BOZ - DC at the output at turn on

[bugg]

Well, I finished my BOZ pre last night and tested it out one channel at a time. I'm getting the perscribed 40v across R104 and everything sounds good as long as I follow the turn on procedure (source -> pre -> amp)

But if I stray from that, or turn the amp on too soon, I see about 20v DC at the output of the amp, which is a LM3875 chipamp. Of course this pushes the speaker's cone out all the way (and I'm sure has damaged it) but that's what test speakers are for

Further investigation proves that it happens when the BOZ gets turned off too. so anytime the amp is either ramping up, or down (power on/power off) it shoves a bunch of DC noise to the amp for about 5 seconds.

The only change I made to the circuit was to remove the output pot (P101) and use a 100k pot in place of R107, with the wiper going to the 1uf C104.

Questions? Comments? Suggestions?

TIA,
Michael.

 

[Netlist]

I guess the purpose of P101 is, apart from acting as a volume control, to discharge C103. Try putting a 47k or so from output to ground and see what happens.

/Hugo

 

[CheffDeGaar]

Turning the BOZ on applies 20V on C103. Although it is reffered to as a DC bias point, reaching that point implies a AC behaviour when charging the cap. This 0-to-20V transient signal (much more greater than a normal transient musical signal at a preamp level) is seen as a big LF signal by the following part of the system. The charge accumulated in C103 is evacuated in the input impedance/resistance of the following amp. The smaller the resistance, the shorter the transient.

Hugo's advice on the additionnal resistance is well worth following, but I would add a turn on delay (and relay) for the output...

 

[bugg]

I tried a 100k and 15k to ground (didn't have 47k lying around at the time) with no real change in behaviour.

Is this problem common/known with the BOZ? I'm a newbie so I didn't just "see" this from looking at the circuit. I can look into a delay, but would like to know if this is typical before I try to fix a problem that shouldn't be there.

Thanks,
~michael.

 

[Netlist]

What cap did you put in place of C103?

/Hugo

 

[CheffDeGaar]

The "problem" is real, and I would dare to say it's a normal behaviour. Just take a 10µF cap, apply 20V to it on one side, and hook up a DMM to the other side, with a 100K resistor between "output" and gnd. And just watch the voltage.... And the same goes when you remove the 20V from the cap... Moreover, the problem will be more present if your amp is DC coupled.
And dumb question, but has your amp an input resistor between "In" and gnd ?

 

[bugg]

Hugo,
The Cap is nothing special, just a 10uf Illinois Capacitor polyester cap. I built this thing with pretty much the cheapest parts possible.

Cheff,
The amp has a 22k from in to ground.

Since this sounds like an expected result from the BOZ I'll search around for a delay circuit that'll suite my needs. (suggestions welcome )

Thanks to both of you for your help, and for anyone with a BOZ that isn't seeing this... let me know how ya did it

~michael.

 

[GRollins]

Must be the phase of the moon...wasn't there just a thread about this same thing?
The charge that accumulates on the backside of the DC blocking cap will take time to go away. That time will be shorter with lower resistance to ground, but the only way to make it go away more or less instantly is to ground the output side of the cap for a few seconds after turn-on. It may start at something approximating 20V, but bear in mind that even if you discharge it down to, say, 2V, that's still enough to send most amps to full power. Your speakers will behave accordingly.
You can either give it time to discharge or you can short it to ground through a relay or a manual switch that you can use as a mute if the phone rings.
Reducing the resistor on the output side of the cap too much will eventually load down the output of the preamp. Dropping the value a little bit won't hurt anything, though.

Grey

 

[CheffDeGaar]

For a little mute/delay circuit, you can use the Pass' one
Look at http://www.passlabs.com/downloads/old product/al_serv_man.pdf (Page 3), the circuit built around Q2 and Q3. Should be a 24V relay, but has to be confirmed. This circuit senses the voltage drop between unreg and regulated supplies, slowly unmutes at turn on (charging C5 takes time to switch Q2 on) , and quickly mutes at turn off. Should suit your needs.

Not exactly the same, but discussed here : http://www.diyaudio.com/forums/showthread.php?s=&threadid=15106

Hope this helps...

 

[wuffwaff]

Hi,

there are also two simple sollutions to this problem:

1. turn it on and leave it on
2. use the balanced output, this way the differential voltage won´t change and the turn on thump will be almost unnoticable.

If you really want to do something I would go for Grey´s suggestion with a mute switch or relay that shorts the output to ground.


William

 

[bugg]

thank you guys very much for your help... I'll let you know how it goes, I may just decide to follow "the procedure" for a while... maybe I'll get used to it. My - soon to be - former pre was passive so it was never really an issue.

Sorry if this has been posted recently, I didn't find it (or didn't understand what I had found more likely) but I'll search harder next time

~michael

 

[bugg]

I couldn't wait any longer, just had to hear this thing so I hooked it up this morning... very nice. I think Nelson said it best in the BOZ article "It sounds marvelous."

A big part of my decision to build this over other preamps was that I felt compelled to take advantage of Nelson's generosity in giving these designs to the DIY community.

thanks,
~michael.