F4 power amplifier

Jump to Latest
#861
Re: Re: F4 power amplifier

Nicola said:

Being ignorant in the matter of electronic, where could I read some literature that would explain me what is a impedance amplifier and how does it works. Same would apply for the current amplifier although texts from Mr Pass on F1,F2 are helpfull

THanks in advance
Click to expand...

Don't under-estimate Nelsons texts.
MC
 
#862
GRollins said:

Safely?
I confess that I don't understand how this amp has caused so much consternation, just by having unity gain.
Let's back up a moment and look at some simple numbers. Nothing kinky. Nothing weird. Just some realistic numbers.
An average power amp has something around 26dB gain. Is there anything sacred about this figure? No. It's just a semi-conventional figure that the industry tends to use. It works out to amplifying the signal by 20x. The Aleph amps have somewhat less gain--20dB if I recall correctly. These days a preamplifier generally has something between 10dB and 20dB of gain. Just to keep things simple, let's assume you're using a CD player (which generally has a fairly hot output) and that your preamp has 14dB gain (5x). Now rouse a few brain cells and do the math: 26dB plus 14dB equals 40dB total available gain.
But--being possessed of super powers--I happen to know that you never use that much. You'd have to crank the volume knob all the way clockwise, something few people ever do. Actually, it's likely that you end up in the 10 to 12 o'clock position...and that's on loud days.
Now, here's your homework: Put a known signal into your preamp and set your volume knob so that the output equals the input. For example, if you put in one volt, then set the volume knob so that you're getting one volt out. Put a piece of masking tape next to the volume knob and make a little hash mark at that point. This is 0dB gain. Your volume knob is now calibrated. Kinda sorta. For extra credit and since you've already got the masking tape in hand, go ahead and mark a few more useful places, such as +3dB, -3dB, +10dB, -10dB...etc.
If your preamp is at -6dB gain at what you regard as a reasonable listening level and you use an Aleph amp (20dB gain), then you're using around 14dB overall gain in your system. This tells you that you'll need a preamp with that much gain and with enough voltage swing so that it doesn't clip when driving the F4 rail to rail. In reality, it wouldn't hurt to have a little extra.
Okay, time for a reality check. The Aleph 1.7 preamp specs show 22dB gain balanced/16dB unbalanced. That sounds pretty good. How about the voltage swing? A quick perusal of the specs shows 20Vrms balanced/10Vrms unbalanced.
Given that the stock F4 is quoted as swinging +-20V, and given that you know it has unity gain, you'd like to be able to swing +-20V (40V overall) from the preamp--that means peak voltage, not RMS.
A little quick math shows that the preamp can swing well over 50V from peak to peak.
Spend a little more time with your thinking cap on and you can answer these questions for yourself. Questions are always welcome here, but people like to see that you're actually putting some effort in, yourself, rather than spraying questions all over the place because you're not thinking things through.

Grey
Click to expand...
I agree regarding this kind questions but I must confess I like your answer.
Thx
 
#863
I spend most of my two channel listening time in what I call
the "4 watt window" on my scope, which is the +/- 8 volts
that you would expect from a 4 watt amplifier - this corresponds
to about 5.7 V rms. Pretty much any preamp can deliver that.

😎
 
#864
Nelson Pass said:
I spend most of my two channel listening time in what I call
the "4 watt window" on my scope, which is the +/- 8 volts
that you would expect from a 4 watt amplifier - this corresponds
to about 5.7 V rms. Pretty much any preamp can deliver that.

😎
Click to expand...
If I may agree.
Before I change to Aleph5 I had a proton Amp with Vue meter. The needle never goes even further than 4-5watt.🙂
MC
 
#865
If it weren't for classical, I'd get by on comparatively modest amounts of power. Organ pieces, in particular, can be black holes for wattage. Yes, I'm quad-amped, so that isolates the rest of the system from the mayhem going on down in subwoofer territory. Want dynamics? Classical is the place to be. Nothing else qualifies. Takes power to do that.
Love Tchaikovsky's Fourth...light pizzicato on the strings, dancing, lilting, lulls you into a false sense of security. BAM! All hell breaks loose. The entire orchestra is on a take-no-prisoners rampage. First time I heard it, I fell in love with the piece. Say what you will about the performance/interpretation, but Slatkin on Telarc goes for the jugular at that transition. Yes!
Jazz? Huh. Realistic listening levels on jazz don't take gargantuan amounts of power, even on relatively inefficient speakers (mine are in the mid-80dB range). It's a non-issue.
Rock? Zilch for dynamics on 99.999% of recordings. The few recordings in that .001% still aren't in the same league as a good classical recording. Other than that, it's mainly a question of how loud you intend to play and how (in)efficient your speakers are. For my part, I don't listen anywhere near as loud as I used to. Hell, played Zeppelin this past weekend on the secondary system in the living room and never came close to stressing a 100W/ch amp.

