Coming right after the SCG. It has turned out quite well.
I'm working on this. If you want to use different devices, I'd suggest using ones that have similar characteristics as J175 and J111. The gain FET has to have a high Vp, that is the main criterion and the input one has to have moderate Vp. But it would be hard to say what would happen with other devices. They would work for sure but may not produce the harmonic profile as the one with the J175/J111 combo. I would suggest trying this combo first and then popping in other devices to see how they sound.
As for operating points, let me post that after the gerbers. I'd like to get those done first.
Eagerly waiting for the latfet MOFO.
Thank you in advance for all the infos, my only concern was about the need/necessity to purchase a (relatively large, 100pcs) batch of each of the JFET types in order to select the proper ones. J175 are getting quite difficult to find (Mouser doesn't sell them by now and from other vendors the price is crazy higher). It's time to sit down for a while waiting for further evolution...
Guido
Thank you in advance for all the infos, my only concern was about the need/necessity to purchase a (relatively large, 100pcs) batch of each of the JFET types in order to select the proper ones. J175 are getting quite difficult to find (Mouser doesn't sell them by now and from other vendors the price is crazy higher). It's time to sit down for a while waiting for further evolution...
Guido
I see they do not expect more until February.
I would worry about heat with SMD. Rahul is not afraid to push these devices to their limits.
I got one hundred of them just to see how much they varied. Nothing like J113. I do not think you need to get 100 of these.
Teabag has PF5102 which is a variant of J113 for sale with his SALAS Ultra BiB kits. He might be willing to sell you some. Not sure if this works as well with Rahul's circuit.
Has the J175 displaced the J176 for the buffer?
I would worry about heat with SMD. Rahul is not afraid to push these devices to their limits.
I got one hundred of them just to see how much they varied. Nothing like J113. I do not think you need to get 100 of these.
Teabag has PF5102 which is a variant of J113 for sale with his SALAS Ultra BiB kits. He might be willing to sell you some. Not sure if this works as well with Rahul's circuit.
Has the J175 displaced the J176 for the buffer?
That is very kind of you, but if you think it’s not worth it, I will not waste your time.
Didn’t see that question, Rick. No, it is still J176 in the buffer. If I spent enough time, I am sure I could make J175 work but it sounds really good with J176 and I’m not changing it.
Just found a vendor on ebay from Berlin, price is fair (<10Eur shipped) for 2x10 pcs j175 and j176, low postage fees for EU, it worth a try. Hope it's a good bet...
Guido
Guido
I have now secured all transistors for this new build with!!! Yeah!
That will be my only build of this end of year!
That will be my only build of this end of year!
Hi
I have not read the entire thread, but some of the things here are quite exciting and another point is rather scary.
One exciting thing is achieving triode-like curves from mosfets. Were those curves generated using a resistive load or a CCS?
Another exciting thing was seeing the jfet version of the line amp.
The scary part is the assumption that for the output buffer one can add together the Vds for two devices. That is patently incorrect. The buffer is a push-pull circuit despite people erroneously calling CCS-loaded circuits "single ended". The CCS actively pulls the signal downward when the upper device is heading towards being 'off'. As it is a push-pull circuit, you have to expect the signal to swing the full breadth of the supply, which means that EACH PUSH PULL device must be rated to withstand that full supply breadth. With this in mind, using 25V jfets means the supply at the drain of the J113 should not be more than 25V. If the devices have dissimilar Vds ratings, you have to limit the supply to accommodate the lower Vds rating.
You may be lucky so far if you have not destroyed output jfets with a too-high supply voltage. This may reflect the fact that if you are driving a unity-gain speaker amplifier that realistic power levels are actually quite low in most living rooms, so the signal swing at the buffer output is low, as well.
I have not read the entire thread, but some of the things here are quite exciting and another point is rather scary.
One exciting thing is achieving triode-like curves from mosfets. Were those curves generated using a resistive load or a CCS?
Another exciting thing was seeing the jfet version of the line amp.
The scary part is the assumption that for the output buffer one can add together the Vds for two devices. That is patently incorrect. The buffer is a push-pull circuit despite people erroneously calling CCS-loaded circuits "single ended". The CCS actively pulls the signal downward when the upper device is heading towards being 'off'. As it is a push-pull circuit, you have to expect the signal to swing the full breadth of the supply, which means that EACH PUSH PULL device must be rated to withstand that full supply breadth. With this in mind, using 25V jfets means the supply at the drain of the J113 should not be more than 25V. If the devices have dissimilar Vds ratings, you have to limit the supply to accommodate the lower Vds rating.
You may be lucky so far if you have not destroyed output jfets with a too-high supply voltage. This may reflect the fact that if you are driving a unity-gain speaker amplifier that realistic power levels are actually quite low in most living rooms, so the signal swing at the buffer output is low, as well.
It is not so dire in practice. Peak signal swings of 40-45V are achieved. I can see it on the scope. The ratings you mention are DC not AC. They are not biased at DC with more than 40V. However, I have seen them survive 70V DC just fine. I have not lost a single JFET despite running all kinda of torturous levels. I have broken their pins through bending but never lost a JFET for exceeding ratings—found them to be quite tough. For the final design, there will be about 30-35V across the output devices.