World-Designs-Forum  

Go Back   World-Designs-Forum > DIY Projects > Sources
Register FAQ Members List Calendar Gallery Today's Posts Search Today's Posts Mark Forums Read

Sources Your DIY source designs (turntables etc.)

Reply
 
Thread Tools Display Modes
  #1  
Old 5th October 2009, 10:16 AM
Chivvyp Chivvyp is offline
Member
 
Join Date: Oct 2007
Location: SW London
Posts: 55
Default DC blocking - hope someone can help

Hi,

I'm still having problems with the DC offset on the output of my old philips cd player.

I'm using 4.7uF soniqs SAX as blocking caps (bypassed with a little 100 nf pp). On the input side of the caps I have 3.5 v dc offset (same as the output from the opamps) , on the output side it's reduced to just under 2v. Checking across the caps I'm reading 5.5uF on a dmm. Both channels are performing the same way.

Does this mean the caps aren't fuctioning properly, I thiught they would either block all the DC or none.

Regards

Pete
Reply With Quote
  #2  
Old 5th October 2009, 12:08 PM
A Stuart A Stuart is offline
Senior Member
 
Join Date: Jan 2006
Location: Selkirk, Scotland
Posts: 251
Default Re: DC blocking - hope someone can help

You would need somewhere (eg a resistor from output end of the cap to earth) for the DC to leak through, in order to charge up the capacitor so that the voltage across the capacitor will oppose the DC.

I reckon (without putting pen to paper) that a 200K resistor would allow the cap to approach equilibrium with a half-life of 1 second.

If you measured with only your DMM there, and your DMM input had a DC resistance of, say, 1M the half life would be 5 seconds; if 10M, 50 seconds.

Alastair
Reply With Quote
  #3  
Old 5th October 2009, 12:26 PM
Chivvyp Chivvyp is offline
Member
 
Join Date: Oct 2007
Location: SW London
Posts: 55
Default Re: DC blocking - hope someone can help

Thanks,

I thought you just needed a cap in the output line to remove the dc, didn't realise it needed a resistor as well.

Does the resistor have to be connected to "real" earth or is the output ground ok (is there any difference)?

Thanks

Pete
Reply With Quote
  #4  
Old 5th October 2009, 01:39 PM
Richard Richard is offline
Super Moderator
 
Join Date: Dec 2005
Location: Notts
Posts: 4,770
Default Re: DC blocking - hope someone can help

Pete,

The reading you get is the voltage formed by current flowing across the resistance of the meter whilst the cap is charging up.

The meter has a high resistance. When you connect it from signal output pin to ground you make a circuit with the DC on the other side of the cap. The cap can now charge through the meter connection and it is the voltage in this circuit that the meter shows. This is not the cap passing DC. It cannot pass DC as its plates are not connected together.

Leave the meter connected and you'll see the voltage drops as the cap charges, it may take a minute or more. Add a resistor across the meter probes as Alistair says and this will happen much quicker as the cap charges quicker to the point that it is fully charged and so no current flows through the meter.

You can understand this more easily with a battery, cap, resistor and meter in a few moments. Charge the cap first by touching it across the battery then put it in circuit with the meter and battery and you'll see the meter reads zero (as no more current flows). Check the cap and it reads the same as the battery. Reverse it in the circuit and you'll get double voltage reading (the 2 are in series). Discharge it with a resistor or wire across it then put it in series with the meter and battery and watch it charge up again mimicking your CDp circ and showing a voltage on the meter as it does so, etc, all good fun

In practice when you connect the CDp to the amp it will have a c. 47k connection to ground anyway.

Rich
Reply With Quote
  #5  
Old 5th October 2009, 02:22 PM
Chivvyp Chivvyp is offline
Member
 
Join Date: Oct 2007
Location: SW London
Posts: 55
Default Re: DC blocking - hope someone can help

So, if I understand correctly, there's not actually any DC in the signal path and I can connect it to my amp without any worries.

Connecting a resistor to earth after the output cap would give me a better reading through the DMM but has no other use?

Pete
Reply With Quote
  #6  
Old 5th October 2009, 03:41 PM
Richard Richard is offline
Super Moderator
 
Join Date: Dec 2005
Location: Notts
Posts: 4,770
Default Re: DC blocking - hope someone can help

Good questions.

