Elektor Q-Watt project

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Hello,
I subscribe to Elektor, noticed this project, so while looking for the documentation, I see that they opened this one up for everyone to download. Looks to be very nice implementation of the LME49811.
Q-Watt Audio Power Amplifier - ELEKTOR.com | Electronics: Microcontrollers Embedded Audio Digital Analogue Test Measurement
I noticed that they used one pair of bjt's in a TO-3P package, so I was surprised at the Q-Watt ratings using only one pair of o/p's. Went looking for the o/p bjt's Semelab MG6330-R/MG9410-R, Farnell has them pretty cheap, as well.
These are 200W/260V/15A parts. I also noted that they make a 300W part in TO-3P and a 400W (MAG6332/MAG9412 in a TO-264.
Might have been nice if they mentioned these facts in the article.
The hfe is very linear, beta droop is good, Ft is good. ( I guess Ring Emitter equiv.) No Spice models. From what I see, cost/performance is a bit better than OnSemi.
One critisism that I have with this Elektor project, is it is only a pcb or assembly, leaves the readers to figure out all the rest, PS,chassis.
At least DIYAudio forum, have got it figured out, offering sources for cases/HS/PS's assemblies, etc.
 
Last edited:
Hmmm???

I just photoshopped the two PCB sides on top of each other,
and realised that they have actually routed the audioground to the star point via the chips GND pin???
I have let myself understand that that pin actually emits noise from the chips internal guts...

Any opinions on that?
 
Hmmm???

I just photoshopped the two PCB sides on top of each other,
and realised that they have actually routed the audioground to the star point via the chips GND pin???
I have let myself understand that that pin actually emits noise from the chips internal guts...
Any opinions on that?

Chip guts does not emit noise. Noise in circuits is generated by unwanted or extended current loops. Therefore when designing PCB, current loops should be avoided as much as possible. This particular circuit topology (by looking at schematic and drive IC datasheet) determines current loops to be minimal if common ground point is designed around audio ground of the chip. And that is exactly what author did.
 
Does the chip use its GND pin for the audio signal?
Or does it use it for something else?
I have let myself understand that it does not use it for the audio signal.
It uses it for something else, which emits noise...

I might have let myself understand the purpose of the chips GND pin wrongly.
But I still try to keep it clear from the audio signals in my designs...
Which should then be a huge problem.
However, by luck I have managed to keep that mistake unmeasurable.
 
Ground pin on integrated chips like this one is the common return for all other signals this chip has. This means, the sum of all out and incoming currents (minding the direction) from/to this chip is equal to common current flowing to ground pin. The LME49811TB itself is just integrated power amplifier, therefore current flowing from its speaker output will return to its source through ground pin. So it is used for audio signal as you refer. There is no "something else" within the LM amp, it is simple integrated circuit with no built in high frequency noise sources which would couple to audio circuits.

I am no sure what you mean by keeping ground clear from audio signals, but you might be misunderstanding the concept of common return in electronics circuits which some of these are named ground like on this example. :cheeky:

As for the subject, I have decided to build this amplifier as its parameters are amazing in terms s/n ratio and distortion level. Simplicity of this design with careful selection of external power transistors should produce same results as listed in Elektor article. In fact, being Elektor reader since 1996, this is the first time power amplifier design is worth checking out. I expect to achieve results which should surpass many commercially available amps I had in past 20 yrs.

The pcb's and components are on the way. :D
 
Hey, I have to agree with you, this amplifier being the best of the Electors implementations ever, I have been reading Electors since the beginning of the -80ies... Need some old issues??? ;-)

Anyhow,
I have let myself understand, that the chips GND-pin is the return to the ShutDown of the chip.
Hence everything injected to the ShutDown will be sent out from the chips GND. So if one wants to return the shutdown directly to the audio ground, the Electors way is a good way to do it, in my oppinion, but hey, I might still be wrong here...
 
Administrator
Joined 2007
Paid Member
This ?
 

Attachments

  • Capture1.JPG
    Capture1.JPG
    134.9 KB · Views: 1,560
Yes, this is it.

Btw, I have compared price for Elektor kit Q-watt combi (110656-72, $244 regular, $219 for member) with set of parts from digikey (with exception of T4,T5 transistors which are available from Allied) I meticulously put together on vendor website and it turned to be $252 including PCB from Elektor but still does not include all small mechanical parts related to heatsink mount of components (screws, washers, thermal pads and paste not included). So at this point I ordered kit with parts from Elektor, as shipping within US ($5 for Ground) is also cheaper for some reason comparing for bare PCB shipping which are freighted from overseas for $20.

Then comes a set of "big" parts not included in the kit, namely: power supply bridge rectifier, bus caps, transformer, heatsink and enclosure. I found 500VA 2x40V toroid on amazon marketplace for $87 ea. plus $19 shipping for two pieces altogether, not bad considering the package weight (19lbs/9kg). The exact Fisher heatsink can be ordered from newark for $34 ea. plus $20 shipping for 2 pcs. As for the bus caps, I found on digikey 10mF/100V model with lowerst impedance of 7mohm which will ensure best dynamic performance free of compression at highest output levels, these are expensive around $36 ea. For stereo amplifier, you are going to need 8 of those so it is probably the highest expense. Bridge rectifier is just $7 ea so it wont break the budget.

Finally the enclosure, I still have not decided yet but I am looking over diyaudio store for one of those full size 12 or 15cm height available for preorder, alternatively there is some selection of these on siliconray website. Will decide after I get all the parts.

