New Frankenproject -

[Mechanic]

HI Pete,

I will sleep better tonight now that the signal flow is straight!!!
However I still don't see where I'm gunna plug my headphones in
You will need to add the bypass caps in for each ic otherwise when you convert from schematic to pcb they won't be there and you do need them, rule no.1! Rule no.2 when you route the traces to the bypass caps, it must go in this order, power source> bypass cap> pin on ic. I don't want to see any bypass caps where the power has been applied to the pin first with a trace going back to the bypass cap Yeah I know, I'm anal!

Cheers Mick

Cheers Mick

[Pete the Builder]

The even newer, more improved, less buggy QED schematic - with ANSWERS!

You're all a bunch of *******s! Talk about tough bosses! Pulling my beautiful schematic to pieces before my eyes! Boo hoo hoo...

I've attached the latest schematic below. I hope I've addressed all your very much appreciated feedback and comments. To make sure, I'll try to answer youse all one list of problems at a time.

If I've missed anyone out, or missed making a change, please let me know. Again. Sigh.

OK, heeeeeere we go...

Hi Pete your schematic shows nothing but 6 red dots in the pink box.

Regards, Ian.

I'm not sure why you're seeing 6 red dots in a pink box, Ian... Do you always see red dots in a pink box? Or do you see pink elephants with red toenails?

Seriously, though, that sounds like an issue with your PDF browser plugin (if you're looking at the PDF in a browser window) or your PDF application (if you're downloading and opening the PDF on your PC/laptop/Mac). It's not rendering (drawing) the objects inside the box.

You'll need to get the latest versions of the PDF software from the developer. If you don't have an "Update" type of entry under the "Help" menu (if it's a standalone program/app), then go to the developer website and get their latest reader. I hope that makes sense.

I have some options to set the compatibility back a few versions, but then other objects don't render properly in everyone's browser plugin! I suppose that's more democratic, but... Update is your best solution, I think. Other folks may have more ideas?

Next.

Just to help you out Pete, Here is a picture with five problem spots circled. If you have the QED schematic you'll be able to quickly see the errors in your version. I hope this helps you.
I should piont out these are not the only errors.

Thanks very much again, Mick. I've addressed the 5 problem spots now, I think. Some of them are duplicates that other folks have mentioned too. Some of them are detailed below. But thanks for going to so much trouble.

Love the wiggly circles... were you drinking (or smoking) when you drew them? OT : I suppose you know about Giotto the painter? The story goes, he was told to present an example of his work to the Pope, in order to be approved for a job in the Vatican. Instead of painting a picture, as you'd do, when the courier came to pick up his sample, he just picked up a brush, dipped it in paint, and in one sweep drew a perfect circle on the canvas. When the Pope saw how perfect the circle was, and how he'd painted it in one swipe, Giotto immediately got the job. There's even a saying, I understand... "As round as the circle of Giotto"...Sorry, where was I?

Our next contestant is :

Ok ,
D 2 Not marked on Drawing above IC1
D2 consists of 2 ea BAV99 ,
C9 should be 100P,
Between R13 and R20 there should be a direct connection, Unless this is one of those mistakes that BW had inserted. Could be taken care of with a marked out line for reference going forward so as not to forget. I do not know ?
Please confirm my findings.

I just forgot to re-display the diodes due to cramped conditions in that part of the circuit. Same with the cap - I moved the labels out of the way and forgot to restore them. D'oh!

Good catch on the R13/20 problem. Mechanic and Midas also caught that too. Quite a popular mistake on my part, I got real creative with those diode packs, didn't I?

And now it's...

HI Pete,

Just a few mistakes I have noticed.
As Kahles has stated the feedback loop on the preamp(ic1) is all stuffed up!
C5 should be 2.2uF not nF.
Pin 20 on ic5 should connect up to the junction between r19 and r24.
I am assuming that all ic's have their own bypass caps on the power rails?
The DAC lines go nowhere, where do the headphones hook up? Must dig out the original schematic.....
C14 should read 10,000uf, not 1,000.Also up the top with the line of ic's, try and line them up so the signal flows straight Corners create harmonics

Cheers Mick

Good save on the missing SNUB_MON connection. Whew! That was a really stupid mistake. I try to triple-check my schematics with unconnected pin triggers (so I get a list of net names that have only one connection), but I totally didn't even see that one. Thanks!

The note on the original schematic indicated that the schematic did not include bypass caps, so I haven't bothered to add them in yet. The power supply is exceptionally quiet, so there aren't as many generated noises on the supply we need to be catching (at least, not many PSU-generated ones!). But I'd be uncomfortable fabricating this circuit without quite a few decouplers. Still, the whole thing is fairly low noise, except for the front end, which is noisier than your neighbours after midnight!

The DAC lines go nowhere, that's correct. Strange, the DACs are pretty quick and useful in that MCU, but obviously BW does the analogue thing better than I do the digital thing...

C14 is fixed too.

I sorted out the lines for you, precious! Now all the left-turning lines (the ones that make odd harmonics) are matched by right-turning ones (even harmonics). So the harmonics cancel, as everyone knows!

And finally (for now, at least) :

Pete,
The error here that Mick has also circled is that the diodes aren't connected correctly. They should be set up so there are two sets of diodes in series, in anti-parallel between the negative input and output. This effectively gives unlimited feedback current once the output exceeds two diode voltages, stopping the amp from going into saturation so it recovers faster.

Midas

Edit: Also and closer to what Kahles was talking about R20 and C9 should be across negative input and output.

Midas, I really appreciate you taking the time to explain what the problem is and how it's supposed to work in the circuit. I had no idea about why I keep seeing antiparalelled diodes (I had to check my spelling on antiparrallellledd) in so many circuits, but your explanation was so clear I had an AHA!! moment. Thank you - I love AHA!! moments. They make my day.

OK, so I think that just about covers it. I hope any further errors are not mine. But if you see something in the schematic below that's still not right, please point it out - and if it's important, if you could take the time to explain for other people (like me), it's very much appreciated.

Thank you everyone. I sincerely apologise for having so many mistakes in one diagram. I swear, I generally check things so many times so there are zero/zilch/nein/none/zippo/no errors in my design work... So when I make something public, pop goes my average!

I really do appreciate everyone's help. This will hopefully make this a great, stable hardware platform we can all use one way or another.

Cheers,
PtB

QED Schematic 1.0003 Jan_2013.pdf

[Davor]

Surge protectors such as ICP-N15 are faster and cheaper, and they work forever

[Mechanic]

Mick, the diode D27 will destroy the battery at reverse connecting. The fuse should blow quickly because the battery is expensive. I'm not familiar with thermal breakers. How quick they break the circuit?
P.S. I searched the WEB. This is not suitable because they have high ESR.

HI Mike,
li-ion batterys can withstand some pretty horrendous short term abuse. I just had a quick look and one thermal breaker I found 1.5A breaker had a 0.12r contact resistance, I can live with that. These things trip pretty quickly and I have used them for this purpose before. Beats carrying a pocket full of fuses out bush!

http://www.jaycar.com.au/productView.asp?ID=SF2281

Cheers Mick

[Midas]

Pete,
I'm definitely over it.

Ok ,
D 2 Not marked on Drawing above IC1
D2 consists of 2 ea BAV99 ,
C9 should be 100P,
Between R13 and R20 there should be a direct connection, Unless this is one of those mistakes that BW had inserted. Could be taken care of with a marked out line for reference going forward so as not to forget. I do not know ?
Please confirm my findings.

The error here that Mick has also circled is that the diodes aren't connected correctly. They should be set up so there are two sets of diodes in series, in anti-parallel between the negative input and output. This effectively gives unlimited feedback current once the output exceeds two diode voltages, stopping the amp from going into saturation so it recovers faster.

Midas

Edit: Also and closer to what Kahles was talking about R20 and C9 should be across negative input and output.

[mickstv]

Thanks Carl.


Here you go Pete. All problem areas I haven't fixed them for you but if you have bug's schematic you'll see the mistakes in the circled areas.

[mikebg]

Or better still a 1A thermal breaker, self resetting.

Mick, the diode D27 will destroy the battery at reverse connecting. The fuse should blow quickly because the battery is expensive. I'm not familiar with thermal breakers. How quick they break the circuit?
P.S. I searched the WEB. This is not suitable because they have high ESR.

[Mechanic]

Pete, connect a fuse in series with lead 1 or lead 2 of supply connector.

Or better still a 1A thermal breaker, self resetting.

[Mechanic]

HI Pete,

Just a few mistakes I have noticed.
As Kahles has stated the feedback loop on the preamp(ic1) is all stuffed up!
C5 should be 2.2uF not nF.
Pin 20 on ic5 should connect up to the junction between r19 and r24.
I am assuming that all ic's have their own bypass caps on the power rails?
The DAC lines go nowhere, where do the headphones hook up? Must dig out the original schematic.....
C14 should read 10,000uf, not 1,000.

Also up the top with the line of ic's, try and line them up so the signal flows straight Corners create harmonics

Cheers Mick

[mikebg]

Re: Fuse for battery

Pete, connect a fuse in series with lead 1 or lead 2 of supply connector.

[KahlesMilcon]

Ok ,
D 2 Not marked on Drawing above IC1
D2 consists of 2 ea BAV99 ,
C9 should be 100P,
Between R13 and R20 there should be a direct connection, Unless this is one of those mistakes that BW had inserted. Could be taken care of with a marked out line for reference going forward so as not to forget. I do not know ?
Please confirm my findings.

[IBGold]

Hi Pete your schematic shows nothing but 6 red dots in the pink box.

Regards, Ian.

[Pete the Builder]

Question and suggestion about moving to the appropriate forum(s).

OK, last time you have to look at my ugly icon mug for a while. (I'll have to change it to something smaller, or else learn how to multi-quote my replies!)

I'm getting a bit confused with all the projects I've got my fingers in right now.

WOuld anyone mind if we moved the discussions away from this thread and into the appropriate forums?

I'll create some new threads in each forum so that the confusing multiple issues can get some attention, rather than continually appending to a single discussion.

I'm thinking that the QED forum could have a new "2013 Hardware" thread to begin with, for discussing this schematic and potentially the PCBs... and possibly also a "software" thread once anyone's able to get a board made to the point where software is an issue.

I'm working on the QED PCB layout in the background - i.e. whenever I can sit up! - but it's a real doozie and I want to offer everyone at least a "potentially" really stable and reliable (as far as it can be) working PCB design to start with (unless anybody beats me to it!)).

I'd think the other projects - UPIM, etc - would also benefit from a fresh start. We don't lose anything by it, and we gain a lot by making the forum more like a forum - where people can post individual questions or problems as separate posts, rather than just appending to horribly long threads. If no-one objects, I guess that could be a good next step.

Thoughts? Problems? Solutions? Money?
-Pete

[Carl-NC]

Mick et al, if you post images please post them as attachments and not links to off-site sources. Those off-site sources disappear eventually (quickly for image-hosting sources) leaving a thread with lots of missing images. Also works smoother, for folks like PTB.

[Pete the Builder]

Pete it is our warped sense of humor they do not understand keep up the good work.

Regards, Ian.

Thank you mate. Glad to know you're lurking around here..

I haven't read your last email yet. It's on my todo list, I swear!
-Pete