Star ground

[naula]

At first, forgive me, too many whiskeys for the post ...

I've build Anthony Holton's sym amp (one ch). It works fine, heat sink is warming slightly at bias current, which is ok. I have some buzzing noise with one channel prototype. This noise isn't very big problem, it cannot be noticed from my listening place. I know this noise is coming from power supply, it's 50/100 Hz.

I'm now asking, is the source of buzzing NOT star grounding in the differential stage. You can see the schematic in here. There is in my DIY PCB (own design) C11, C7 (Anthony's schematic: C8, C9), differential stage zener D3 (A's sch: ZD2 and resistor R35 (A's sch: R9) sharing the same ground wire. It's not sub star in my own made board

Is it worth a try to make sub star for these? Or would I try couple of more wires to main star?

And the last question: Where is the best place for main star point, center tap of the transformer or the center of the supply capacitors (as close as possible)?

There is couple of pic of the project:

Testing with one pair of expencive fets

All fets assempled on the circuit board

Measuring the bias current (20.2mV/0.22ohm=91,8mA)

If you guys know good star grounding articles please tell me the link.

- Antti

 

[rwbrown1]

Try:

http://aikenamps.com/StarGround.html

for more information, check out:

www.tnt-audio.com/clinica/ssps1_e.html

Both of these are good reading..

Good luck!

 

[Centauri]

Main star earth point should be at centre of main filter caps. Earth return from speaker should be run here seperately. The earths of the caps and zener should be a separate return as well as that of R35 (this is the critical one which should not share any other earth current).

Cheers

 

[carlmart]

Main star earth point should be at centre of main filter caps. Earth return from speaker should be run here seperately. The earths of the caps and zener should be a separate return as well as that of R35 (this is the critical one which should not share any other earth current).

Main star point should NOT be at centre of main filter caps, as there are currents acting there that might make it noisy.

It should be a few centimeters away, though equidistant from both caps.


Carlos

 

[naula]

... R35 (this is the critical one which should not share any other earth current).

I think there is the worst mistake on my PCB. Maybe I have to do new circuit board.

Main star point should NOT be at centre of main filter caps, as there are currents acting there that might make it noisy.

It should be a few centimeters away, though equidistant from both caps.

I got the same info from another forum. I will do this. At this time my star point is about 15 centimeters away from center of the caps.

Thanks for the info! Now I know much more about star grounding

- Antti

 

[naula]

Hi!

Now I have designed new PCB layout.

An externally hosted image should be here but it was not working when we last tested it.

Part numbering is the same as Anthony's shematic.

I made star ground for the signal ground. This is connected to main star with 10 ohm resitor (R9). From there a wire will go to the main star at the power supply. In this version MOSFETs are not placed on the board, they are placed at the heat sink with short wires.

Will this work? ... and maybe you have some comments for the board.

 

[Ryder]

online articles on grounding

If you guys know good star grounding articles please tell me the link.

There are a couple on star grounding. Some are tube related.



Grounding/System Interconnection

* Kurt Strain: (from diyparadise site) Grounding and Shielding Primer for Tube Circuits: http://www.sonic.net/~ktstrain/Groundtutor.htm
* R.G. Keen, Star Grounding in Tube Amplifiers: http://www.geofex.com/Article_Folders/stargnd/stargnd.htm
* Advanced Energy Ind., Grounding: http://www.q-audio.com/grounding.pdf
* Ken Simmons, Ground Loops and Home wiring, An Introduction: http://www.smr-home-theatre.org/Ground-Loops/
* Pete Goudreau, Superconductor Interconnects (from SoundStage): A Technical Discussion: http://www.sstage.com/articles/pete01.htm
* Epanorama: Tomi Engdahl, Ground Loop Problems and How to Get Rid of Them http://documents.epanorama.net/documents/groundloop/index.html
* Randall Aiken, Star Grounding: http://www.aikenamps.com/StarGround.html
* Rane, Grounding and Shielding Audio Devices: http://www.rane.com/note151.html
* Rane, Sound System Interconnection: http://www.rane.com/note110.html
* Pete Millett, The Power Line Interface: http://www.pmillett.addr.com/images/ga_powerline.PDF
* Rane: Practical Line Driving Current Requirements: http://www.rane.com/note126.html
* Audio Cable: http://www.euronet.nl/~mgw/homepage/uk_index.html
* Bill Whitlock, Interconnection of Balanced and Unbalanced Equipment: http://www.jensen-transformers.com/an/an003.pdf

 

[peranders]

naula, check the point where the feedback is connected. I recoemmend that you move R40(?), R11(?) to the connector of the speaker. This may lower the dist becasue you inluce all resistances from the scource resistors and their traces. Just move the feedback to the "SPK" in your layout.

I would also remove the ground "wings". The acts only as antennas. Remove also the copper islands with no electrical connection. If you have the opportunity, use a real groundplane. Just put tape over the pcb laminat on the upper side. Real easy! Look at my hand made pcb's below.

http://home5.swipnet.se/~w-50674/hifi_pics/hifi_100pr/qrv01_old.jpg
http://home5.swipnet.se/~w-50674/hifi_pics/hifi_100pr/qro_pcb_top.jpg

 

[naula]

Just move the feedback to the "SPK" in your layout.

This means redesigning the whole board. Is it really worth of it? My earlier design sounded fine (much better than my current amp) even if the feedback was very much longer away from the speaker connector. Only problem was the hum/buzzing noise.

 

[naula]

Peranders:

I have made the changes you recommended. Here's the pic of the new layout.

An externally hosted image should be here but it was not working when we last tested it.

Is this better than earlier one? (There is still long trace in the feedback circuit.)

Ryder: Thanks for the grounding links.

- Antti

 

[Kilentra]

I'm using Anthony's boards and there still is a little bit of hum. I have imperfect power supply wiring myself. The problem may not be specific to your PCB layout. I'd look at perhaps isolating the earth grounds with a diode bridge, or ensuring a good power star ground[?]

 

[traderbam]

Hi,
Running fb from SPKR is good. But move R11 close to R10 - otherwise you have a long aerial on the base of your diff pair.

Where is your Zobel network? This is an HF load for the amp.
BAM

 

[naula]

Hi,
Running fb from SPKR is good. But move R11 close to R10 - otherwise you have a long aerial on the base of your diff pair.

Where is your Zobel network? This is an HF load for the amp.
BAM

I think I will move the SPK connector to center of the board, next to C12. This will shorten the feedback circuit.

There is no Zobel network in the Anthony's design.

- Antti

 

[traderbam]

think I will move the SPK connector to center of the board, next to C12

Yes you could do that, but my point is that the length of the track off the base is critical not the length of the whole feedback path. In fact I'd say leave the spkr term as far away from the input stage as you can to minimize inductive feedback - the currents in the speaker wire will be very large, or rather the transconductance of the amp will be very high compared with the voltage gain.

IMO the circuit should have a zobel. Otherwise, what is providing damping for the output stage at high frequencies?