Hi Trileru,
You are rigth, jp6 is so close of the Vout rail than the solder melted and jumped it.
I come back later after remelting it.
What is the purpose please of the two little vias between C4 on the pcb and close to the bigger ones that are for the output plug ? The two little vias are shorted (continuity) , pcb populated or not.
You are rigth, jp6 is so close of the Vout rail than the solder melted and jumped it.
I come back later after remelting it.
What is the purpose please of the two little vias between C4 on the pcb and close to the bigger ones that are for the output plug ? The two little vias are shorted (continuity) , pcb populated or not.
Thisnis the single LM317 from this post https://www.diyaudio.com/community/...grade-any-317-based-v-reg.331491/post-6942943
Attachments
There is no link between the LED and trimpot. The LED is from the CCS circuit. R3 trimpot is for setting Vout.
You have built an earlier version without the Nonoiser option. But should work fine anyway.
The latest version has better layout for JP6 where it shorted on yours.
Those vias are actually holes so the board can be DIYed, you can solder a wire between those holes. The gerber files have a trace between them, on the top layer. For DIY you make only bottom layer so you need to connect between those holes with a wire.
You have built an earlier version without the Nonoiser option. But should work fine anyway.
The latest version has better layout for JP6 where it shorted on yours.
Those vias are actually holes so the board can be DIYed, you can solder a wire between those holes. The gerber files have a trace between them, on the top layer. For DIY you make only bottom layer so you need to connect between those holes with a wire.
Attachments
Ah thanks it is easier to follow with the rigth shematic.
So the version from the post I linked has the ccs nd must called DieNoizer? Which BC3xx is the the ccs please, the 337-40 ?
R8 is either 2 or 3k. And R4 =100R is ok in the 20V ?
What should be done with R7 please : J3 not jumped and R7 alwas equal to 22R ?
Sorry for my confusion as I was reffering to the sim than came later
So the version from the post I linked has the ccs nd must called DieNoizer? Which BC3xx is the the ccs please, the 337-40 ?
R8 is either 2 or 3k. And R4 =100R is ok in the 20V ?
What should be done with R7 please : J3 not jumped and R7 alwas equal to 22R ?
Sorry for my confusion as I was reffering to the sim than came later
Ok so based on your 20Vout need and the pcb schematic from my last post you should populate it like this:
C1 - 100uF/35V
R1/R2 combo - 120R/1.8K, R2 value depends on R1 value
R3 pot you install if you don't install R2.
C2 - if you use the Panasonic FC cap then short JP6
D4 - 5mm normal red LED
R4 - 100R
R8 - anything between 2K-3K
JP1 - open
R7 - I'd put 10R max or unpopulated and JP3 closed
for highest gain Q1 populated and JP2 open with R5 about 330R-360R.
If you're sensing from output connector then JP4/JP5 closed. The rest from the schematic, I didn't mention the obvious parts.
C1 - 100uF/35V
R1/R2 combo - 120R/1.8K, R2 value depends on R1 value
R3 pot you install if you don't install R2.
C2 - if you use the Panasonic FC cap then short JP6
D4 - 5mm normal red LED
R4 - 100R
R8 - anything between 2K-3K
JP1 - open
R7 - I'd put 10R max or unpopulated and JP3 closed
for highest gain Q1 populated and JP2 open with R5 about 330R-360R.
If you're sensing from output connector then JP4/JP5 closed. The rest from the schematic, I didn't mention the obvious parts.
Last edited:
Q1 addition to increase the gain is Diego's contribution hence the circuit version is called Dienoiser.
The CCS uses a BC327 (Q3) as it can be seen in the schematic. Gain version for Q3 doesn't matter in that position, either gain version you have is ok there.
... and the jumper post rigth above.... I follow on a smartphone so not easy.
yeah all versions have that overshoot right at startup, settling time differs a bit depending on the C3/R2 value combo.
I started doing some measurements for the 5Vout discrete version and I see the same thing, around 5.5V peak at startup then it settles at 5V. Seems typical to this denoising circuit type.
Hi Trileru,
I am confused with my low skills.
With 180R for R1 and 2k7R for R2 and 2k for one board and 10k trims for the other it stays at 13.7 V and the red diodes are not ligthning. The led is netween V output and the bc327/40 gate. Maybe the diodes were kaput as they were not brandnew from my refurbished parts.
R7 is 2k3.
The filtering cap are 330 uF
Traffo is 18V with a 1200/50 smoothing cap.
I am confused with my low skills.
With 180R for R1 and 2k7R for R2 and 2k for one board and 10k trims for the other it stays at 13.7 V and the red diodes are not ligthning. The led is netween V output and the bc327/40 gate. Maybe the diodes were kaput as they were not brandnew from my refurbished parts.
R7 is 2k3.
The filtering cap are 330 uF
Traffo is 18V with a 1200/50 smoothing cap.
Attachments
You either install R2 either R3. You could install both, but you need to take that into account. For 20Vout R1=180R and R2=2.7k. You can parallel R2 with the pot as long as you can reach the desired value.
Also what is Vin measured when you turn the supply on? If the LED doesn't light up with voltage on Vout then maybe rotate it? You might have installed it the other way around.
Also what is Vin measured when you turn the supply on? If the LED doesn't light up with voltage on Vout then maybe rotate it? You might have installed it the other way around.
Last edited:
Also close both JP4/JP5 before rotating the LED. JP4 is GND for denoising circuit and LED would be off without a path to ground.
If you want to make it simpler remove C3 to disconnect the denoising circuit, then make sure the regulator is working normally at the desired Vout. Then connect C3 and see if everything is stable.
If you want to make it simpler remove C3 to disconnect the denoising circuit, then make sure the regulator is working normally at the desired Vout. Then connect C3 and see if everything is stable.
Hi, J4 populated and red ligth is shinning....hurra!
Stupid me, twice: the 18V R-core is not 1,41 factor after smoothing but a 22V.... so with the drop of the LM317...which needs around 3V from memory plus the rectification diodes, my 20V can not be reach for the moment. But I will measure it with 18 to 19 Vout with the trims )
Stupid me, twice: the 18V R-core is not 1,41 factor after smoothing but a 22V.... so with the drop of the LM317...which needs around 3V from memory plus the rectification diodes, my 20V can not be reach for the moment. But I will measure it with 18 to 19 Vout with the trims )
In your (20Vout) case simulation shows that you need at least 1.8Vdrop on the regulator for performance to not be affected. I'd say make sure you have at least 22VDC going into the regulator for 20Vout. That's 22VDC without the ripple.
If you like tinkering you can put an oscilloscope on the input and load the supply at max load it's ever going to see, and check the DC and AC components, and set Vout of the regulator 2V lower than input DC. That should work out fine.
If you like tinkering you can put an oscilloscope on the input and load the supply at max load it's ever going to see, and check the DC and AC components, and set Vout of the regulator 2V lower than input DC. That should work out fine.