I recently built the 12AU7/IRF510 from here -
NP-100v12: DIY 12AU7 (ECC82) Tube / IRF510 MOSFET Headphone Amplifier
However, when put on a power supply I built using the basic LM317 with 1N4007 diodes and only 1000uF of capacitor after the rectifier, I found the buzz to be horrible. When I placed the circuit on a 12V battery, it was super clean. As such, there seems to be many people simply using a wall wart and filtering, but I would really like to nail down a good, clean PSU using the LM317's I have laying around. Based on some recent suggestions and with some good grounding practice, I think this circuit should have enough PSRR. Is that probably true?
Thanks.
NP-100v12: DIY 12AU7 (ECC82) Tube / IRF510 MOSFET Headphone Amplifier
However, when put on a power supply I built using the basic LM317 with 1N4007 diodes and only 1000uF of capacitor after the rectifier, I found the buzz to be horrible. When I placed the circuit on a 12V battery, it was super clean. As such, there seems to be many people simply using a wall wart and filtering, but I would really like to nail down a good, clean PSU using the LM317's I have laying around. Based on some recent suggestions and with some good grounding practice, I think this circuit should have enough PSRR. Is that probably true?
Thanks.
Hi milsman2. Great to see that this headphone amp still gets built. I built my first one in 2015 from the same diyaudioprojects pages. It was my very first diy audio project, and wow – what a delightful little amp. I built it point-to-point and was so impressed I built a second one for fun. I still use them often when I long to hear that warm tube sound.
Getting back to your problem (and I am probably not qualified to hand out advice here), I can just relate my experience. I used a very similar LM317 based regulator schematic I got from Pete Millett’s audio pages. If I remember correctly I had to change the one resistor RR2 from 2K to 600R in order to dial into the required 12,6V for the amp. I feed the regulator with a 19V open frame Meanwell smps. This setup is dead quiet when I use the amp with my desktop PC.
I subsequently built (2 years later) another headamp - xrk971’s famous Pocket Class A headamp (supposed to fit in a Altoids tin), using the exact same LM317 regulator. Another forum member, BYRTT, spotted this and suggested an improvement. You can see my point-to-point implementation and BYRTT’s suggestion here: xrk971 Pocket Class A Headamp GB
However, having two of these NP-100v12 12AU7 amps I also ran into horrible buzz problems and my conclusion is that it is not really the PSU or LM317 regulator, but rather a ground loop issue. Both my amps have the same LM317 regulator and both are powered with smps supplies - but the one supply (Meanwell open frame) is earthed and the other one (laptop smps brick) not. So, here are my findings – it depends on the combination of your source and PSU grounding (or earthing) scheme. When I use amp No1 with a smps psu that is connected to earth (third prong on the power outlet) running from a desktop computer that is also earthed – no buzz. When I run the same amp from a laptop computer with a smps brick (no earth connection) – horrible buzz.
When I run amp No2 from a laptop computer – no buzz, because both the amp and the laptop is powered with smps bricks that are not earthed at all.
Hope this makes sense and can help you. But please feel free to ask more questions or PM me.
Getting back to your problem (and I am probably not qualified to hand out advice here), I can just relate my experience. I used a very similar LM317 based regulator schematic I got from Pete Millett’s audio pages. If I remember correctly I had to change the one resistor RR2 from 2K to 600R in order to dial into the required 12,6V for the amp. I feed the regulator with a 19V open frame Meanwell smps. This setup is dead quiet when I use the amp with my desktop PC.
I subsequently built (2 years later) another headamp - xrk971’s famous Pocket Class A headamp (supposed to fit in a Altoids tin), using the exact same LM317 regulator. Another forum member, BYRTT, spotted this and suggested an improvement. You can see my point-to-point implementation and BYRTT’s suggestion here: xrk971 Pocket Class A Headamp GB
However, having two of these NP-100v12 12AU7 amps I also ran into horrible buzz problems and my conclusion is that it is not really the PSU or LM317 regulator, but rather a ground loop issue. Both my amps have the same LM317 regulator and both are powered with smps supplies - but the one supply (Meanwell open frame) is earthed and the other one (laptop smps brick) not. So, here are my findings – it depends on the combination of your source and PSU grounding (or earthing) scheme. When I use amp No1 with a smps psu that is connected to earth (third prong on the power outlet) running from a desktop computer that is also earthed – no buzz. When I run the same amp from a laptop computer with a smps brick (no earth connection) – horrible buzz.
