| wish |
What is usual square wave of Zen4 amplifier?
I looked at 1KHz square wave at output and notice 10 peaks on top of square. There was also noticable fall before first peak.
I increased fallback capacitor c12 value approx. 2 times and number of notable peaks decreased to 4.
I can't deside if it really helps to improve sound.
What are usual and desired square wave forms?
Should I increase c12 value more to see no peaks and falls? |
|
|
| Prune |
| quote: | | What are usual and desired square wave forms? |
Squarish. |
|
|
| Blues |
| Hi wish! Are you looking at an actual zv4 circuit or in simulation? I saw the peaks on the leading corner of a square wave using simulation with c12 at 5pF. Increasing c12 to about 11pF rounded it off up to 10khz. In Nelson's zv4 article he described how he watched the square wave as he trimmed a couple of twisted wire he used for c12...he most probably got 5pF to work well with the stock zv4 circuit board. Other values might work well with other DIY boards. Tell me more on what you find out...I don't have an oscilloscope. Looking at the amplified square wave tells us how close it resembles what you feed it...it certainly affects the sound...how much you like it subjectively? Look at amplified sq. w of tube circuits!:bigeyes: |
|
|
| wish |
Hi!
I'm looking with oscilloscope at actual built circuit. Q4 is replaced with IRF9610, R2 and R3 lowered to 10K and 47K, C12 initially increased to 22pF. PCB is reworked. Now C12 is increased twice, to 40pF.
I'm confused with the fact of multiple small ringing peaks even after doubling C12 and suspect some flaw in my design. Any advice?
| quote: | | Look at amplified sq. w of tube circuits! |
I've found some wave samples:
http://www.icl.co.jp/audio/english/2a3/2a3amp.htm
http://www.hometheaterhifi.com/volu...amp-1-2001.html
I see that form isn't ideal, with single or multiple peaks, did you mean that?
Thanks. |
|
|
| Blues |
| my advice is to sit down with every change you make and then let your ears be the judge (subjective). then you'll come to the conclusion which values and measurements (objective) sounds best. make sure you take note and let me know:)...I don't have the luxury of test equipment and time right now. I lot of tube designs have far from ideal bench test results but sound good to many. |
|
|
| GRollins |
Those little ripples indicate that the amplifier isn't quite as tame as it should be. When in doubt, roll things off a bit. It's better to have a slightly rounded leading edge than one that looks like a pond where a fish has jumped.
Grey |
|
|
| Nelson Pass |
I have a stock Z4 here, and I'll fire it up and look, and
report back.
Later:
No overshoot. Source impedance (and that resistance in series
with the input that sets the gain) have an effect here. Are you
using less than stock value for input resistor? |
|
|
| jarek |
| quote: | | I looked at 1KHz square wave at output and notice 10 peaks on top of square. There was also noticable fall before first peak. |
Silly question. Have you tested signal itself? I mean your source of the square walve. |
|
|
| Blues |
I noticed in simulation that lower values of R2 and R3 tend to require a higher value of C12 to tame the overshoots.
Nelson, in your experience does a lower value of feedback resistor and higher cap across it robs the output of some microdetails (treble extension)? And does the ckt board layout have an effect on c12 value? Thanks! |
|
|
| Nelson Pass |
My wanders over the value terrain does not seem to have
covered that plot of ground, but probably it is the case. I chose
the input impedance to satisfy the whining of preamp owners
who think that only 47 Kohms is compatible with their tubes. |
|
|
| wish |
Thanks for All. Thanks for Your great support!
| quote: | Originally posted by Blues
my advice is to sit down with every change you make and then let your ears be the judge (subjective) |
Well, I think I need to clean up my design first.
Now I know for sure that I should get very close to square response to square wave input with my Zen.
I made some experiments and found that twisting "+" and "-" wires from PS large capacistor to PCB and increasing C10 (capacitor across power rails before regulator) from 0.47 to 4,7uF lowers number of noticable ringing peaks from 10 to 4. Much the same result that I got increasing C12 from 22 to 40pF.
I doesn't understand to the end what's going on, but something is wrong. I definetely should go further and resolve this.
| quote: | Originally posted by jarek
Have you tested signal itself? |
Yes. In fact it's calibrating output of my oscilloscope. Square wave, 1KHz, 1V. I don't know for sure, does it drive 10K load of my Zen v4 without problems, but square wave at Zen input definitely doesn't have any overshoot or ringing.
| quote: | Originally posted by Nelson Pass
Source impedance (and that resistance in series
with the input that sets the gain) have an effect here. Are you
using less than stock value for input resistor? [/B] |
Yes.
I've read in Zen v4 article: "This can be improved by lowering the values of R2 and R3. If you can accept a 10K input impedance, you could consider using these alternate P channel Mosfets".
Since I use IRF9610, I set R2 (input resistor)=10K, R3 (fallback resistor)=47K, and C12 (fallback capacitor)=22pF. Choosing C12 I thought I should get R3*C12 value not changed.
Mr Pass, You say output impendance of signal source plays role here. I don't know impedance for calibrating output of my oscilloscope, hope it's low enough. I should check this also. |
|
|
| Nelson Pass |
If you want to round off the square wave, set the input and
feedback resistors to a higher value. |
|
|
| Blues |
| Wish, try R2=20k; R3=93k and c12~15pF or so...works well in simulation. 40pF for c12 seems high and might be a factor at high frequencies especially if you have a supertweeter in your loudspeakers. |
|
|
| wish |
Thanks for the response!
I've just added 220uF capacitor in parallel with C10 (across power rails on PCB) and problem has disappeared. I see no ringing at all now with my default values (R2=10K, R3=47K, C12=22pF) :cool: .
I believe the real sources of my problem may be:
- poor power supply capacitor quality (I used cheap one)
- poor connection between power supply capacitor and connected wires (the capacitor isn't of screw terminals type)
- improper C10 type (I used cheap metallized polyester film type) |
|
|
|
