup, but then again, with high dynamic range material, peak power outputs of a few hundered watts are easilly achieved even with pretty listenable and moderate room SPL.
A guy call Bob Cordel did some measurements at an RMAF show and measured something like 280W peak in a hotel room with material playing at an average SPL level that probably wouldn't disturb the folks in next room too much.
Now if you were to throw into that mix a "difficult load" speaker with resonant dips to 0.25 times (just for example) the nominal impedance and a more spirited listening level, then the situation starts to become a lot more demanding on the poor amplifier.
It has been suggested as a rule of thumb that a HiFi amplifier should be capable of driving on a transient basis loads 1/4 that of the rated nominal impedance load for full rated power.
With a class A power rating of 200W into 4 ohms, I went one step better at 1/8 the impedance.
My amp will pump out 1600Wrms (3200W peak) and 100A peak into a low enough impedance load with the emitter currents of the multiple 16A output transistors only reaching 5A each.
This avoids beta droop non-linearity, and fT droop along with the attendant surge in junctions capacitances. WRT the latter, the output devices and driver devices are also voltage clamped to prevent saturation. In other words, the output stage alone will happily pump out this level of transient power with less non-linearity than a typical 100W max HiFi amplifier at the onset of clipping.
But this is probably steering a bit off topic.
Cheers,
Glen
G.Kleinschmidt said:
Absolutely agree with everything you said, but look careful at what you did say.
You aren't intending to hook 1 ohms speakers on for everyday use. What you want is the stability and reserve so the amp holds together when a normal impedance just happens to dip that low.
In other words, a sustained 1 ohm load is not the goal. The goal, in routine use, is for the amp to hold rock solid for transient 1 ohm loads.
Now, I conceded in my long rant that 2 ohm loads are possible for Hi-Fi consumers, and a few consumer amps (NAD i think) are 2 ohms stable. So, this isn't that far out.
For your own personal use, I really don't see a need to demonstrate the amp at massively low impedances. If there is no commercial venture aspect to it, then it is really $4,000 just to be able to say to your friends, 'look what my amp can do'. Seem a bit steep for that.
I think a far better application of the money is to look at this as designing the best possible speakers to produce the best possible music with your fantastic amp.
If you simply want to demo the current and load ability for friends, use a bank of power resistors.
Some serious research should turn up speaker systems that will do this amp justice musically.
Though, in some sense, I think we need an out of the ordinary speaker to do justice to this amp. So, finding the exact right speaker design is going to be more difficult, but when you find it, it will be worth it.
I wish I was there to hear it when you get it hooked up. That would be spectacular.
steve/bluewizard
BlueWizard said:
Steve, not as low as 1 ohm. I was hoping for a nominal impedance of around 2 ohms. That still meets the ¼ impedance rule. But I guess that may be unrealistic unless I throw a couple of ribbon speakers into the mix (haven’t looked into ribbon drivers yet or have any idea of their availability or cost).
The good thing about designing a speaker for this amplifier (even with a nominal impedance of 2 ohms) is that one is not required to go to great lengths to make the speakers amplifier friendly. I'm guessing this will relax the complexity of the crossover design somewhat.
Thanks for the compliments. It’s looking like finding / building a pair of suitable speakers is going to be as much (if not more) of a project as building the amplifier!
Cheers,
Glen
G.Kleinschmidt said:
Hi,
So the class A part is an irrelevance ?
FWIW you can make a speaker that is a pig to drive, its called poor design.
So not to second guess .....
One would think the point of the demonstration would be the
class A quality, not the fact DC-coupled with current source
feedback it would make a very good arc-welder.
first lets talk voltage levels :
Output appears to be +20dB i.e. :
100W into 8R class A
200W into 4R class A
400W into 2R class AB
800W into 1R class AB
etc...
The ~ doubling depends on the power supply dimensioning.
To demonstrate class A you need an 8 ohms nominal loudspeaker.
The amplifier will clip before it leaves class A (good 8 ohm design).
A 4ohm nominal speaker will leave class A near clipping.
FWIW a 4 to 8 ohm nominal speaker does not dip to 1 ohm.
Worst case dimensioning is ~ 1/3 nominal resistive.
i.e. for 8 ohms nominal a 2.7R resistive load, 4ohms - 1.3R.
(Anything worse is poor loudspeaker design.)
For good speaker design it is ~ 1/2 nominal resistive, i.e. every
attempt has been made to make the speaker easy to drive.