When I run amp No2 from a laptop computer – no buzz, because both the amp and the laptop is powered with smps bricks that are not earthed at all.
Hope this makes sense and can help you. But please feel free to ask more questions or PM me.
Attachments
You could take your three paralleled caps before the regulator and split them up to make a CRCRC filter.
How would that help? Not being sarcastic just still learning. And what values would you suggest?
That does make sense but my PSU was not earthed? I just had one bus line for the PSU and amplifier using the breadboard and the hum was pretty pronounced. Do you have any design that worked well as a standalone power supply not using an SMPS? Or rather a picture of a buzz free setup not using an existing PSU?
Last edited:
A regulator will offer a certain amount of hum rejection. If that's not enough, you'll still have power supply noise. A CRCRC will knock down the power supply noise before it enters the regulator.
How much DC voltage do you have at the "IN" terminal of the regulator, and what's the dropout voltage and DC current drawn through the regulator?
How much DC voltage do you have at the "IN" terminal of the regulator, and what's the dropout voltage and DC current drawn through the regulator?
That would be interesting thanks, I would also like to know. For this application – 19 to 24V DC input and 12,6V output after regulator. Current is probably 1A max I guess. Then there are also CLC filters, but I don’t really understand when CRC vs CLC filters are best used.
We are trying to solve your particular problem here. Best you supply us an in-depth description of your complete setup and even photos if you like. That way more members here can understand your setup and chime in.
I was always against low voltage tube use. But I will try this mess, and will see, but I think that this is not good for my HD800 🙂
Could you send photos of layout so we can look at your grounding around the power supply. It could be a ground loop. The input caps should return directly to the rectifier with no signal path there.
Last edited:
If you have such strong convictions and beliefs, why bother to try this? Your beliefs are also OT here.
I can certainly say that this little amp sounds fantastic – I really love it. Yes, it is low voltage and a very simple design – therefore a perfect beginner project and I am forever gratefull to the people who posted all the info and the complete build guide over on the diyaudioprojects site. Being a tube amp, the sound is also affected by the tubes used, so tube rolling is fun here too (once it runs buzz free).
Agreed on the sound quality. Like I said, when I put this on a 12V battery it sounded stellar. That is why I really want to put together a good PSU design so I can call this one all my own from mains to output. Thanks for the reassurance.
I would, but since my original design was producing buzz and I believe part of it was from a flawed design, I will be building a new PSU from scratch. Given that that is true, I want to be sure and implement the changes to layout and how it was built as well as new the design. I could really use help with that part.
I have about 18-19VDC from my readings that I remember. I already gutted the board as I know there were some issues with how I built the first one. But again, I also want to rework the design so I am not going in circles.
Do you have a CRCRC design you would recommend?
From the LM317 datasheet -
8.4.2 Operation With LowInput Voltage: The device requiresup to 3-V headroom (VI– VO) to operate in regulation.The device may dropout and OUTPUT voltage will be INPUT voltage minus dropout voltage with less headroom.
Load current - 10mA maximum
That being said, do you have a CRCRC design you would recommend?
Hi, I was just wondering if you have found a solution to run this amp silently? Would be interested to know what worked for you.
Mine worked ok with just a 12vdc regulated wallwart.
I found some valves wouldnt work at that low a voltage and I had to "persuade" them by using a little positive grid bias.
I found some valves wouldnt work at that low a voltage and I had to "persuade" them by using a little positive grid bias.
I did manage to get it up and running and it sounds quite good. If you listen very carefully with sensitive headphones you can hear a tiny bit of hiss but I also still have to ground the input cable shield and put this in case. For a point to point build on perfboard using a solderable breadboard PSU, I am very happy.
Built it a time ago, one of my first makings. I remember that at the first I had a slight buzz me too, but then I fed it with a simple SMPS (15 V) and then used a simple LM317 regulator and worked fine.
You should check solder joints at PSU or maybe the ground connections, I could be also a ground loop.
You should check solder joints at PSU or maybe the ground connections, I could be also a ground loop.
