There's a lot more between the AJ and the J2 than just the outputs right?
So how can we be sure that the SQ differences are due mainly to ss vs mosfet ?
Not doubting anyone's findings - just curious ....
So how can we be sure that the SQ differences are due mainly to ss vs mosfet ?
Not doubting anyone's findings - just curious ....
Thank you everyone. The bias pots are installed but I've apparently overlooked any proper pre-adjustment recommendations. Will power up with the bulb in place and if it works, adjust the bias and offset as I've seen described in this thread. Thanks!
Well, if it was only based on AJ vs J2, than yes, but I tried them in the same topologies as well. Similar observations 🙂
Zen has kindly teed me up, so I'll offer a non-quantitative thought for now. My thinking combines what buzz and lhquam have to say quoted below. As I described in Part II, the shape of the amplifier transfer curve is governed by the properties of the transistor curves themselves, interacting and working against each other in a push-pull amp or against the load in a single-ended amp. At the risk of ignoring the speakers (already pointed out by buzz to be an essential element in what you eventually hear), the residual amplifier distortion is what I think is differentiating things, and part-to-part variation might explain some of the different experiences people are reporting. (Is it a coincidence that I get the most glowing reports about SS performance after sending well matched parts?) I have not looked at enough MOSFETs to see how much their curve shapes vary, but I know I see some in the SS JFETs for sure. Also, we usually only speak of the distortion in terms of the magnitudes of harmonic components, but I think it is essential to consider phase distortion, especially when we are talking about stereo, to understand "sparkle, clarity and space." That's something I have not quantified yet, and care must be taken in any case because magnitude, frequency and phase become abstract mathematical concepts when discussing complex signals in real amplifiers that are unavoidably non-linear, however slight. Might hint at that in Part III (phase delay is an unavoidable subject when talking about negative feedback) and try a Part IV if I get enthralled. After I was blown away listening to the spatial coherence of my F2J's (when biased right) I came to suspect that this is something to reckon with.
Too many variables, as Ilquam says. Perhaps everyone thinks mosfets sound better, but they are a higher distorting part than SS, so perhaps folks like distortion. I do. You can play with sonic signature of these amps too much to do side by side comparison. Only way to know for sure is to build two amp using same circuit topology, equalized to same measurement specs, and then compare.
Some folks say Accuton are cold and clinical. Go figure.
Beauty of diy, build what you like and don't worry about it.
Could it be that both answers are correct?Perhaps the SS fets have (had) some much variability that different individual devices or batches have significant differences in input capacitance and other essential parameters.
Hope your doing well, SS Fan. I a sure enjoying my matched fets. I think that what I want most now is to visit the lab and use your equipment. Oh yeah, maybe take a little professional advice here and there🙂
Ha ha! Come on down. Although tomorrow would not be the best day to get a hotel room. 80,000 folks milling around the football stadium tailgating
while waiting for the #4 team to take the field. Folks in Starkville like their football anytime, but especially lately!
while waiting for the #4 team to take the field. Folks in Starkville like their football anytime, but especially lately!
If I ever find myself on the west coast I will most certainly have to come by. So many folks I need to see over in that foreign land.
on a more mundane note,
if you mounted an F6 pretty much inside (except the mosfets)
a 2-way sealed cab (4'-5' tall x roughly 1' square),
how far would woofer magnets
(2 peerless 830668 is the general idea, not monstrous, anyhoo)
need to be to avoid goofing around with the transformers?
and just how long a jumper could one run (without running into problems)
from the board to each of the mosfets?
if you mounted an F6 pretty much inside (except the mosfets)
a 2-way sealed cab (4'-5' tall x roughly 1' square),
how far would woofer magnets
(2 peerless 830668 is the general idea, not monstrous, anyhoo)
need to be to avoid goofing around with the transformers?
and just how long a jumper could one run (without running into problems)
from the board to each of the mosfets?
ac/dc multimeter won't give a hint?
mebbe with 3 bare wires and sliding contact for mosfet? adjust, listen and check readout?
adjustable mount for speakers to get near/far to bare circuit board? adjust, listen and check readout?
rinse and repeat?
mebbe with 3 bare wires and sliding contact for mosfet? adjust, listen and check readout?
adjustable mount for speakers to get near/far to bare circuit board? adjust, listen and check readout?
rinse and repeat?
not trying to be a smarta**,
just we're talking SQ effects here, right?
not recipe for magic smoke as long as I keep things within reason
given the kind rules of thumb already offered,
(and I don't slide the alligator clips around with power on)
or mebbe it is a bit dangerous for mosfets/other components?
just we're talking SQ effects here, right?
not recipe for magic smoke as long as I keep things within reason
given the kind rules of thumb already offered,
(and I don't slide the alligator clips around with power on)
or mebbe it is a bit dangerous for mosfets/other components?
25 W/8 ohm means ~2.5W / 100 ohm for starters unless total runaway,
pick LEDs appropriately,
and keep hand near power switch/eyes open from the start.
voltmeter either side of 1K steady or turn off immediately,
if nothing turn up test tone to check that LEDs work,
then turn back down below threshold,
watch awhile,
rinse and repeat for next test condition, if any?
pick LEDs appropriately,
and keep hand near power switch/eyes open from the start.
voltmeter either side of 1K steady or turn off immediately,
if nothing turn up test tone to check that LEDs work,
then turn back down below threshold,
watch awhile,
rinse and repeat for next test condition, if any?
Last edited:
LED on/off time will determine what oscillation freqs it can detect;
F6 (if it construct it perfectly 🙂 is spec -3dB at 50 kHz,
how far into MHz should I make sure I can cover?
F6 (if it construct it perfectly 🙂 is spec -3dB at 50 kHz,
how far into MHz should I make sure I can cover?
que ?
ya see that magic cap in series ?
it's ruining entire power equation , leaving it there just for real HF
however , your scoping (with real one or dummy , as enclosed small circ is) must cover well into MHz range