Happy new year to all Fellas here.
Only one little question about CA-9.
I was searching from post 12.04.2020 when Mr. Miles has showen a nice
build and layout, 100 sides now.
Have i missed some gerbers from layout?
Is it possible that to share here, otherwise i will work on own, close to this, layout.
Regards.
Only one little question about CA-9.
I was searching from post 12.04.2020 when Mr. Miles has showen a nice
build and layout, 100 sides now.
Have i missed some gerbers from layout?
Is it possible that to share here, otherwise i will work on own, close to this, layout.
Regards.
Hi everyone
I made 4 Active servo PSU with LM317 and all of them works.
But, I have fluctuation in output voltage on second decimal.
I got 15.09 and 15.10V. It changes in time every few seconds.
That means at least 50mV gap.
I built this, because I think that is more stable then LM317. I was thought that servo is for better ripple rejection.
After some digging on net I found, that this servo is for rejection of HF noise.
NE5532 here act as voltage comparator. Plus is on ground, and minus is also on 0 volts because it is linked via 10uF to output voltage.
If on output we got some AC signal it pass over 10uF and 1K resistor and the difference voltage is amplified via IC and pass into adj of LM317.
Here I am lot of questions.
If this is servo, why we amplify the difference signal ?
With 33K/1K we have amplification of 33 times. If we pass this to adj pin we make another difference 32 times greater.
I think that there must be combination of 33K/33k or 10K/10K for unity gain. Diferrence signal then be oposite sign, same intensity and theoretically on output of supply we got original voltage without noise.
Second question is why 33p and why 10uF.
33K/33p is filter which don't pass frequencies above 50Khz
10K/33p is filter which don't pass frequencies above 100Khz
33K/3.3p is filter which don't pass frequencies above 110Khz
10K/10uF pass from 10Hz
1K/1uF pass from near 1Khz
1K/10uF pass form near 100Hz
10K/1uf pass from near 100hz
If we wish eliminate noise from line, or all what come from load, why only HF ?
I think it must be all frequencies.
With 10K/33p in feedback and 10K/10uF on minus input we have unity gain and passing frequencies from 10Hz to 100Khz via opamp to adj pin of LM317.
I tried different combinations, 33K/33p with 1K/10uf, with 10K/10uF and with 33K/10uf on input, and at end 10K/33p and 10K/10uF and got always same result. Voltage 15.09 and 15.10 which is not stable.
I was measured with 3 voltmeters, and in one put new duraccell batteries.
I noticed some small HF noise in speakers, which should not be there with this circuit.
I had problems on 2 of them, and solved with Ne5532 +/-Vcc decoupling with 22R/10u as on schematic.
I put low ESR capacitors for 10uF capacitors and out capacitor . All of capacitors have 100n bypass capacitors.
I don't have oscilloscope to see is there oscillations or not.
My load is about 60mA.
In spice simulation all works fine, i put voltage generator 3V/8000Hz on load, an got ripple near 0.15mA.
Have the protect diodes or low ESR capacitors some influence ?
Have someone idea why the output voltage is not completely stable ?
I made 4 Active servo PSU with LM317 and all of them works.
But, I have fluctuation in output voltage on second decimal.
I got 15.09 and 15.10V. It changes in time every few seconds.
That means at least 50mV gap.
I built this, because I think that is more stable then LM317. I was thought that servo is for better ripple rejection.
After some digging on net I found, that this servo is for rejection of HF noise.
NE5532 here act as voltage comparator. Plus is on ground, and minus is also on 0 volts because it is linked via 10uF to output voltage.
If on output we got some AC signal it pass over 10uF and 1K resistor and the difference voltage is amplified via IC and pass into adj of LM317.
Here I am lot of questions.
If this is servo, why we amplify the difference signal ?
With 33K/1K we have amplification of 33 times. If we pass this to adj pin we make another difference 32 times greater.
I think that there must be combination of 33K/33k or 10K/10K for unity gain. Diferrence signal then be oposite sign, same intensity and theoretically on output of supply we got original voltage without noise.
Second question is why 33p and why 10uF.
33K/33p is filter which don't pass frequencies above 50Khz
10K/33p is filter which don't pass frequencies above 100Khz
33K/3.3p is filter which don't pass frequencies above 110Khz
10K/10uF pass from 10Hz
1K/1uF pass from near 1Khz
1K/10uF pass form near 100Hz
10K/1uf pass from near 100hz
If we wish eliminate noise from line, or all what come from load, why only HF ?
I think it must be all frequencies.
With 10K/33p in feedback and 10K/10uF on minus input we have unity gain and passing frequencies from 10Hz to 100Khz via opamp to adj pin of LM317.
I tried different combinations, 33K/33p with 1K/10uf, with 10K/10uF and with 33K/10uf on input, and at end 10K/33p and 10K/10uF and got always same result. Voltage 15.09 and 15.10 which is not stable.
