Dynalo (Rev.C) construction thread
May 9, 2009 at 1:02 PM Post #106 of 188
Quote:

Originally Posted by digger945 /img/forum/go_quote.gif
bidoux, the 7815 and 7915 have different pinouts.

kvant, it would probably cost him less for two more diodes than many times larger caps. I was only trying to help the guy out a little as it seemed he was online for quite a long time waiting for someone to answer him.



I'm not waiting for anyone to answer me, but it is kind to do so.
regular_smile .gif

You just seemed to be so out of the topic when you arrived...
You are saying large caps costs a lot but they are like 3 or 4 € each it is not that much.

PS: The PSU looks a bit like this one.
 
May 9, 2009 at 3:02 PM Post #107 of 188
Quote:

Originally Posted by bidoux /img/forum/go_quote.gif
So yours provides a ground, where the one I linked would need something like an op-amp or something else to have a ground.


Yes. There are certainly more ways to get a dual rail DC power from a single AC supply. You can rectify the AC into a single DC and then split this into two halves and create virtual ground; tangent has a detailed description of this technique on his web site. I sketched an alternative approach where I do the splitting directly from AC - the connection is also known as voltage doubler.

Quote:

I have a last question about the L7915 and L7815. There are three pins on it... ground, and the two others can be used in no particular order ?


No, the other two pins are not interchangeable.

Quote:

Originally Posted by bidoux /img/forum/go_quote.gif
PS: The PSU looks a bit like this one.


Not only do the two circuits look a bit alike, they are the same circuit. The Silicon Chip incarnation uses smaller caps as their target current draw is considerably lower than what you aim for. It is a nice picture with identification of pins etc... saves me some drawing. I also did a bit of googling and found yet another page with this power supply. These guys provide a pictorial explanation of the circuit functions. They also have PCBs for sale but I am not sure there is enough room on it for heatsinks you will need on the regulators.

Quote:

Originally Posted by digger945 /img/forum/go_quote.gif
to give an understanding about how to arrive with positive, ground and negative using 4 diodes...
kvant, it would probably cost him less for two more diodes than many times larger caps.



Absolutely, there are better plans to get bipolar supply. The issue with them is that bidoux expressed strong preference not to work with high voltages - and AC walwarts do not seem to come with a center tap (and ready made dual rail DC walwarts are also very scarce).
 
Jul 11, 2009 at 11:04 AM Post #113 of 188
I've finished the Dynalo using the power supply we talked about earlier. So far I have some questions to ask:

-When I don't plug any kind of source to the amp, the DC offset is something like 6V for both channels, when a source is plugged (zune) it goes down.
-For the left channel it is always under 5mV so it is ok I think but for the right channel it is more 50mV.

The BJT pnp and npn are well matched, I checked this a few times, I don't think the problem would come from there. I swithed the LEDs but it doesn't change anything.

The DC offset from the Zune is a bit less than 10mV, the DC servo should take care of it
icon10.gif
.

Is there anything less I should check before I plug my Headphones (DT880 250 ohm BTW).

Thanks a lot.

Edit: Didn't populate R65 and R66, don't use a pot yet and don't use the trimpot (I have 4 of them if it can fix my DC problem).
Edit: Breizheau suggested to shortcut the input while doing DC offset measurement. This way I get 50mV for a channel and 30mV for the other (I switched and changed some things meanwhile to have the best matching I can).
 
Jul 12, 2009 at 9:26 AM Post #115 of 188
Ok, so I have to solder 500K resistors to R65 or R66. The problem came from the LED, I didn't match them, have to buy some more now, thanks Justin.

Does upgrading the PSU is worth it ? The only difference I see between GS-1 and Gilmore Lite is the PSU so I guess it should make a difference, but i'm skeptical about high end power supply. Can anyone share his experience ?
 
Jul 12, 2009 at 10:52 AM Post #116 of 188
i believe PSU could make a big differnece. got a chance to listen to a gilmore lite with hd580/hd600. the gilmore lite just ran out of a wallmart PSU thingy (i dont relaly understand). i feel that the gilmore lite synergies really well with the hd580/600 and has a good hold onto them, but in complicated passages you really feel that the gilmore lite just needs more power for a much stronger control
 
Jul 16, 2009 at 9:02 AM Post #117 of 188
Quote:

Originally Posted by kvant /img/forum/go_quote.gif
attachment.php



I have some heat issues with the regulators. Does anyone knows how to calculate the size of Headtsinks I need on these regulators ?
 
Aug 16, 2009 at 6:08 PM Post #120 of 188
Quote:

Originally Posted by bidoux /img/forum/go_quote.gif
Since the dynalo is passive ground, would an active ground make it sound better ?


Sorry to up this, does anyone have a clue ?
For the ultra-reg power supply, is a center taped transformer 48VAC as good as 2*24 VAC ?
 

Users who are viewing this thread

Back
Top