Elvee, so if using denoiser for LM337 I need to use a PNP instead of the NPN? What about the dienoiser? Do I just swap NPN with PNP and PNP with NPN? I thought (wrongly) I need to just flip C-E on the same footprint.
RickRay remember for LM337 you also need to flip the polarity for C3/C2 as well.
RickRay remember for LM337 you also need to flip the polarity for C3/C2 as well.
Last edited:
Yea, going to need to replace the filter caps. I removed the two 100nF caps and measured the ESR at the regulator with an Atlas ESR70. The regulators are seeing between .10 and .14 ESR.
BTW, there are 8 of those filter caps in parallel, two for each channel and a set for the phono and a set for the line stage.
BTW, there are 8 of those filter caps in parallel, two for each channel and a set for the phono and a set for the line stage.
I hate to change the 220uF to 100uF, but I see why. Those 100uF are like an anomoly in the middle of the catalog, one of the only ones in the .20-.24 range. Could I just add a .1 ohm resisitor in series with the rail? That way I would need to change the caps and it wouldn't change the voltage very much.
Sure, if you can manage then that is even better so you keep the original caps. LM317 doesn't need an output cap. Hence the 0.1uF used in your application. Input caps esr doesn't matter for the stability of the regulator, they are ok if they have a low esr.
Add a 0.1ohm resistor in series with the output cap to ground.
I was wrong with the bjt flipping, you just need to swap between them for lm337.
Add a 0.1ohm resistor in series with the output cap to ground.
I was wrong with the bjt flipping, you just need to swap between them for lm337.
I think I remember that being an option as well but the recommendation is to have it in series with the output cap to ground.
That is a phono preamp, means there's low current draw. Those caps are meant as local decoupling, I don't think the value matters. They are there to filter noise not supply energy to the output stage of the phono on peak currents. Since you're using the denoiser, the overall noise should be lower so their value doesn't need to be that large. Using a single cap with a known esr is simpler. I may be wrong and if someone knows different do correct me.
Thanks Elvee! So I was wrong, I asked a mod to delete the previous posts where I got it wrong for lm337.
That is a phono preamp, means there's low current draw. Those caps are meant as local decoupling, I don't think the value matters. They are there to filter noise not supply energy to the output stage of the phono on peak currents. Since you're using the denoiser, the overall noise should be lower so their value doesn't need to be that large. Using a single cap with a known esr is simpler. I may be wrong and if someone knows different do correct me.
Thanks Elvee! So I was wrong, I asked a mod to delete the previous posts where I got it wrong for lm337.
Last edited:
Knowing the length of wire/trace between the output of the reg to the phono decoupling caps would be useful as well. You might only need to replace the 100nF caps on the output with something like that FC cap or even FR series has 47uF/63V which fits for the reg output. The wire/trace length might be all that is needed to get the esr of the phono caps in the correct region.
You are modifying the original circuit so having those caps might not benefit your application as it did originally.
You are modifying the original circuit so having those caps might not benefit your application as it did originally.
Last edited:
Adding the 0.1R resistor in series between reg output and load actually ruins the ultra low output impedance of the reg+denoiser. Hence it's better to add it between cap and ground. The output impedance is way lower than than the 0.1R resistance you add. Ideally you'd have thick traces/wire between reg output and load, and around 0.2R esr for the output cap. I say try it with a .2R esr cap at the phono stage and listen to it. You might like how it sounds. If not then you can tinker with the series resistor + larger caps. Two extra caps for reg output + two extra caps for phono decoupling should be cheap overall.
Last edited:
A 40mil trace at 30cm length for 1oz copper has a resistance of around 150mOhm using an online calculator. If the phono caps are 100mOhms should be ok. To not complicate too much just remove the 100nF caps from the output of the regs and see how it does with the denoiser circuit. You can tap the VCC/GND from the removed 100nF caps. Start with that and see how it goes. So just completely remove C35/C37 when adding the denoiser.
View attachment Scan20201018163240.pdf
There are 4 caps in parallel on each rail/regulator, L/R linestage, L/R phonostage. The regulator feeds both the phone and linestage. I didn't copy the second page of the schematic. I attached it above. The schematic is only showing one channel, so everything times 2 except power supply.
There are 4 caps in parallel on each rail/regulator, L/R linestage, L/R phonostage. The regulator feeds both the phone and linestage. I didn't copy the second page of the schematic. I attached it above. The schematic is only showing one channel, so everything times 2 except power supply.
Check IC3, is it powered by the same rails? If that chip has +- >25V and caps on both rails are 35V rated then you can use the denoiser, which swings for a bit at startup.
Best is to measure ESR at the reg pins and see what it "sees". If it's around 0.2R then it should be ok, if it's around 0.1R and lower then you might need to expect oscillations. Tracks may dampen it but you need to keep that in mind.
Best is to measure ESR at the reg pins and see what it "sees". If it's around 0.2R then it should be ok, if it's around 0.1R and lower then you might need to expect oscillations. Tracks may dampen it but you need to keep that in mind.
IC3 is an LM393 comparator, max voltage 36V. It is fed from the 23.1V rails for a total of 46.2. However, look at the schematic in post# 1476, pretty sure zener diode D12 is going to keep the positive rail at +10 volts going into the chip and the negative side will still be -23.1 for a total of 33.1V. I'm pretty sure D12 will protect the chip during the start-up swings.
I also measured the ESR at the regulators in an earlier post, they were .10 to .14 ESR. I may have issues, I'll find out.
I also measured the ESR at the regulators in an earlier post, they were .10 to .14 ESR. I may have issues, I'll find out.