I was measured with 3 voltmeters, and in one put new duraccell batteries.
I noticed some small HF noise in speakers, which should not be there with this circuit.
I had problems on 2 of them, and solved with Ne5532 +/-Vcc decoupling with 22R/10u as on schematic.
I put low ESR capacitors for 10uF capacitors and out capacitor . All of capacitors have 100n bypass capacitors.
I don't have oscilloscope to see is there oscillations or not.
My load is about 60mA.
In spice simulation all works fine, i put voltage generator 3V/8000Hz on load, an got ripple near 0.15mA.
Have the protect diodes or low ESR capacitors some influence ?
Have someone idea why the output voltage is not completely stable ?
Attachments
APEX VU8S
Hello,
For those who are interested, I extracted the electrical diagram of the VU8S vu-meter.
An error occurred in the APEX VU12S wiring presented by apexaudio, namely the AO LM339 inputs were reversed.
The correct connections are shown in the attached diagram.
Thanks apexaudio for the tips.
All the best.🙂
Hello,
For those who are interested, I extracted the electrical diagram of the VU8S vu-meter.
An error occurred in the APEX VU12S wiring presented by apexaudio, namely the AO LM339 inputs were reversed.
The correct connections are shown in the attached diagram.
Thanks apexaudio for the tips.
All the best.🙂
CA9 Amp Board
Hello again
i have done a bit homework and put a little board.
Hope it works.
Please take a look to correct contacted traces.
Inspired by following posted schematic, pic and layout
i came up to this one.
If it is OK i will share gerbers as next.
And mostly cloned Prasi's AA9MD board without fuses.
Thanks for sharing Prasi.
Best Regards. 🙂
Hello again
i have done a bit homework and put a little board.
Hope it works.
Please take a look to correct contacted traces.
Inspired by following posted schematic, pic and layout
i came up to this one.
If it is OK i will share gerbers as next.
And mostly cloned Prasi's AA9MD board without fuses.
Thanks for sharing Prasi.
Best Regards. 🙂
Attachments
Apex AX11 Bimo mod problem
Hi guys, I have tried to reactivate my diy activities this winter 🙂 and I have chosen Apex AX11 Bimo mod as a project.
Home etched pcbs and did one channel, it sings but I cant adjust bias, there is no voltage on 2sc5200's emiter resistor 0.22r (0V), on 2sc1943's there is 12mV voltage drop. No reaction to trimer (used 1k trimer).
Since then, I changed all actives and checked all passives and couldnt identify the problem.
Has anyone made this build succesefully by layout in attachment?
I checked pcb traces and they seem ok, same as in pdf.
Any ideas?
(Another problem is a little noise in highs, not much but like constant wind in tree's canopy 🙂, maybe it is a nusproduct of defunct bias setup?)
Thanks!
Hi guys, I have tried to reactivate my diy activities this winter 🙂 and I have chosen Apex AX11 Bimo mod as a project.
Home etched pcbs and did one channel, it sings but I cant adjust bias, there is no voltage on 2sc5200's emiter resistor 0.22r (0V), on 2sc1943's there is 12mV voltage drop. No reaction to trimer (used 1k trimer).
Since then, I changed all actives and checked all passives and couldnt identify the problem.
Has anyone made this build succesefully by layout in attachment?
I checked pcb traces and they seem ok, same as in pdf.
Any ideas?
(Another problem is a little noise in highs, not much but like constant wind in tree's canopy 🙂, maybe it is a nusproduct of defunct bias setup?)
Thanks!
Attachments
Have you or anyone else made a Sprint Lay6 file of this or posted gerbers?
Thank you,
David.
Can you give the post number so I know what you are referring to? I probably have the Sprint file.
solved
I have just solved a problem with my build , it was inverted bd139 🙁. AX11 Bimo mod now works fine. In the meantime, I built AX11 vanilla and it works like a charm. Now I have 2 versions of AX11 to compare with. 🙂
I have just solved a problem with my build , it was inverted bd139 🙁. AX11 Bimo mod now works fine. In the meantime, I built AX11 vanilla and it works like a charm. Now I have 2 versions of AX11 to compare with. 🙂
Attachments
Last edited:
This post,
https://www.diyaudio.com/forums/sol...mate-fidelity-amplifier-1121.html#post5799206
Unless you have a more current Sprint version of the Apex speaker protect.
Do you also have a Sprint version of the Apex Start/Stop?
Thank you,
David.
apex ax11
Both channels finished and started without problems. Sound is wonderfull, bass deep and pleasant.
Now, going to finish second channel of ax11 bimo mod 🙂.
Both channels finished and started without problems. Sound is wonderfull, bass deep and pleasant.
Now, going to finish second channel of ax11 bimo mod 🙂.
