As I don't have good enough power supply it was just 1 V/div, 8 ohm load and input filter and input cap as in schematic.
BR, Damir
Dear Damir,
either your input square wave isn't good/sharp enough or your rfi filter is a bit strong. Your 20kHz square wave figure looks like my 200kHz small signal figure (2V/div) of my 2stageEF with 8R load, output zobels and input rfi capacitor.
Can you check this?
BR, Toni
Attachments
Toni, you are right, my function generator is very poor one. I have built it many years ago with one chip(don't remember now the name) and output from that generator is just the same as the output from the amp. That why I don't show square wave test to often as it does not show much of the tested amp. I just wanted to show that there is no overshoot. I think that is time to invest some in that direction, not sure how good are those quite cheap function generator on ebay?
BR Damir
Dear Damir,
you mean:
Code:
if (output == input * 27db) {
print "very good amplifier\n";
}
BTW: today you will get really good programmable function generators e.g.:
- ...
- Rigol DG1022 ~ EUR 380
- ...
- Hameg HMF 2525 ~ EUR 1430
BR, Toni
Dear Damir,
you mean:
Does it display "very good amplifier"?Code:if (output == input * 27db) { print "very good amplifier\n"; }
BTW: today you will get really good programmable function generators e.g.:
My Hameg HMF 2525 is for every day use. A GW Instek SFG-2104 is currently backup only.
- ...
- Rigol DG1022 ~ EUR 380
- ...
- Hameg HMF 2525 ~ EUR 1430
BR, Toni
Hi Toni,
Thank you for your advise, but that is a lot money for me. I am retired and this is just a hobby and can't afford something like this. I am happy if I have enough for new project. All my tools a god enough only to confirm that my amps are working but how good I can evaluate only by listening.
BR Damir
Is your scope's trace always that thick? Otherwise, there might be some high frequency content. Is it also that way at the sig gen output, without the amp connected?
You could crank the timebase up to its fastest sweep speed, and set the v/div low, to attempt to zoom in on the top or bottom of just one of the square wave traces.
This is my scope non focused trace, I can't find a time to do some maintenance on it.
Chris and Toni will receive the PCBs in next days and here are corrected schematic, BOM and the layout.
Some errors exist on the layout where some components are with wrong values or wrong denotation.
1. All output mosfets gate resistor should be of 150R (R12, 13, 27, 28, 59, 60, 61, 62)
2. 1uF cap instead C13 should be C18.
I hope that was all.
Soon there is going to be some real life measuring and listening impressions.
BR Damir
PS. OK some measurement has been done already: http://www.diyaudio.com/forums/loun...ch-preamplifier-part-ii-6613.html#post4306908
Some errors exist on the layout where some components are with wrong values or wrong denotation.
1. All output mosfets gate resistor should be of 150R (R12, 13, 27, 28, 59, 60, 61, 62)
2. 1uF cap instead C13 should be C18.
I hope that was all.
Soon there is going to be some real life measuring and listening impressions.
BR Damir
PS. OK some measurement has been done already: http://www.diyaudio.com/forums/loun...ch-preamplifier-part-ii-6613.html#post4306908
Attachments
Last edited:
The CCS used in this amp is thermally very stable, but the current is dependent of the LED, and LED voltage drop can fluctuate a quite.
Here I suggest a novel (?) CCS, also thermally very stable but uses only bipolar transistors. It has lower dynamic resistance then ordinary feedback bjt CCS, but not much.
Damir
Here I suggest a novel (?) CCS, also thermally very stable but uses only bipolar transistors. It has lower dynamic resistance then ordinary feedback bjt CCS, but not much.
Damir
Attachments
Actually no, it's the same, R8 is there to get more headroom for sensing transistor.
Hi,
I just want people to know that I have such a circuit buit and being tested.... but I have on order a larger heat-sink.
But so far some simple low power tests -- preliminary:
The 1K-5Khz at 1-10W/8 is below -100DB thd. The noise is -82db at 1W and -92 at 10W. Unweighted numbers - 30KHz BW. Thd at 10KHz is just .00145%.
More to come. But geat numbers... cant wait to listen to it.
THx-RNMarsh
I just want people to know that I have such a circuit buit and being tested.... but I have on order a larger heat-sink.
But so far some simple low power tests -- preliminary:
The 1K-5Khz at 1-10W/8 is below -100DB thd. The noise is -82db at 1W and -92 at 10W. Unweighted numbers - 30KHz BW. Thd at 10KHz is just .00145%.
More to come. But geat numbers... cant wait to listen to it.
THx-RNMarsh
Hi Richard
At what power level is the 10kHz THD taken?
I am always surprised at the high 10-20kHz distortion of CFAs since I thought part of the point of CFA is exceedingly good high-frequency response. Doesn't that mean it should have better / lower THD at those frequencies? Doesn't seem much different than VFA.
Have fun
At what power level is the 10kHz THD taken?
I am always surprised at the high 10-20kHz distortion of CFAs since I thought part of the point of CFA is exceedingly good high-frequency response. Doesn't that mean it should have better / lower THD at those frequencies? Doesn't seem much different than VFA.
Have fun
perhaps I'm missing something in understanding #771, but, Dadod, would a fluctuation or noise in the supply voltage change the CCS current? Is this circuit requiring a highly regulated, low noise supply? What would the current I(R1) be if the supply is elevated from 30V to 40V? If the current changes with voltage, perhaps it is no longer a CCS? Too many questions from me. Thanks.
Yes you are correct. I was thinking to use this "CCS" with stable supply voltage as in this 200W CFA where input stage uses +-15V stabilized by zener.
LOL, no it is not, it shouldn't, and will never be
Of course virtually all high speed high resolution IC op-amps are (so-called) CFA. I think 14bit SFDR at 100MHz is done now. So next time you answer your cell phone just remember those nasty little CFA's in the base station are making it all happen.
Last edited: