advertisement


New PCB for shoebox size NAC preamp part I

Status
Not open for further replies.
Wilky said:
P.S. Where are all the others?
Click to expand...


In my case, I've put all the parts in bar the transistors, I really don't know why I haven't finished it off, just finished my 250s second rebuild, another CD player on the way to mod and some new crossovers to build not to mention modding the second system in the lounge.

The Starfish is such a good project it deserves full attention, my mind only on that when playing around with it, my mind is elsewhere at the moment, not sure where exactly, just elsewhere. :D

Be sure to be done when my karma says so.
 
Hi Jiim

Greetings to you from Kuala Lumpur. I have been somewhat busy...so have not had much time for all my hobbies and projects :mad: !

Anyhows.... i had been busy making sense of your ground regen thingie mod drawing. And so i ordered the bits from Farnell and started scratching my head how to plant them into the board neatly.

Your drawing was clear enough it seems mate ;). I managed to build it. Yay! I hooked it all up and i am glad to say that i there were no bad smells and sounds. Phew :p !

testingtracofish.jpg


The Traco seems to draw just under half an ampere when fed 12 ish volts. In turn it puts out just a whisker under 30V. So all is well there!

Checking the rail balance showed that the 0V was no where near where it should be. I can see why you say 1K5 "ish" in your drawing Jiim.

I was measuring 17V on one rail and -13V on the other. It was somewhat of a pain in the bum to be pulling out the "regenie", and so i thought i would work it out empirically.

Pulled out my bag of resistors and started some poke and see. What i reasoned out was that i should try some high ohmage resistors in parallel with your / my 1K5. Slowly slowly... i managed to bring that 17V down to nearly smack on 15V. I found that the magic value to add in parallel with the 1K5 for my preamp was 10K. There is probably 0.1 or 0.2 V difference now between rails :)

Checking all the regulators shows a voltage spread from 11.4 to 11.7V.

Now i am starting to work out how to case it into my 42......

Question for you Jiim.... Not sure how to word it...

Can we hook up the Traco to the Naims 4 pin din socket and get the power fed from a Snaic 4?

The reason i ask is..... sure, our wide band Traco will support the 24V from a standard Naim power supply or lower end power amp.... But will be compromising our optimised starfish earthing by connecting the traco's earth input pin to our preamps signal out earth at the signal out socket ?

Do you know what i am getting at?

Elsewise...here is a nice closeup shot of the power supply arrangement. A big thanks to you two, Jiim and Hacker. You guys left a good trail of info for the seekers to dig out.....

tracofish.jpg


Cheers

mark
 
markzb said:
Can we hook up the Traco to the Naims 4 pin din socket and get the power fed from a Snaic 4?

The reason i ask is..... sure, our wide band Traco will support the 24V from a standard Naim power supply or lower end power amp.... But will be compromising our optimised starfish earthing by connecting the traco's earth input pin to our preamps signal out earth at the signal out socket ?
Click to expand...


Nice looking build you have there Mark, very tidy!

You could certainly use a hicap to power the Traco input.

However - in this configuration the audio signal should not go via the hicap used to power the traco.

It would be a bad thing to couple the audio ground to the 'power input' side of the traco - one of the main reasons the traco is there is to provide isolation between it's input side and it's output side.



How does it sound so far?


Cheers,

Jim.
 
Hi Jiim!

I have no idea how it sounds. I am going to put it into my 42 case. That's still a work in progress. I think it should be ready next week :)

Till then...

Cheers

mark
 
Hi Jiim

I have managed to put the preamp board into the 42 case. It's all looking very nice and coming along very well. I want to go for the title of "Best Built" Starfish.... :)

I have one problem, somewhere around the volume pot. Signal is not getting out of the volume pot on one channel.....so i need to look into this next.

Whilst i had it on the bench, i tried out my old Picoscope ADC 212 on it. What i was looking at here was the quality of the signal rail. I have not looked critically into the Picoscope software settings... but i suspect the waveform captures of the positive rail which you see below are pretty representative of whats going on....

Perhaps you and Martin and Piglets Dad can comment on these waveform captures...are they good or bad?