Grey
 
#866
I've had a couple of inquiries about the Telarc Tchaikovsky from people who were unable to locate it after reading my mini-review above. That's because I goofed. It wasn't Slatkin, it was Lorin Maazel conducting the Cleveland Orchestra. Stock number 10047. I've got it on vinyl. No idea whether it's available on CD, but they'd be fools not to have it.
Mea culpa.
Now, go forth and listen.
Have fun.

Grey
 
#870
Sorry, I'm only jumping in here for a short question without have read the whole thread.
Is it a better choice to use a balanced driven F4 with the X1 or perhaps an X-poweramp (one with excluding the output stage from the feedbackloop and a quad-sym front end) with a volume control at the inputs?
Both configurations under same bias and voltage conditions and perhaps reduced closed and open loop gain by the X-poweramp front end?

I can imagine that Mr. Pass has tried it out.

Is there some information out there, which I can read?

Dirk
 
#873
BOM fixed.
The mistake was concerning R3 and R4 : 22.1R - R8 : 22.1K
C5 - C6 caps are optional, printed on Peter DANIEL PCB for PS filtering.
According to Mr Nelson PASS User Manual r0 - 4/17/07.
Many thanks to you Mr PASS !
Value for LED is supposed to be thrue... maybe someone to confirm ?
 

Attachments

#874
Question
According to Fairchild's cross reverence engine
The following fet's are equivalent to the standard IR parts

IRFP9240 = FQA12P20
IRFP240 = FQA19N20C

There are two reasons for the question
Somewhere on the forum some one sed that Nelson actually uses the Fairchild part’s

The Fairchild parts are much cheaper via my local distributor
 
#875
korben69 said:
BOM fixed.
The mistake was concerning R3 and R4 : 22.1R - R8 : 22.1K
C5 - C6 caps are optional, printed on Peter DANIEL PCB for PS filtering.
According to Mr Nelson PASS User Manual r0 - 4/17/07.
Many thanks to you Mr PASS !
Value for LED is supposed to be thrue... maybe someone to confirm ?
Click to expand...
There is one more typo. Q11 should be TL431.
Thank you for maintaining the BOM, and of course, Nelson and many others, too.
 
#877
GRollins said:
In an attempt to short circuit the inevitable questions:
--Yes, you can increase the bias as long as you have sufficient heat sinking/power supply/etc.
--Yes, you can decrease the bias.
--Yes, you can substitute the usual MOSFETs for the IRF outputs.
--Yes, you can substitute the input JFETs, but the field of candidates is much more limited.
--I don't have the datasheets for the input JFETs with me at the moment, but I don't think you're going to want to increase the rail voltage much. The outputs can take it, but the inputs can't. If someone knows where to find 100V low noise, complementary JFETs with reasonable current capability, I'm all ears.
--Yes, you can decrease the number of output MOSFETs, but watch the heat dissipation per device.
--Yes, you can increase the number of output MOSFETs, but bandwidth and distortion will suffer as a result--though not fatally.
What'd I miss?

Grey
Click to expand...

Hey, the F4 is a great concept. You could of course increase the rail voltage, in principle run it in class B or G or whatever. But as it is basically an emitter-follower (source-follower, OK then) it has gain of slightly less than one. So it would need a large-swing preamplifier to be of any use (with high rail voltages).
This means simplifying circuitry, really. One may say that the traditional input stage of the power amplifier is moved to the preamplifier. In other words less cascading of gain stages.
With a preamplfier able to swing +/-40 Volts you will get quite some wattage out from an emitter-follower. 🙂 Yes, 100W in 8 ohms.
BTW, all this might be said in the tread somewhere, but I just could not read through all 93 pages after Nelson's original post, so please excuse me for just dropping in like this.
Anyway, nice thinking, NP.

Rolv-Karsten
 
#879
One could use the concept to increase the current capacity of whimpy 100W or 200W power amps, so the whimpy one could provide the - hopefully - clean voltage - and the huge follower could provide the muscle. I MUST have heard about a similar philosphy many moons ago... somewhere.... ;-)
 
You must log in or register to reply here.
Top Bottom