Article: Beast With a Thousand Jfets

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
this year(2013) they said:D
and they know it very well:) the fets are maked as audio fets. maybe papa gave them a tip about this beast:D

Buying a 100 of the above fets. Will try them in PP buffer to see how they perform against Toshibas. Maybe good enough for input stage of power amps, leaving the Toshibas for small signal stuff. I think Linear has been releasing these thigns for about 5 years. Their 170's compare pretty well from what i have heard.
 
I can't really understand that there is NOT a marked for 2SK170 and 2SJ74, or the Linear Systems counterparts. I guess Parasound, Ayre, Pass Labs/First Watt, and probably several other audio corporations have bought shiploads of these JFETs now, to be able to build amplifiers for at least one decade. Yes, I know, in a world where you have to sell billions, a few 10.000s of JFETs does not matter. Still, I think it could be good business to sell these. Maybe if they repackaged them. Maybe SOT23 SMD or something.
Well, let's hope Linear Systems can do them.
 
understand

Lamborghini had a special tire developed by Pirelli in the '80s for their 4wd monster.
Apart from the huge width and the very high payload number, it had a wide edge on the side that prevented the tire from digging itself in on very loose sand terrain (as abundantly found in desert surroundings)

Extremely expensive, a set of four would set the owner back the cost of a compact automobile, and they didn't last very long with highway use.
Production of the super-4wd lasted half a dozen years, and Pirelli stopped manufacture of the Scorpion BK tire due to lack of demand.

Nowadays Pirelli manufactures a tire for various full-size 4wd's, even so bold to use the same model name.
But not in the original width, not for the same payload, and it's sand edge is pure cosmetics.
Pirelli will still have the mold of the original lying around somewhere, I'm positive that they'd be willing to manufacture a custom series if someone ordered a thousand at the right number of $/pc.

Similar deal with JFETs, not enough exotic 4wd makers around to justify the deal.
Not unless someone has access to the mold on the cheap, without the initial development cost, and 2nd stage development plus manufacturing cost levels enable him to cater to a niche market.
 
In that case, here's an original =>
 

Attachments

  • Scorpion tire.JPG
    Scorpion tire.JPG
    85.6 KB · Views: 477
Hehe, When I think twice, seems that, even if we had a bag full of Jfets, few among us would have the time and patience to solder, solder and solder...what comes to mind is a prefabricated board with 2000 smd jfets each. Then you could go and build amps with 1, 2, ...10 boards..Mega beast bridge/parallel with 20000 jfets. Fiction? MUAHAHA.... :devilr:
 
I built a headphone amp using the OPA2134PA which sounds amazing when paired with my Sennheiser HD280, so would paralleling a load of these produce the same when paired with some full size speakers?

What would be the best way to safely parallel these op-amps? Anyone got an idea component values and how much each you could drive each op-amp? You can get them for as little as $1 each if you buy from aliexpress.com: 10PCS/LOT OPA2134PA OPA2134P OPA2134 2134 TI/BB DIP 8-in Integrated Circuits from Electronic Components & Supplies on Aliexpress.com, minimum order is 10, but as loads are needed, no problem :)
 
OK. I found this circuit about paralleling op-amps:

Increase output current of the op-amp IC LM324 in 85mA size | Eleccircuit.com

OK, do my calculations work:

Now, if we had say a +/-18V supply, our input pk/pk was 2Vrms, could we set up each op amp with a gain of say 12 and use a resistor on the output of each to limit the current and parallel them all together? To limit the current to 25mA, we could have an output resistor on each op-amp of (18v / 0.025A) = 720ohms.

If you want to power a 4ohm speaker:

Power = I² R, so I² = P / R, so I = sqRoot (P / R), to get a 20W rms, we would need sqRoot (20 / 4) = 2.236A

Total current required is 2.23 amps, each op-amp will give us 0.025A, so 2.236 / 0.025 = 89.44 (45 x dual op-amps) per channel.

Would this work, and do the calculations look right?
 
Just re-read the data sheet. The OPA2134PA can give +/- 35mA at the output, so we could safely use it at 30mA, so you could 560ohm resistors on each op-amp output, giving you 18 / 560 = 32mA. If you used 50 dual op amps per channel, you could get a max of 3.5A (100 op-amps in parallel), that would be 3.5² x 4 (P = I² R) = 49W into 4ohm load.

Any thoughts?
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.