They were measurements of the positive star rail with the scope ground lead at the power ground and the positive lead at the leg of the last tantalum cap between the positive and negative rail.

starfishpositiverail.jpg


starfishlowspectrum.jpg


starfishhighspectrum.jpg
 
Hi Jiim

Looking at the lower spectrum analysis, whilst observing it in "real time", the peaks that you see tend to "bounce" up and down somewhat. But i notice that the peak around 13 Khz is somewhat prevalent most of the time..... Not sure if its effects are noticeable.

Cheers

mark
 
Hi Mark,


I'd expect to see some crud on the 'raw rails' - particularly as I havent tuned the filter at all yet.

13KHz is unexpected.
The traco switches at around 300KHz and there should not be any other oscillating sources in the circuit.
It could be that it's an artifact of your measuring setup or equipment?
Alternatively, something in the board could be oscillating :confused: although smack in the audio band like that is very unexpected...

Luckily, it will easily be filtered out by the local-regs. I also note that it's very low-level, looking to be around 0.05V P-P from your graph.

Cheers,


Jim.
 
Hi Jiim and PD

Perhaps you guys could help me. I have myself a function generator. Its not an audio oscilator...but a good general purpose one. I connected it up to the input of the starfish. I then scoped the output of the Starfish.

The first problem i had was no output on one channel. Probing around like an amateur that i am.... i noticed that the signal in the faulty channel stopped at the pot. I removed, cleaned and resoldered the volume pot. That took care of that fault.

I measured the DC offset at the preamp's output. 1 and 3 or so mV. So i connected my AKG 501 headphones to the output of the preamp. I can hear Sine waves from approx 20 Hz to 16.25 kHz. The volume pot works well...

The problem i have is that i have an EXTREMELY hashy sine wave when i probe the preamp. I thought it might have been the function generator. So i unplugged it out of the preamp ( a long time ago i made a BNC to DIN test lead :) ) and scoped the lead. The signal going in is very clean.

But..when i plug it back in and scope the INPUT signal wires...there is a lot of hash on those signal wires. When i switch of the Traco, the signal becomes a clean sine wave again.

Any ideas what could cause the Starfish to put out a horrible waveform on the input of it??? :confused:

I will take some pictures of what and how i am scoping it.... could be interesting, some teleaudiomedicine :)

Cheers

mark
 
I suspect the Traco is radiating some out off band (300kHz) rubbish that is being picked up by your scope probes.

It is really hard to get good measurements in this kind of situation; as it is not easy to tell harmless pickup on the scope probes from a real problem.

One method is to try measuring the outputs from outside the box (say at the connector), with the case closed up ; this should shield your scope leads from what is happening round the Traco. If that looks clean, I wouldn't worry.

If there is breakthrough of more than a few millivolts, somebody who has worked with a Traco needs to comment; I have always had worries about adding switching supplies so close to sensitive circuits, so I never tried it myself.
 
Mark,

A few pictures might help us to diagnose.

Set your generator to something sensible in the middle of the audio band - say 10kHz at 1V P-P.
Take a picture of what you see at the input to the preamp.
Set the vol pot to get an output from the preamp of a similar level, and take a picture of that.

Post 'em on here and us armchair experts can blather on for ages ;)

Some good pictures of both sides of the board might help too.


Cheers,

Jim.
 
Hi Jiim

I had a brief look at it. I have been a bit busy of late, so i havent had time to take many pictures of it all.

But, in the few moments that i have spent on it.... it seems to me that the cause of my problem may be the ground regenie thingie.

Whilst nulling my picoscope, i have to short the earth wire of the scope probe to the scope tip and then via the software, null it. This gets me pretty much a "flat-line" at 0 volts.

Whilst the probe earth wire was still shorted to the tip, i touched the signal star ground point at was suprised at the amount of electrical activity there.

So the next thing i did was connect the ground lead of the probe to the signal star ground, and the probe tip to the output of the ground regenie that synthesizes the ground..... basically the begining and end of the star earth track...

Here is what i am getting....

noisyregenie.jpg


What are your thoughts?

Cheers

mark
 
Hi Mark,


I'm not sure I understand this statement: "Whilst the probe earth wire was still shorted to the tip, i touched the signal star ground point at was suprised at the amount of electrical activity there."
If the scope ground is connected to the scope tip, you should not get any activity on the scope?

