Hi all
I finished my V9 and am now testing it. It heats normally and seems to be very stable in that respect. However, I feel disappointed with the shape of output signal when I feed it with square signals. There is a big amount of 'overshoot' and it gets better only with a 100 pF feedback cap. The same if I reduce the Voltage gain of the V9.
Is this 'normal' ? I tried to tweak some resistors and it is the same. This overshoot which could be 2nd order harmonic, appears at low level as much as at high level. It gets somewhat worse around 20 khz, but I suppose it is because the wave length of this harmonic has then about the same duration as the input square.
I read that one can get as low as <1% total distortion, I am miles away ...
Your ideas ?
Thanks
I finished my V9 and am now testing it. It heats normally and seems to be very stable in that respect. However, I feel disappointed with the shape of output signal when I feed it with square signals. There is a big amount of 'overshoot' and it gets better only with a 100 pF feedback cap. The same if I reduce the Voltage gain of the V9.
Is this 'normal' ? I tried to tweak some resistors and it is the same. This overshoot which could be 2nd order harmonic, appears at low level as much as at high level. It gets somewhat worse around 20 khz, but I suppose it is because the wave length of this harmonic has then about the same duration as the input square.
I read that one can get as low as <1% total distortion, I am miles away ...
Your ideas ?
Thanks
What is causing the overshoot and how to fix it may be two different functions? Fixing it with some other method than correcting the cause may complicate things. Gain would be one thing to look at but, since that had no effect? Can you post the actuall circuit schematic you built? with actual values?
What is the output Z of the signal source?
What is the output Z of the signal source?
Well, it's kind of solved.
There where a number of small issues:
- indeed the Gate resistor of Q4 was 220, too high, brought it gradually down to 47. Improved the signal a lot. Howzit possible ...?
- I tried IRFP 044, IRF 140, and an ALFET. The best signal (at 20 khz full swing) was with the Alfet but it had the lowest bandwith (about 80 KHZ). The 044 had a good bandwith but the worse signal, the 140 is the best compromise. the bandwith is about 150 khz estimated. After 20 khz the square is slightly out of shape but it does not show too much on triangles.
- max output is 20 Vpeak to pk.
- now the strange things: the two channels running, exactly the same parts and PCBs and PSU. One needs 18pf feedback the other none at all ! how is that possible ? I tried among a batch of 20 LU1014 and did not see much difference. the famous R3 supposedly between 1.02 and 1.05 ohms : no apparent difference between 0.9 and 1.2 ohms...? my Vds is at 2.95 Volts.
another one: at 20 khz, the squares are square enough on the top but the bottom is somewhat not horizontal, something integrating ? what could it be ?
Any explanations about these ?
I should be connecting the bugger soon, after checking wether it needs a delayed output switch.
Regards
There where a number of small issues:
- indeed the Gate resistor of Q4 was 220, too high, brought it gradually down to 47. Improved the signal a lot. Howzit possible ...?
- I tried IRFP 044, IRF 140, and an ALFET. The best signal (at 20 khz full swing) was with the Alfet but it had the lowest bandwith (about 80 KHZ). The 044 had a good bandwith but the worse signal, the 140 is the best compromise. the bandwith is about 150 khz estimated. After 20 khz the square is slightly out of shape but it does not show too much on triangles.
- max output is 20 Vpeak to pk.
- now the strange things: the two channels running, exactly the same parts and PCBs and PSU. One needs 18pf feedback the other none at all ! how is that possible ? I tried among a batch of 20 LU1014 and did not see much difference. the famous R3 supposedly between 1.02 and 1.05 ohms : no apparent difference between 0.9 and 1.2 ohms...? my Vds is at 2.95 Volts.
another one: at 20 khz, the squares are square enough on the top but the bottom is somewhat not horizontal, something integrating ? what could it be ?
Any explanations about these ?
I should be connecting the bugger soon, after checking wether it needs a delayed output switch.
Regards
This is certainly a wise counsel.
http://www.diyaudio.com/forums/pass-labs/164003-so-my-quad-zv9-2.html#post2138229
That's what I do.
Who cares for the small mismatch, if that you love to listen to this?
http://www.diyaudio.com/forums/pass-labs/164003-so-my-quad-zv9-2.html#post2138229
That's what I do.
Who cares for the small mismatch, if that you love to listen to this?
Hi everyone
Gyuri, my output signal looks exactly like yours. Therefore I will consider it is nominal and listen. It will hve to compare with a triode set which I published with Mr Hiraga some years ago. Original and very good indeed in my opinion, and in his (I'm still so proud).
Next step: replace the BOZ by a triode Vamp, gain needed is only 3, that is not so easy with triodes. Or try the F4 version... ? Read somewhere that the difference is the F4/5 bring in some 3rd rank harmonics and that I really don't want... but who knows ...? maybe worth it.
Gyuri, my output signal looks exactly like yours. Therefore I will consider it is nominal and listen. It will hve to compare with a triode set which I published with Mr Hiraga some years ago. Original and very good indeed in my opinion, and in his (I'm still so proud).
Next step: replace the BOZ by a triode Vamp, gain needed is only 3, that is not so easy with triodes. Or try the F4 version... ? Read somewhere that the difference is the F4/5 bring in some 3rd rank harmonics and that I really don't want... but who knows ...? maybe worth it.
I'm preparing some photos, just warning you: maybe you'll be horrified because it really does not look like the usual big and heavy. It all gets into a 2U rack, because I used everywhere some fanned dissipator, so much smaller and efficient. Their price is ridiculously low, come from the computer world.
I ve read this night that one could parallell a diode on the Gate resistor to allow a faster discharging of Gate current. Has someone tried this ?
I ve read this night that one could parallell a diode on the Gate resistor to allow a faster discharging of Gate current. Has someone tried this ?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.