We need inrush if we use a 1000VA main transformer and 72 000 uF of the filtering caps. Powering such a PSU with no inrush results in shorting the mains and power down in the whole house. I have tested this scenario.
so where does the DC offset really happen? I mean which part of the circuit the DC gets generated the tubes?
Hi Vizai I see the super pair at the output of the VAS but what is the use of it in VFA? in CFA it will increase the slew rate substantially but VFA dont you have stability issues in the circuit? Any sonic benefits of having the super pair in the current design?
When the tube is cold it is off so there is no feedback. You can see full rail at the speaker. The most important thing is that the tube has time to fully warm before the rest of the amp is engaged. I would still not run without some type of DC protection.
Hi rhythmsandy,
The use of it is actually the same - high speed and linearity. TubSuMo front-end is very fast and low-distortion - one on my favourites.
In terms of stability - combination of the right base stoppers, local compensation and global compensation ensures unconditional stability of the circuit. It's been built many times, used with different output stages and never had stability issues.
See the measured performance with lateral MOSFETs:
TubSuMo with Laterals
as well as with non-switching NS-OPS:
TubSuMo with NS-OPS
If you're looking for a great sounding amp - the one with laterals is strongly recommended - ease of build and top performance.
Cheers,
Valery
Terry, absolutely right. Input LTP needs some seconds to settle even when the filament is already hot enough.
In any case, I would never connect my speakers to anything DC-coupled without the proper soft-start/protection system. That's how the 21-st century board started 😉
Cheers,
Valery
I have tried using the ALF16N20W and they are very good. I do wan to know more about the NS OPS. How much bias have you used to get the low distortion at 20Khz. In general does the OPS switch with conventional simple EF driver? For the bias of 60ma per transistor?
NS-OPS is a EF3 with CCS-ed class A drivers and mechanism, preventing the output transistors from switching off when inactive. It runs at 65mA per output pair (3 pairs of MT-200 Sankens in total) and shows extremely low distortion - see the measured spectrums and THD / IMD numbers in combination with Vertical front-end here:
Vertical CFA + NS-OPS measurements
There are not too many amplifiers, showing 0.0007% THD at 20KHz (measured at 50W @ 8 ohm) and 0.0005% IMD 14+15KHz in the same conditions. This is a virtually "no distortion" amplifier.
Very low distortion of the OPS allows relatively low global loop gain in the amplifier (35-37db), still showing very good distortion performance. See X4 series of the low gain ultra-linear current-drive front-ends at the link above - combination with NS-OPS forms an outstanding amplifier with exceptionally natural sound. X4 JFET version is especially cool.
It's published in the other thread:
http://www.diyaudio.com/forums/soli...e-old-ideas-1970s-ips-ops-35.html#post4554894
Post #1 in that thread contains a very good index with lots of good developments listed.
Non-commercial use only, please 😉
was trying to simulate the ckt with the NS OPS with Kypton v2 but im unable to get higher bias no idea why. Here is the .asc can you tell me whats wrong in the above schematic?
From what I see right away:
Q13, Q14 are wrong - must be MJE340/350
Q17, Q18, Q19, Q20 are also wrong - must be MJE15032/15033 or 2SC4382/A1668
R50, R61 = 22R
updated with the models you have mentioned but still no increment in idle current in emitter resistor.
I triple checked the circuit its not working. I would be interested to know where exactly the circuit is wrong. Is it working circuit?
What's not working: your simulation, or an actual amplifier that you've built? Countless people have built the amp based on Valery's circuit boards so the circuit is known to work, and work extremely well.