So here is a complete amplifier cost breakdown, for those who are interested:

$244 ($5 s/h) for Q-watt combi (110656-72) stereo kit from Elektor
$87 x 2 ($19 s/h) for 500VA 2x40V toroid transformer, 2 pcs
$7 x 2 ($5 s/h) for bridge rectifier, 2 pcs
$37 x 8 ($25 s/h) for 10,000uF/100V bus caps B41560A9109M, 8 pcs
$33 x 2 ($20 s/h) for Fisher heatsink
----- Total: $863 (including s/h)

and that does not include enclosure which may vary but just for the sake of example picking nice one 4U size in Deluxe finish from the website store for around $350 (including s/h), the total comes to little over $1,200.

If this amp will truly perform as advertised, this might be a bang for the buck considering prices for 300W per channel (@4ohm) off the shelf designs.
 
Art-Tech

No Spice models. From what I see, cost/performance is a bit better than OnSemi.One critisism that I have with this Elektor project, is it is only a pcb or assembly, leaves the readers to figure out all the rest, PS,chassis.At least DIYAudio forum, have got it figured out, offering sources for cases/HS/PS's assemblies, etc.
Art-Tech
 
The Elektor article contains all required information about the parts recommended for power supply suitable for this amplifier. As for power supply topology, I am sure folks who know how to use Spice models can figure this out, for those who do not, here is initial diagram of power supply I am planning to use in my build.
QWattPWS1.jpg

The 120/240 voltage selector switch S101 is optional and shown here for the purpose of hard-wiring power supply for either of the voltages. The initial value for the F101 fuse is 6.3A (for 240VAC) or 11A (for 110VAC) slow blow for both voltages. The exact current rating for this fuse will be determined later. The main reason for it is that with such high capacitance of bus capacitor bank, I expect high inrush current when amplifier is turned on. If selected transformer impedance will turn to be very low, I might need to use inrush current limiter which will be designed after the need is determined.

As for enclosure, I have initially selected Deluxe 4U amplifier chassis available from diyaudio store. However, before I spend $300 I need to make sure, all main components of this amplifier will fit inside the enclosure. From store website I found the internal enclosure real estate is limited to 360 width by 300mm depth. This should allow me to install amp guts in few different configurations - some shown below. Only one most optimal configuration will be chosen to provide best supply performance with minimum interconnection length as well as interference from adjacent modules and parts. I have not marked connections yet, neither decided about front and back of the amp in each configuration. This will be determined next.

Two configurations with amplifier board positioned horizontally:

verA.PNG


verB.PNG


and three configurations with amplifier board positioned vertically:

verC.PNG


verD.PNG


verE.PNG


and additional check for space on the enclosure heatsink to fit amplifier PCB in arrangements C, D and E:

HeatsSide1.PNG


Note: all drawings based on internal dimensions of Deluxe 4U enclosure and its heatsinks. I have abandoned idea to use Fisher heatsinks as diyaudio the enclosure has these already.

Seems like Deluxe 4U chassis can be a good choice.
 
Forgot to mention, the shown power supply except S101 and F101 is just for one channel amp. For stereo configuration we need two of these, hence two transformers and 8 large caps shown on preliminary chassis layout drawings. The fuse rating I mentioned is for two supplies in parallel fed through one fuse of course.
 
Finally after 5 wks of waiting, the Q-Watt combo kit has been delivered. I already have all other accessories including: power supply part, awesome large VU meters, speaker biding posts, RCA and headphone jack sockets, even golden knobs and chassis feet jacks except chassis itself which design is being wrapped up as we speak and soon be sent to shop for custom machining.

I am planning to hold final design of the power supply, specifically the large capacitor bank carrying board until chassis is made and delivered. This is to make sure all components and assemblies fit where they supposed to.

QwattConcept1x.jpg


More coming soon. :D
 
Last edited:
$87 x 2 ($19 s/h) for 500VA 2x40V toroid transformer, 2 pcs
$37 x 8 ($25 s/h) for 10,000uF/100V bus caps B41560A9109M, 8 pcs
That is a significant turn-on surge current, could blow a breaker in your distribution panel.
I suggest a soft-start ckt, to limit the inrush I
A few references for you:
1) Pioneer high power amps such as Spec-4, SX-1250 or SX-1050 and others( You can get SM from www.hifiengine.com)
2) TI/National app note, AN1849-An Audio Amplifier Power Supply Design
3) This forum has Soft start ckts for sale as well

Good luck
Rick
 

Attachments

  • Power-Entry&soft-start.pdf
    26.9 KB · Views: 289
  • POWERENTRYASSY-A0-27-top-assy.pdf
    37.5 KB · Views: 217
  • powerentryassy-a0-bom.txt
    1.4 KB · Views: 170
Last edited:
Thanks for good suggestion Rick. Indeed inrush current can trip the breaker, that is why I include inrush limiting circuit as I mentioned btw on my initial post. I happen to design SPS circuits from time to time for living and I have couple of soft start solutions here so implementing one of these in Q-Watt amp should not be a problem. Again, I appreciate the comment and accompanying materials which are always very helpful. Will keep you posted and appreciate your input.
 
I finally received long waiting enclosure. It looks like $million$ dollars, champagne aluminum front with professionally printed labels and plexiglass window w/artwork for UV meters, looks like hi-end gear from top audio manufacturer. I will be doing final assembly for few weeks now, lot's of fastening prep and work as this is 1st amp of its kind and not everything was designed in Solidworks. Once completed, I will post some pictures. For now just a teaser of what's comin' :)

20140613_222607_1.jpg

~Patience is the key ~
 
Last edited:
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.