DCB1 Build

jls001 said:
So should I parallel two caps as per above values?

1) 15pF + 100pF for live
2) 15 pF + 1000 pF for gnd

I didn't know that we should use on live lead too!

25V rated caps should be fine?

And not needed for output, right?
Click to expand...
Yes parallel them - small and big values. 25V is fine at input. Nothing required at output as there is no gain/feedback circuit in this. In place of caps, I used ferrite bid for 400KHz and above.
http://www.hifivision.com/diy/19629...ass-t-100-100w-diy-stereo-amp.html#post271676

But for DCB1 its overkill.


BTW, why you thought of protecting DCB1 from RF/EMI? Any environment issue?
 
Last edited:
Reverted back to 47K log pot for the DCB1 and I got my bass back:)

I measured the resistance across the two tracks of the pot and one measured 48.67K and the second measured 52.30K. I thought the difference was large enough so I put a series resistor of 3.3K to the lower resistance. The resistances of the two tracks after inserting the resistor was 52.53K and 52.27K, but on playing, one channel dominated completely. Also, when the pot was at zero, it did not fully attenuate the incoming signal and music was still playing at zero pot position. So I desoldered the 3.3K and reverted back to the original. 3.3K was in series with the CCW lead. Is this correct? Or should I have it in series with the Wiper? I can hear a very slight channel loudness imbalance and would like to correct it.

Will try the input caps another time as I need to plan properly how to wire it without causing a mess. Got all required caps.

Adding: I installed a red LED on the front panel for on/off indication. I had no indicator so it was sometimes hard to see if the unit is on/off against the black fascia. For those who want to implement it, connect a 4.7K/1W resistor in series with the anode of the LED and power it from the 24V output of the OPS.
 
Last edited:
How to correct imbalance?
Step 1: Measure resistance between W-CCW at zero volume. If both are zero then go to step 2 below. Other case, One will have some non zero value, 'R1'. Add that small resistance 'R1' at CCW pin on channel which was zero with W-CCW. Then both channel will have equal non zero value between W-CCW at zero volume, means practically no zero volume.

Step 2: Now measure difference of both channels CW-CCW, R2. Add difference of both channel R2 to lower value channels CW pin. So both have some equal volume at lower level, practically no zero volume if R1 is non zero.
 
Last edited:
@Om: so if I don't have perfect zero, attenuation won't be zero and some signal will pass through?

BTW, right now there's no sound when pot is at zero.
 
jls001 said:
@Om: so if I don't have perfect zero, attenuation won't be zero and some signal will pass through?

BTW, right now there's no sound when pot is at zero.
Click to expand...
Yes, if you have no perfect zero volume in any channel means wiper error at one end. There is some signal still passing through. If that is more than amp sensitivity then you will listen low volume playing.
Now you have zero volume as lowest means you have matched zero in both channel without error. So need to go step 2 above.

Then add difference on CW pin and not on CCW. That means till max volume both will have near equal volume.

Consider to circular paths around same centre. One outer is covering more distance than inner one which is same carbon material. So increase level may not always match.
 
Last edited:
Fitted the 3.3K resistor to the CW of the wiper with lower resistance, and it works perfectly. Thx for the guidance, Om.
 
Build pics for magma's DCB1:

1) Top view of completed DCB1:

ZsO4sAn.jpg



2) Close up of the DCB1 board:

SDAYu3k.jpg



3) Volume Pot Extender bracket:

sL3EmUu.jpg



4) R-Core and IEC/Power Switch

skTU8Lm.jpg



5) OPS Power Supply Board + LED

KRvoZEP.jpg



6) Inputs/Output and Input Selector DPDT

Kml8dRi.jpg



7) Rear panel (label direction to be corrected:))

JyaUCPX.jpg



It's playing nicely.
 
Nice build! That r-core is looking interesting.
BTW at input section, from toggle switch to DCB1- you twisted both channels together. Rather I suggest twisting ground and live wires (signal+return path) of each channel. Keep ground wires 2" bigger than live wire (towards switch end.) then bend ground wires towards RCA grounds.
Consolidate ground wires at input like L-L AND R-R together. Now you connected LR grounds of each source together.
What is output offset mV?
 
Last edited:
omishra said:
Nice build! That r-core is looking interesting.
BTW at input section, from toggle switch to DCB1- you twisted both channels together. Rather I suggest twisting ground and live wires (signal+return path) of each channel. Keep ground wires 2" bigger than live wire (towards switch end.) then bend ground wires towards RCA grounds.
Consolidate ground wires at input like L-L AND R-R together. Now you connected LR grounds of each source together.
What is output offset mV?
Click to expand...

DC offsets are R: 0.14-015 mV, L: 0.39-0.40 mV.

I have changed the following:
1) Separated the input Ch 1 and 2 L +ve leads.
2) Twisted the leads of the primary of the transformer. Ditto the secondaries.

Regarding grouping grounds, I gave it a thought while deciding what length of jumper to cut. It did cross my mind that I should jumper grounds of L & L of Ch 1 and 2, and jumper grounds of R & R of Ch 1 and 2. But then I thought all inputs grounds are at a common potential, so I ended up jumpering L & R grounds of Ch 1, and L & R grounds of Ch 2 input. I was wondering if my understanding is wrong?

@Sachin/Sarith: I have left the input +ve and -ve leads untwisted but physically separated from surrounding wires to avoid interferences. The reason for this is because the DPDT switch I use to select between Ch 1 and 2 is a make/break switch for one lead. So only +ve lead from RCA socket is diverted via this DPDT and the -ve lead goes directly to the DCB1 board. There is no provision in the DPDT to accomodate both leads of the signal. So the +ve and -ve leads go separate ways the moment they emerge from the RCA socket.
 
Regarding ground - think that all current have equal return path. If two current path use same channel, naturally they cross talk. Separate grounds eventually end up to PS reference 0 volt. Practically all shorted together. But that's current sink, Central point which sits in between the cross talk. So that gives more separation, low noise floor.
 
Posting pics of my completed DCB1:





The ACA and HTPC are visible in the background.

I also implemented the remote control for source selector using Quad's circuit diagram. Used different colour LED to indicate the current source - green for Tatasky and red for HTPC. Got a bunch of remotes now :)




In addition have android apps for FoobarCon and XBMC on my mobile...

Cheers!
Shantanu
 
That's it! This is the last time I'm using CAT6 wires for internal wiring. I had been hearing crackling noise from my speakers and sometimes one channel goes off completely. I suspected all sorts of things and eliminated them one by one, but in the end reflowing solder on all internal wiring of the DCB1 did the trick. Next time I will stick to multi strand wires:sad:
 
anyone tried this (or B1 for that matter) with an unregulated, stiff supply. Something like a 5a trafo with 10kuf caps, to ensure minimum voltage drop when a preamp type of current is drawn.
 
Follow HiFiMART on Instagram
Follow HiFiMART on Instagram for offers, deals and FREE giveaways!

Similar threads

NadirT
DIY Slant Wall Mount for Surrounds.
Replies
0
Views
4K
NadirT
NadirT
F
Build This Tube Amp With A Screwdriver
Replies
10
Views
7K
Amarendra
A
S
Acoustic portrait Swara Tube buffer V2 (A review)
Replies
9
Views
4K
Shammi Dang
S
S
How to connect earth pin on SMPS?
Replies
0
Views
1K
sim_tcr
S
F
My 10 Year Old Granddaughter Builds a Tube Amp
Replies
10
Views
6K
linuxguru
L
Back
Top