8-Channel Gainclone Amplifier

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
I'm a fan of the new Daftpunk album. It's a great recording with awesome dynamics and deep synthesized bass.

daftpunk.jpg


At normal living room listening level I have no issues with this 8-channel GC to drive my 4-way speakers. However sometimes my house is empty and I'd like to listen at 'discotheque level'. The sub runs out of steam without ever reaching xmax. So it's an amp problem.

DSC_2576.jpg



Hence another 4-channels are added, making this amp a 12-channels. The 4-channels are bridged to drive the Subwoofer pair. In theory the amp will be happy to drive 2 ohms load.

DSC_5400.jpg



I'm testing it with my much updated 4-seconds finger touch method :). So far so good with temperature not passing 70°C albeit the tiny heatsink. But this is winter with in-room temperature of 19°C.

DSC_5411.jpg
 
Is that the same as a puddle on the floor?
What has that (liquid wall) got to do with audio sound quality?

Well it's just a throw away subjective thought bubble - I like it because it seems to have great dynamic range and I can turn it up loud and the simplicity of the sound and great base notes make it feel from an SPL point of view like I'm at a great concert - and my pot addled memory of a great concert was that it was like standing in front of a wall of sound.
These concepts are probably far too random for you - just accept that I might not be able to explain it to your satisfaction and move along.
 
With Melbourne weather this week I wish I was in Sydney too! :)

Peter Daniel's 4780 kit is awesome both in quality and size. Can't complain really.

I have problems now with the new transformer.

- It hums and buzzes mechanically. Some quick research it seems due to DC voltage. I deployed DC blocking circuit here but it does not go away.

- The speakers also hum. I need to find out why. Already have single point of ground in the system but it still hums.

Some sleepless night this week I guess :(... I hope it's not due to the poor quality of the Transformer. Cost me heaps.
 
I have problems now with the new transformer.

- It hums and buzzes mechanically. Some quick research it seems due to DC voltage.

I have the same problem. I have dual mono amps and one transformer humms, the other doesn't. It is mechanical humm and as I red there is not much I can do, but mechanically fix, impregnate or something like that. Dip it into molten tar or smg like that :D
 
Indeed it's a mechanical buzz. ... Some luck... After opening the enclosure and poking here and there, lifting the Toroid etc. I noticed that the vibration buzz came from inside the Toroid itself. Also, when not under load it's extremely quiet after I put the DC blocking. So that circuit had done its job.

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


In my case I use a 10A diode bridge and 4700uF 25v

So I proceeded to remount the Toroid and *tighten* it using proper 'garage' tool. This seems to compress the winding and the buzz is gone!... :) I still can hear it if I put my ears 2-3cm from it. So no problems at all!

I'm quite happy :cool:
 
Good to see you are happy, makes me want to consider bridging in my system.
I have not used bridging before and dont understand whether different gains across the driver set would result in amplitude variation between the drivers as the volume is varied? I understand MiniDSP only attentuates the input of Max 0.9V or 2 V (balanced 4V or 8V ) and am not clear on whether the change in signal to noise ratio is significant when using higher gain amplifiers? Would you please explain the gain structure you have used for the full system? Does the bass with its bridged amp go louder relative to the other drivers when the system volume is increased? Is there any detectable change in the noise floor amplified by the higher gain at low volume?
 
I have to correct myself, they were not bridged but paralleled!

I had thought that the load was putting a strain to the chipamp which it resolved, but not fully. What I need is actually a bridged - paralleled configuration. Perhaps something like this, although I would do this non-inverted:

DIY BPA300 6x LM3886 300W audio Amplifier

So two paralleled amps, then bridged. But that's for later.

I might even try to do single amp bridged.
 
You might want to think about adding a few tens of thousands of uF on the power supply. I have a stereo three way (8 amps) Gainclone with about 40,000uF and a 5 channel two way (15 amps) with 75,000 uF on Chipamp 3886 boards, both with hardwired active crossovers. The extra capacitance helps make up for sag on the power supply rails.
 
- The speakers also hum. I need to find out why. Already have single point of ground in the system but it still hums.

I am no expert on this though am currently rebuilding my amps and recently revisited the guidance provided when I set up my system:

Place the supply components and wiring away from the signal components and wiring.

The transformers center tap connection to the filter caps and signal earth requires heavy wiring or a bus.

Take the amp supply from the caps, not the bridge.

Twist the signal cables. Re-reading my notes it doesn't say to twist the power cables, but I did that.

Parallel the bridge diodes with a 100nF capacitor.

Use the correct transformer connection to the bridge. My notes show the schematic from here: Power Supply for Power Amplifiers
 
A second point after looking at the pictures again.
Like Fenris I was surprised not to see supply filter capacitors half the size of a fist in your system.

Rod Elliot indicates filter capacitors of 4,700 uF / Ch are required.
Elliott Sound Products - Linear Power Supply Design

The Sigried Linkwitz LM3886 supply calling for 2 off 56,000 uF for 10 channels is consistent with such sizing.
http://www.linkwitzlab.com/images/graphics/3886pwr.gif

I am building Heao LM474C for which the manufacturer recommends 22,000 uF for two off LM4780 which is about the same capacitor sizing.
Heao.it

The Audiosector schematic shows an outlier filter cap size at 10uF. Only when an optional snubber is applied are the filter caps recommended to be like the others at 10,000uF.
http://www.audiosector.com/lm4780 psu.pdf

I'm no expert though understand that the LM4780 can manage supply variation and the design would be what is needed for the amplifier.

What is not clear to me is any system effects of permitting what must be a 'higher' level of ripple current noise near the other components and wiring?
 
Well I'm almost ready to throw in the towel. While the 8-channel setup was practically set-and go since day 1, adding another 4 channel (2x parallel) had been a completely different story :(

I'm sure its due to additional ground loops introduced by adding more channels. I've tried star grounding etc. to no avail.

I'm back to the old setup and wait for another go when time permits. I might even simply separate the PSU or enclosure for the additional 4 channels.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.