The cap won't let any DC through but might cause a DC voltage across a high resistive input whilst it charged at switch-on. Most likely it will be connected to a c. 47k input such as a preamp which will allow the cap to charge near instantly, and allow it to stay fully charged, but there might still be a plop or other sound if the amp was on first.

If it was connected to a 1M input, such as a valve power amp (say the CDp has a volume control and is being used directly into a Leak ST20) maybe there would be an issue at switch-on if the amp was on and warm before the CDp was switched on. Interesting, and I don't know, maybe that's why some CDps have muting transistors before the outputs.

Looking at existing designs,

my Pioneer CDp has 100uF op coupling cap followed by 470k to ground and muting transistors.

WD Phono has 2.2uF followed by 1M to ground.

Rozenblit Pre has two cap coupled outputs, 0.47uF and 2.2uF, and neither have any shunt to ground after them.

You could put 220k after them as Alistair suggests and/or switch the CDp on first.
Reply With Quote
  #7  
Old 5th October 2009, 04:00 PM
Chivvyp Chivvyp is offline
Member
 
Join Date: Oct 2007
Location: SW London
Posts: 55
Default Re: DC blocking - hope someone can help

Thanks.

That's made it a lot clearer.

I've just looked at the schematic for the player and it looks like it uses a 10K res after the output cap in the original path. I'll maybe start with a similar value and work my way up.

Pete
Reply With Quote
  #8  
Old 6th October 2009, 09:26 AM
Richard Richard is offline
Super Moderator
 
Join Date: Dec 2005
Location: Notts
Posts: 4,770
Default Re: DC blocking - hope someone can help

Pete,

We should really ask now, what problems are you having with the offset, any symptoms as such or was it just the meter reading that was a concern? And, what input is the CDp working into?

What was the value of the coupling cap in the original circuit? If you use 10k shunt after the 4.7uF and feed it into 47k input the -3dB point will be 4Hz which is a bit higher than usually set for CD so I'm thinking the original cap was perhaps a higher value. Or it may have been a 10k shunt to ensure the output stage wasn't driving too high a load as some opamps don't like that.

As you can see there's not really a universal or perfect way that all gear can interface as, in this case for example, the cap is part of the coupling and the input resistance is the other part, so it's best to use the known values in each bit to optimise the coupling.

Rich
Reply With Quote
  #9  
Old 6th October 2009, 10:24 AM
Chivvyp Chivvyp is offline
Member
 
Join Date: Oct 2007
Location: SW London
Posts: 55
Default Re: DC blocking - hope someone can help

Hi Rich,

Original caps were 100u.The player is an old Philips CD373 I rebuilt from a couple of broken donor players. I've converted it to non os (removed the 7220 chip) and I've effectively removed all the output circuitry after the first half of the opamps. Really just playing around to see if I like non-os and what else can be done.

I have a few symptoms, slight static at start and end of track, echo of last bars at end of track before muting kicks in but I'm not convinced they're related to the dc offset.

I just didn't want to connect it to my Arcam amp while there seemed to be all this offset or spend much more time tracking down the other problems until I'd got rid of it.

Regards

Pete
Reply With Quote
  #10  
Old 9th October 2009, 04:32 PM
Richard Richard is offline
Super Moderator
 
Join Date: Dec 2005
Location: Notts
Posts: 4,770
Default Re: DC blocking - hope someone can help

Hi Pete,

It doesn't sound like the offset will be causing the problem and the caps will be blocking it ok into the Arcam. That amp has 20k input I think so using another 20k across the op of the CDp will give 10k for the op to work into. I wouldn't go any lower than that and it will also predominate if you use the player into a higher input such as 50k to keep the overall load a lowish value.

I noticed on my circ that they'd gone to quite a bit of trouble to screen the op connections between pcb and rca's. They're braided screens connected to ground with 0.1uF at both ends. The caps are rf ceramics and even they are supported by foam plastic to stop them vibrating I suppose. Sounds overkill I know but will be done for a reason. I wonder if you're picking up RF noise at the start and end? Maybe have a look at your connections from chip to rca's if you've rewired it.

The echo you mention, is it possibly recorded "pre-echo" you can hear, is it audible between tracks, or could it be actual room echo if the volume was high?
Reply With Quote
Reply

Thread Tools
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Forum Jump


All times are GMT +1. The time now is 03:36 PM.


Powered by vBulletin® Version 3.7.3
Copyright ©2000 - 2020, Jelsoft Enterprises Ltd.
Copyright World Designs