His goal may have been to get the HAT off the switcher. Normally a HAT takes power off the pi-board. You can easily power the HAT separately by using a cable to connect to the pi instead of using it as a HAT. The power requirements on the HAT are probably tiny (10ma or so) and he could even use the xformer and regulator he already has to do the job.
I think part of the reason I am skeptical of using a linear for the pi is it has a switcher on the board to take 5-3.3 to power the broadcom part anyway. So you are really not removing a switcher from powering the pi.
I think part of the reason I am skeptical of using a linear for the pi is it has a switcher on the board to take 5-3.3 to power the broadcom part anyway. So you are really not removing a switcher from powering the pi.
Last edited:
I will, next week I should have a different transformer and will test it again. I hope it works this time 🙂
I would like to have a LPS for the the hat, and since I can get the LPS to supply both with no really extra (or relevent) costs wI will give it a try, later on I can go back to power the pi with the official switch power supply but for now that is not the objective 🙂
I had a feeling it was the HAT you wanted on the LPS. Is the HAT pulling 5V or 3.3 off the pi though? If it is 3.3, you are still getting a switcher (the pi-board) as your power source.
Hi Mike
The hat has two very different power options:
a) it has a DC input (2.5mm jack) that takes 19V to 32V (the supplier offers an optional 24V 1.67A switching power supply). the power regulated down to 5V 5A and sends the necessary power to the pi and to the hat from that (see image)
b) it can be powered with 5V (doesn't draw more than 2A) directly from pin 2 and 3 (or 1 and another ground pin) bypassing the voltage regulator in the board but, in this case the pi must be powered separately.
I wanted to try both configurations and since in a) I would need 24V this explains why the transformer with 24V output but my first experiment was 5V to hat and 5V to the pi and this is when I discovered I messed up the current equation big time as the transformer I have chosen is not capable of delivering the current the two boards required. I tried 19V directly to the hat input jack but it didn't work either. I didn't try 24V.
Does it make sense to power the pi with the LPS, well... since I had two power supply modules I gave it a try 🙂 I have no issues of changing my mind later on, but for now I would like to power both units with the same "clean" power.
This is my first try on a linear power supply so you see how little I know about the topic, even if I spent the last 2 or 3 months reading all I could about the topic, but most people, even when they have doubts and ask in the forum, they are several steps ahead of me so I may have better understood the part of the process but not that well some basic concept on current ...
let me know if I could explain my self
Regards, Nuno
So I took a look at the website that sells your box. Nice box. They don't seem to have a problem with using a switcher to bring the 24V down to 5 to power the box and the pi. But they are using linear low noise regulators for 5->3.3 on their board. The board itself appears to be a well designed low noise design. Nothing wrong with trying a linear for the 5v, I just don't expect you'll see much difference. Especially since it is all digital anyway. At first I thought the board might have a D/A or A/D, but it does not.
I am still unsure why they chose 24VDC as their input. They could have gone 5V and skipped the 24->5 DC/DC converter. They mention the 24V can be used on an expansion header, but could not figure out what that expansion header would go to. Nothing on the board needs 24V from what I can see and the block diagram only shows it power the DC/DC converter.
I am still unsure why they chose 24VDC as their input. They could have gone 5V and skipped the 24->5 DC/DC converter. They mention the 24V can be used on an expansion header, but could not figure out what that expansion header would go to. Nothing on the board needs 24V from what I can see and the block diagram only shows it power the DC/DC converter.
Hi Mike
Thx for taking the time to check out the pi2aes. Indeed it is a nice board. I too never really got the 24V input by design but I'm not able to dig deeper as I lack the knowledge on the topic.
Initially I wanted to put all in a nicer case, Pi+Pi2aes and I thought... why not a LPS to power this ? 🙂 So here we are.
Maybe in the end I will see little to no big difference but at least I have this little project and I learn something.
I should get another transformer this week or next week (7V+7V 50VA) and will give it a try and let you and Jean-Paul know how it went.
PS: my next project of this kind will be an active crossover with two little boards from XKITZ, that will be fun as well, but probably this project is further down the road, maybe at the same time I try to build a subwoofer.
If I remember correctly pi2aes only uses 100mA @5V (and it is a very steady load). I've tried your experiment many times switching back and forth between clean 30V and clean 5V and I consistently seem to prefer 30V, though I have no explanation on why it is better. Dynamics seems a tiny bit better, but I bet you will need a very high quality chain to hear any differences and they will still not be obvious at all. I would have liked to hear it with an ultracapacitor @5V (i.e. ucconditioner), but I've been too cheap to buy one.
🙂 experimenting is part of the fun, even when the differences are not that evident. I will try this one day too. Do you think it would work with a LPS with 24V and 1,25A ? what do you think?
It will definetly work, and it can provide for both units if you want (with the jumper in place at the side). Looking forward to your impressions 🙂
Thx to all. With the 7V+7V 50VA it worked, powering up the two units separately. Sound quality is great, haven't tried the 24V configuration yet, for now I'm happy with this for now.
I learned a lot with your help.
Can't wait for my next diy project
I learned a lot with your help.
Can't wait for my next diy project
Hi
A quick update,
I never got to make this work properly, even with the 7V+7V 50VA feeding the pi4 and the hat separately with 5V. It was working for a few days and stopping... I was getting always the message of low power even if the output voltage was within specs.
I went back to the 24V initial plan and it is working fine, I can't say for sure if it sounds better but... first impression is very positive. Maybe the low power or the instability of the previous solution was hurting the sound quality, but it looks like it sounds more dynamic 🙂 so I'm happy I tried and thx again for all support and suggestions.
Last edited: