I am so very excited by the prospect of this project. It will be perfect for powering the uppers on my lxMini's. If I use class D on the lows I can probably convince myself that leaving it on all the time uses about the same power as the 6 channel lm3886-based distribution amplifier (bridging 4 channels for the lows) I'm using now...
I bet this amp sounds fantastic. It's incredible what Pa can get out of such a simple circuit and very few parts.
Papa, babies need new shoes. The ACA Mini is wonderful and for 90-95% of my listening is fine, but just listened to Lindsey Buckingham's solo on "I'm So Afraid", and I wanted to turn it to 11!
Thinking of ordering a Pearl 3 chassis from Hifi2000, and want to order a F5m chassis at the same time. In the video Pap says this could run on a 2u chassis? That seems small for 25 watts, am I missing something? I have a toddler around, should I just bit the bullet on a 4u deluxe?
Yes, should you decide to build another amplifier down the road the 4U Deluxe has you covered for most builds.
Besides who builds only one amp....
Besides who builds only one amp....
I have found that 3U 300 sinks from Modushop work very well for the 10-25 WPC amps in the Pass forum. I’ve built a few different ones using those sinks and all of them have performed well, within design spec for bias current, without getting too hot (staying well under 60C, for sure). The MiniDissipante 3U 300 is my favorite chassis for class A amps. If I get nervous about the heat, I add external fans blowing into the sinks.
However, with little kids around (mine are teenagers currently), I understand and support going up in size for safety.
However, with little kids around (mine are teenagers currently), I understand and support going up in size for safety.
Very similar to the BZ3213 we used for the F5Pi.
The BZ has slightly better heatsink (40mm fins compared to 30mm) but less space inside in width.
You can see a thermal simulation here of the Mini Dissipante :
https://www.diyaudio.com/community/...tage-perfectly-integrated.406024/post-7537388
I would not be using them to dissipate much more than 60W per heatsink though.
Patrick
Hi Jim, it would be a very interesting solution (I'm thinking of a beefed-up version of Mark's AmyAlice), LRS-150-24 are selling for 20 Euro each 🙂. Any update on this? Thanks
A question for the SMPS experts: how can I check from the datasheet of a p.s. whether the V out pin is connected to safety earth (that would prevent putting two ps in series for a +/- supply)
Guido
I have used the LRS as a floating power supply in an M2. I used four and tied the + output of one the -ve output of the other to provide two independent bipolar supplies (for L & R). I then tied the signal ground back to the common rail and then used a diode bridge to lift this but provide a connection to safety ground. Rightly or wrongly, as I'm not clear on whether capacitor energy storage provides much benefit for a SMPS supply, I used 4 CapMXs to charge 20mF of capacitors per rail per side. Power supplies were in a separate Galaxy enclosure and coupled to the heatsinked chassis and Caps via Neutrik connectors & cable. Works and sounds great.
Attachments
Many switching supplies have minimum and maximum capacitive loads (though this one doesn't say beyond 47uF to meet spec). If you add some series R to make it an external filter, you should probably be able to exceed or maybe multiply that and also improve filtering.
I recently acquired a pair of Mean Well LRS-150F switching power supplies, which is adequate to power two stereo F5m channels at 24V 1.3A.
I was interested in determining the noise level from 60Hz harmonics and the SMPS switching frequencies and what steps could reduce that noise.
My measurements were performed by custom spectrum analyzer software using M-Audio Audiophile AP192 (192kHz, 24-bit, balanced I/O) sound card. Carefully utilizing the balanced I/O, the measurement system has a noise floor down around a few microvolts.
Here are the primary results from my measurements and experiments:
All of the spectra shown below are for 1kHz, 1 Watt into 8 Ohms output.
The first image shows the spectrum before attention was given to the SMPS wiring to the F5m PCBs. In absolute terms, the noise level is that bad.
By careful implementation of the star ground topology shown in the second image, the noise levels were well below 100uV rms.
The third image shows the spectrum with both channels driven.
The fourth image shows the output spectrum with the other channel input grounded and its input ground floating, thus breaking any input ground loop. Little is changed, indicating that the star grounding implementation is working well.
Finally, the fifth image shows the output spectrum with both channels driven, and the power to the channel +24V/-24V rail filtered using DIYAudio Store SMPS filters. The noise in the 40kHz to 60kHz region has been eliminated, but there are more 60Hz related harmonics than before.
Bottom line: These Mean Well supplies should present no significant noise issues, particularly when good grounding practices are used.
I was interested in determining the noise level from 60Hz harmonics and the SMPS switching frequencies and what steps could reduce that noise.
My measurements were performed by custom spectrum analyzer software using M-Audio Audiophile AP192 (192kHz, 24-bit, balanced I/O) sound card. Carefully utilizing the balanced I/O, the measurement system has a noise floor down around a few microvolts.
Here are the primary results from my measurements and experiments:
All of the spectra shown below are for 1kHz, 1 Watt into 8 Ohms output.
The first image shows the spectrum before attention was given to the SMPS wiring to the F5m PCBs. In absolute terms, the noise level is that bad.
By careful implementation of the star ground topology shown in the second image, the noise levels were well below 100uV rms.
The third image shows the spectrum with both channels driven.
The fourth image shows the output spectrum with the other channel input grounded and its input ground floating, thus breaking any input ground loop. Little is changed, indicating that the star grounding implementation is working well.
Finally, the fifth image shows the output spectrum with both channels driven, and the power to the channel +24V/-24V rail filtered using DIYAudio Store SMPS filters. The noise in the 40kHz to 60kHz region has been eliminated, but there are more 60Hz related harmonics than before.
Bottom line: These Mean Well supplies should present no significant noise issues, particularly when good grounding practices are used.
Here are images of the F5m stereo amplifier implementation.
The Mean Well LRS-150F supplies, the star ground, and the wiring.
In the upper center is a barrier connector block used for the star ground.
Immediate below is the power entry module.
A view of the left side of the chassis.
The Mean Well LRS-150F supplies, the star ground, and the wiring.
In the upper center is a barrier connector block used for the star ground.
Immediate below is the power entry module.
A view of the left side of the chassis.
Hi
On your post 78, the first picture shows a black wire joining the 2 power supplies, sort of the center tape but I don’t see this in your detailed ground schematic (post 77, 2nd pic)
I see that both supply have their ground connected to chassis gnd 👍
I often use smps for my class A, if this works I will for sure give it a try.
Thanks for sharing.
Eric
On your post 78, the first picture shows a black wire joining the 2 power supplies, sort of the center tape but I don’t see this in your detailed ground schematic (post 77, 2nd pic)
I see that both supply have their ground connected to chassis gnd 👍
I often use smps for my class A, if this works I will for sure give it a try.
Thanks for sharing.
Eric
This is a really good illustration of the importance of a proper grounding scheme.
I have been watching the store and waiting for the kit to show up. I already have a couple pairs of Meanwell chassis mount SMPS. My inclination would be to follow this scheme and add bulk storage capacitance for each channel - 15,000 uF to 30,000 uF on each rail.
I have been watching the store and waiting for the kit to show up. I already have a couple pairs of Meanwell chassis mount SMPS. My inclination would be to follow this scheme and add bulk storage capacitance for each channel - 15,000 uF to 30,000 uF on each rail.
In post 78 there are two black wires (grounds) twisted together from each SMPS connected to the star ground.