12 minutes with lid on and 17 minutes with lid off. The heatsink was to hot to touch. The heatsink is tiny and it needs to dissipate about 20w, I don’t have much faith in the fan. I will try the fan tomorrow and if that does not work I will get the 1000w smps from aliexpress.
Why not try the Meanwell? Cheap and good enough for Papa. And why not linear? A hassle, but always a fun experience.
Linear supplies do take building, but they're not too bad, though it does mean working with line-level voltages, which can of course be dangerous. There's the VRDN supply that's discussed in its own thread, and AMB (amb.org) has some nice linear supplies, too.
If you should ever want a good starter project that incorporates such a supply, the Whammy headphone amp could be a good choice. Everything is on one board, including the transformer, and the supply itself is pretty simple. Not to mention that it's an absolutely killer headphone amp.
If you should ever want a good starter project that incorporates such a supply, the Whammy headphone amp could be a good choice. Everything is on one board, including the transformer, and the supply itself is pretty simple. Not to mention that it's an absolutely killer headphone amp.
I wanted to use an smps for a few reasons. Less noise at audible frequencies, simplicity, cost, weight and just because. I know smps is not perfect, producing its own noise and rf interference, catching fire, shutting down when overheating and it is not popular with lots of DIYers. My smps is external so hopefully its interference will be minimal.
I will get REW out and see what is going on with my amp when I am done with my smps. My amp so far is producing no noise that I can hear and is crystal clear so I am happy with my take on Papa’s amazing circuit.
I will get REW out and see what is going on with my amp when I am done with my smps. My amp so far is producing no noise that I can hear and is crystal clear so I am happy with my take on Papa’s amazing circuit.
Last edited:
These are very quiet but not powerful, with a maximum of rated 1.5 A per rail you would need 1 per channel.
When the designer writes 3A he actually means the total current of both the positive and negative rails...basically 1.5A max.
I spent some time last night looking at Meanwell psu’s and then this morning I had an idea. I will get a second 500w as they only get warm running one channel and it is also the cheapest option for me.
My second smps will be hear in a couple of weeks🙁. That will be a long two weeks. I got a fan, it keeps the smps cool but is too noisy (it is a silent one apparently ) and I stuck a potentiometer in there so it is slower. I will play around with it and see if I can improve it. I will put the two smps’s together along with a 12v trigger and no fans ( or a giant ultra quiet slow fan if required).
I got a 60mm fan. It might be quieter if I move it to another position in the enclosure. I had my AJ on for over 3 hours this evening, much better than 15 minutes. I will try to sort out a nice face plate (probably oak), attach some feet, spray the aluminium panels and drill a hole for another Powercon socket before the 2nd smps arrives. When it is a dual mono I will get some measurements for everyone with REW and a Focusrite.
Hi All. I made a boo-boo and need some help. While desoldering the jumper placed at R30, I accidentally removed one of the solder pads. I would like to repair this by soldering a wire onto the surface of the board that makes the required connection that has been removed. I looked at the schematic, and tried to figure it out with a DMM set to continuity, but I am stumped as how to do this. I really don’t want to take the board off the sinks, as this build was a bit of a wiring nightmare, and this would add a lot of work. I have attached an annotated photo that shows the place on the board where the solder pad came off. Any guidance would be much, much appreciated.
Based on a photo I found online of the back of the DIY Audio PCB, it appears that the connection in the photo should be made. Is this correct?
Thank you very much for your speedy reply. R30 was a jumper, and the right side (in the photo) is torn off, while the left side is good. Can I put a jumper from the left side to C of Q2 as in the photo in my second post?
You can, if the PCB via/eyelet/hole marked with the red cross, is good. It seems it is, so go for it.
I have finally managed to find some spare time and decided to give different bias settings' a go. I will use the voltage drop across one of those 0.47 ohm / 5W resistors as a reference. The 3 values of interest were 0.310mV, 400mV and 0.450mV. The first one should give negative H2, the middle value should give... I actually don't know... and the third one should give a positive H2 harmonic spectre.
310mV:
The sound was placed away from the listening position and had very good extension top-to-bottom and left-to-right. Very nice, pleasing, natural. With my 4 ohm speakers, I lost a lot of bass drive/control. Other than this, the sound was very nice. I expect this bias setting to be probably the best option for easy-to-drive speakers. In my case, the price to pay with regard to loss of bass control and extension was too severe.
450mV:
The sound was congested... the high and low frequencies collapsed horizontally. The sound was more in-my-face. However, the bass grip, extension and control was great. I am not 100% convinced that the low frequencies were portrayed naturally, though. It was like the bass extension/control was really good, but the bass frequencies sounded too tight, too stressed. The whole soundscape, in fact, was sounding stressed, but with a lot of drive and control. Again this was with my 4-ohm Dynaudios.
400mV:
This setting gave beautiful extension top-to-bottom and left-to-right. The soundscape was just right, placed where it should be... not in my face and not too far behind the speakers. The details are allowed to breathe, instruments had their own placement. The sound was simply correct... something I really enjoyed. The bass was well controlled and natural sounding. The only thing that I am kind of missing... is that bass control. However, I definitely appreciate the bass that seems to flow freely, with a 400mV bias setting. This is my new bias setting (it used to be 450mV before)
Edited by moderation as requested.
Last edited by a moderator:
It’s alive! It has been biased at .4 Volts for an hour. One of the LEDs is dead and needs replacing. The washers securing the mosfets are at 27-30 degrees Celsius, and the heat sinks, where they are hottest close to the mosfets, are at 46-48 degrees Celsius. The measurements are made with the heatsinks flat. They should be cooler when in their optimum upright position. Does this mean that I can bias hotter?
The amp.
The temperature of the washer securing the mosfets.
The temperature of the hottest part of the heat sink after an hour.
The amp.
The temperature of the washer securing the mosfets.
You probably could though I would defer that decision until after the case is put together. Your heatsinks in the vertical position should be more efficient, but the interior of the chassis will likely be hotter because of the more restricted air flow.