(And the manafacturer is not lying about its impedance)
If the amplifier is an "extravangance" to drive difficult loads class
A really does not come into it, it becomes a nicety at low levels.
Really difficult loads are due to the way the speaker is
built, electrostatics or transformerless ribbons usually.
For "standard" speakers very low impedance is very bad,
it causes all sorts of difficulties. There are speakers with
bad impedance dips, this is poor design, not an inevitable
consequence of the technology.
What are you trying to demonstrate :
The normal class A quality into a sensible load ?
The class AB quality into poor loads ?
I'd go for the first and then add parallel "powersoaks", either
resistive or modelled on real speaker impedances, how many
and what impedance they represent is up to you.
FWIW you could build a Watt/Puppy clone - 4 ohm nominal.
A 2 ohm nominal version is not difficult if desired.
The latter would use 4 ohm nominal drivers.
🙂/sreten.
Well, 4 4ohm drivers like the Peerless 830952 in series parallel should have a f3 and Fcb of 36.5 Hz in roughly a 240 liter box.
This would be "better" from a smaller box size standpoint and lower f3 standpoint. If you allow a lower impedence (2ohms nominal), you could use only two drivers and cut you box size in half.
Would the sound be better? Some think high Xmax subwoofer drivers don't sound as good as large "Hi-Fi" woofers. (I might be one of those people).
JJ
Make a pair of dipole monster subs with each 4 AE or Exodus drivers in paralel and apply lots of Eq...that may squeese the last juice out of your monster amp
Well, you will have to build another more normal amp fore your mains...anyway, fore top mids and tweet small amps with a single pair of transistors will do the trick nicely, you just need efficient and easily driven speakers fore that 😀
Well, you will have to build another more normal amp fore your mains...anyway, fore top mids and tweet small amps with a single pair of transistors will do the trick nicely, you just need efficient and easily driven speakers fore that 😀
sreten said:
Sure as heck don't see how you figured that out.
sreten said:
FYI there are a number of very highly regarded speakers out there with significant impedance dips. This has been discussed at length such as in the PGPamp thread on the solidstate forum.
I can't remember the model/brand numbers ATM. John Curl has quoted 0.5 ohms for his speakers (Watt/Puppy I think).
sreten said:
Well that is pretty incoherent. With a 5A bias current class A operation is pretty much guaranteed at low levels and beyond.
But I guess you are right. An amplifier of this specification really isn't extravagant at all.
sreten said:
I already stated quite clearly the intentions for the design in the very first paragraph of the ******** post that you quoted.
G.Kleinschmidt said:
Hi,
Well you obviously know so much you do not need any help.
Or to try to understand the point of a question.
You solid state guys obviously must know more about what
it takes to drive a loudspeaker than the loudspeaker guys.
🙂/sreten.
sreten said:
Not on designing amplifiers, no. On a decent 2-4 ohm nominal impedance high power speaker system, yes. That is why I started this thread.
Knowing what it takes to drive "difficult load" speakers of established design is not equivalent to knowing how to design good speakers, "difficult-load" or not.
jupiterjune said:
you seem to be confusing the value of the part with the power handling. Yes a low impedance load will use a lower value inductor, but my concern was the large inductor value and the power handling combined. Since this part would be in series, you would want the smallest amount of dcr you can get with a low chance of saturation. Any thing over 1mh for an inductor is large enough where the dcr can begin to become a problem. In this particular case I bet he will need an inductor over 2mh. Though the capacitors will be increasing in value because of the 4 ohm impedance, I agree, an electrolytic is fine, as the part is in parallel. The only potential issue is power handling, most electrolytic are rated at between 80 and 160 volts. I would much prefer to see the high end of that range for this purpose.
Wouldn't a simple way of reaching the goal be a line-array?
Something like a tweeter, one mid, then eight (8 ohm) bass drivers wired in parallel for a 1 ohm nominal load.
I think this would be easy enough to design (normal sort of 3 way) and would provide staggering bass output. These could be in sealed boxes for clout or open baffle for transient quality. Something like 8" bass drivers. Or 4 drivers for a merely "fairly ridiculous" amplifier load.
Simon
Something like a tweeter, one mid, then eight (8 ohm) bass drivers wired in parallel for a 1 ohm nominal load.
I think this would be easy enough to design (normal sort of 3 way) and would provide staggering bass output. These could be in sealed boxes for clout or open baffle for transient quality. Something like 8" bass drivers. Or 4 drivers for a merely "fairly ridiculous" amplifier load.
Simon
- Status
- Not open for further replies.