I decided to implement the post filter feedback. This needs a higher gain and supply voltage of the input stage.
Well without post filter feedback the inductors have a high influence in performance. While the MA5172/3 had a pretty good performance, the ML1565 had huge problems with lower frequencies. With post filter feedback these two inductors had almost same performance results (of course not over the same power range). There is also no more need for choosing the right inductors depending on the desired load.
But the main point: The sound got better
Yes it still fits inside the Fischer case, but without RCA plugs. You need to use the JST connectors instead.
I will show you a cased amp when I got everything infront of me ;-)
So does the upright RCAs makes any sense then at all? Just asking for practical reasons.
Actually I'm not a huge fan of RCA jacks inside a housing but the many private messages from the DIY community here showed me that they totally want to have it.
For the Fischer case yes, it's useless. But for many other cases or mobile applications it's a good way.
Plus, this board is replaceable. It's not married with the amplifier and I bet there will be more boards soon. The pinout of the connector is not a secret so everybody can build boards by himself. Even on lab cards, because it's a regular pinspacing of 2.54mm (0.1").
Yes this "old" board will not longer fit to the new amp.
But you can use this "old" board as standalone, too. It will not end up in trash.
BTW: anybody interested in buying the old revision of this amp? With or without DSP and µC maybe?
I will make a good price
Hi irribeo,
I didn't notice any change in temperature change in combination of PFFB. I also didn't notice a drop or rise of bass gain with temperature change but I can have a look at this.
I added the fan control because there are some situations where both revisions (with and without PFFB) need a active cooling. For normal listening I bet the fan will never power up but under heavy load the fan is neccessary.
See pic attached. Because of a mistake in board layout the DSP board needs to be soldered 180° turned around. Because of this the board has 10mm more width at this point and two connections need to be made by wires.
The height of the DSP board is 40mm. Plus the spacers and the amplifier board itself we end up at 46mm.
Without DSP Board the total height of the amp is 32mm.
The DSP is not from TI. It's a ADAU1701 from AnalogDevices.
The Software is called "SigmaStudio" and can be downloaded for free. No need for a licence.
For connecting the DSP to the PC you can use a FX2LP CY7C68013A Mini Board (see pic attached) from ebay (around 5USD) or you can buy this board directly from me.
Regards
Bernhard
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
I have seen frequency or gain graphs for temperature, probably for tpa3244/3250, regarding benefits selecting pbtl mode. But my regular laptop had major meltdown, literally, so no easy browsing back option at the moment.
Continuous or burst/audio, I don't recall. Because bassoutput was limited when temperature went up, I wondered about that a little reading about bass performance little inductors and little inductors quickly being the hottest components on the ampboard too. Bit far-fetched probably, like you said
Nah, it's about power derating when pushing continuously at high levels. This is normally irrelevant when playing music.
The plots don't mean that output power gets lower when temperature rises, they just show what frequencies would heat up more than others.
(Of course one could say output power decreases with temperature when taking Rds_on of the output FETs into account, but this would really only matters when having 2R loads connected. When doing so, supply voltage should be decreased anyway and the 3251 is the better choice then)
The plots don't mean that output power gets lower when temperature rises, they just show what frequencies would heat up more than others.
(Of course one could say output power decreases with temperature when taking Rds_on of the output FETs into account, but this would really only matters when having 2R loads connected. When doing so, supply voltage should be decreased anyway and the 3251 is the better choice then)
Good evening guys,
I'm sorry for no updates the last weeks.
I told you that there will be a new revision and I already showed you 3D simulations but I decided to upgrade the first revision instead.
Upgrades are:
- Post Filter Feedback
- DSP Board is now exchangeable
- Fully symmetrical Input Buffer Stage with real symmetric power supply!
I also finally started writing the firmware for the µC. Still not finished but the most of work is done.
First pics of new revision are attached
I will update you with more info soon.
Regards
Bernhard
I'm sorry for no updates the last weeks.
I told you that there will be a new revision and I already showed you 3D simulations but I decided to upgrade the first revision instead.
Upgrades are:
- Post Filter Feedback
- DSP Board is now exchangeable
- Fully symmetrical Input Buffer Stage with real symmetric power supply!
I also finally started writing the firmware for the µC. Still not finished but the most of work is done.
First pics of new revision are attached
I will update you with more info soon.
Regards
Bernhard
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.