Hi Brian! No good reason other than I have plenty of them, and they have a good hfe. Cob is one of many things I have zero knowledge about.
The original transistor are not in my drawer, but maybe you could suggest other alternatives? I have a reasonable amount of "older" transistors...
The original transistor are not in my drawer, but maybe you could suggest other alternatives? I have a reasonable amount of "older" transistors...
So I think I see what you mean now. Distance from pin hole to PCB edge is about 8mm which gives plenty of space for mounting hole to be well off the edge.
The parts bin - That makes sense.
If you use those, might want to reassess compensation. With their higher Cob, I assumed that the compensation caps could be lowered. I don't have simulation models for those devices, otherwise I'd give it a shot.
Generally speaking, low Cob is preferred for the VAS. This is why @lineup and me both focused on SC3503/SA1381. These about the lowest Cob TO-126 devices available. But even now the last source of these (Onsemi) are going EOL. TTC004B/TTA004B are probably the next best alternative.
If you use those, might want to reassess compensation. With their higher Cob, I assumed that the compensation caps could be lowered. I don't have simulation models for those devices, otherwise I'd give it a shot.
Generally speaking, low Cob is preferred for the VAS. This is why @lineup and me both focused on SC3503/SA1381. These about the lowest Cob TO-126 devices available. But even now the last source of these (Onsemi) are going EOL. TTC004B/TTA004B are probably the next best alternative.
BTW - I like your idea for the jumper on the feedback cap. I'm going to add that to my version.
Thanks for sharing! I have neither of those, but here is a list of what I could use. Would you give it a glance to see if there are any good (or at least better) candidates for the VAS?
2SC3503E / 2SA1381E would probably be the best. I see you have 2SC3423 / 2SA1360 listed - those would also be very good depending on what VAS current you plan to run (due to their 50mA limit).
I'm not familiar with many of the others. I entered this hobby after the "golden age" of through-hole components.
I'm not familiar with many of the others. I entered this hobby after the "golden age" of through-hole components.
The VAS current according to rule of thumb, I do not remember who's thumb, should be about 10 mA. (I remember John Curl or maybe Bob Cordell in a thread about 10 or so years ago, I adopted that stance and it lived with me ever since) I guess it was for a normal home amp not running 24 transistors and 1500 watt output in class A!
I recall once building a BIG amp and the VAS could drive a set of speakers pretty loudly. 😡
I recall once building a BIG amp and the VAS could drive a set of speakers pretty loudly. 😡
Last edited:
I believe the regulator/noise reduction circuit, Q11/Q12 should be moved to the other side of the VAS....
And laterals with their nonlinear ciss like some drivers.... Or you can run the VAS with higher current....
And laterals with their nonlinear ciss like some drivers.... Or you can run the VAS with higher current....
Hi Soekris, long time.
The VAS section was fed from its own supply though (#313) , if I am not mistaken, but it is still a good point you make if the VAS is on the same rail as the output.
The VAS section was fed from its own supply though (#313) , if I am not mistaken, but it is still a good point you make if the VAS is on the same rail as the output.
Last edited:
Now I see that it is on the same rail as the output. Sorry. May I ask your opinion, Will you not loose output swing if the VAS is on the same power supply rail as the output, thus you are negating what you tried to achieve by having higher rail to drive the mosfets from.
Morton, please run your simulator on the last schematic, I think you just lost some output power.
Morton, please run your simulator on the last schematic, I think you just lost some output power.
Last edited:
Sorry to say, but simulation is not my thing. I din't know how... But if the educated persons suggest I put the VAS inside the cap multiplier, I will do so!
Thank you! If it is better to include the VAS behind the cap multiplier, I will do so. But drivers? Ouch! More work, and I don't know how to caculate drivers... 🙂
I did an analysis of the impact of adding drivers. They do seem to reduction THD quite a bit. However, the reduction appeared to be all 2nd order. I wasn't sure that was a good thing.
Yes I see it is in the last schematic. In that case you are absolutely correct, maybe the capacitance multiplier even becomes moot. Maybe it suppresses transient response, maybe some one with the schematic already saved in his simulator can check. Brian, I trust you to do it, you probably sitting waiting for some more work.
You are probably right! It was a relic from the original HEXFET version - and I generally like cap multipliers as a practically free noise remover.
But with the regulated supply, wich in my case will include a cap multiplier anyway, it may be redundant.
On the other hand, the board is routed and not much real estate is gained by removing those parts. Lazy? Yes, whenever I can!
But with the regulated supply, wich in my case will include a cap multiplier anyway, it may be redundant.
On the other hand, the board is routed and not much real estate is gained by removing those parts. Lazy? Yes, whenever I can!
The point is valid on whether this is needed. If the IPS-VAS PSU has its only filtering and is putting out clean power, the cap-x probably isn't needed. You can probably just get away with a 100uf + 100nf bypass cap on boards.