On your graph, I see you are managing to measure 24mV P-P between signal-ground and power-ground?
This immediately rang alarm bells, as that link should be very low impedance.

I looked at your pictures in post 682 - it's difficult to see, but I think I can just make it out in the first picture - It's looks like you are missing the wire link between signal-ground and power ground? This would certainly account for your circuits odd behaviour....

Jim.
 
After spending many happy weeks with the locally regulated asymmetric version its time to take the plunge and go symmetric. I've got the tracos and am starting to wind the toroids. The manual says "It is critical the number and direction of turns is as per the diagrams in order for the choke to function as intended." but I'm sure I've seen somewhere that the more turns the better just make sure the three windings have the same number of terms. Anyway I've wound this jobbie which is about an inch and a half with three lots of 21 turns - anyone see any problems other than making it fit?

DSC01549.jpg
 
Hi Tim,

Apologies - my terminology is a little 'off'.
The critical part is that the the windings all have the same number of turns and the same direction.
Your choke should work fine - just be carefull to wire the coils into the pcb as per the diagram.

Cheers,

Jim.
 
Hi Jiim

Just thought i would let you know.... i just rebuilt my earth regenie. I found a rather elegant way to add it to the motherboard. It really looks slick n trick and works well now. YAY ;) !

Yesterday, i hooked it up to my modest system. I am going towards an all Quad system and currently use a Quad 34 and 405. Currently refurbishing my 57's.

In comparison to my 34, your / my Starfish really woke my system up indeed!

I don't really like describing audio things in "flowery" language.... but the first impressions were that the symmetric starfish as i have built it offers superb levels of presence and drive. It has a wonderful sense of grip and authority. And it offers really wonderful musical insight.

Alas, i will be selling it, due to the fact that i need to raise some finances.

Hmmm guys, what would such a thing be worth?? :confused:

I think i can claim the title for best built Starfish. Let me post some pictures tomorrow to back up such a bold claim!

Jiim... superb piece of kit there! You should be proud!!! :p

Thank you so very much indeed!

Cheers

mark
 
Hi Jiim and everyone else....

Here is what she looks like now.....

starfishaerialview.jpg


starfishsideview.jpg


starfishfrontview.jpg


starfishregenie.jpg


Not too bad for something that's been "bodged" eh Mick! :D

I really tried to build this baby as best i could. I bought 300 or so transistors and then matched all the transistors for gain between channels. Same goes with all the coupling caps. I matched them as best i could between channels. I'm sure you will recognize the feedback caps courtesy of an earlier group buy here. The board needs no introduction ;)! The Noble pot was bought from LesW some time ago. Oscon SP's for the symmetrical power rails. 2K ? resistor change in the feedback loop. Ground regenie. 9 - 36 V wideband Traco.

So here is the big question....i need to raise some funds to hopefully finance purchasing some nice speakers for myself. What do you guys think such a thing would be worth?

I guess it depends how you look at it. If your looking from the value of a standard 42... perhaps not much. But...if your looking at it from the point of view that this preamp stands a great chance of being every bit as good as naims top of the line preamp.... then it might be worth somewhat more ;)!

I would appreciate any feedback on what a fair price would be to try selling such a thing.... markzb@ yahoodotcom

Cheers guys

And Thank You Jiim!!!!!

mark
 
I've no idea, but the result - like atll the posters here - looks stunning and I'm sure it sounds superb. Damn well done! I suspect if you sell though, you'll regret it; for all the effort input and pleasure to be had from such a nice job, gaining a few shekels now would probably be better sought other ways.



(wish I'd bought a board now!)
 
My word, oodles of StarFish shock Pr0n :D


Glad to hear you got it working well Mark, nice job it looks too!



I've got not idea what one of these would be worth, but I must admit to at least a passing interest in what results you get - I might be persuaded to build one or two myself in exchange for sufficient pecuniary enticement.


Good luck with your sale. I'm always a little sad when I see a StarFish go for sale. I hope it goes to a good home :)
 
Status
Not open for further replies.


advertisement


Back
Top