MOTM

Hi all,

Just had a question about the 960: are the bypass caps that are there  
essentially a just-in-case precaution?

Bypass caps, from what I understand, are there to reduce noise. And  
correct me if I'm wrong, but most of us that are using the 960 would  
probably be also using a power supply with its own bypass caps.

Do the caps then only apply to certain of the power connectors, like  
the one on the side of the caps? Or is it the case that if you  
connect from a PSU to one of the .156 connectors - or any other spot,  
for that matter - that it would pass the caps?

You'll have to forgive me if there's a schem out of there somewhere  
that would explain this - I didn't see one.

thanks,
Matthew

a) technically, they are not 'bypass' caps, they are 'bulk' caps. A 'bypass' 
cap is a small value cap (like 0.01uf) that is very close to an IC's power 
pin. It is there to 'bypass' RF frequencies to ground cause by the lead 
inductance of the part. The caps used in the '960 are there to reduce 
voltage drop when wiring large cabinets, where the '960 is say 20-30" away 
from the main supply (or if you are wiring say 2-4 of them up in a big 
cabinet.

b) the caps are across the +-15V lines, so every connector "sees" the caps.

Both bypass caps and bulk caps are needed in 'real systems' to counter 
effects of the 'real world'.

Paul S.

Just theorizing here: could a faulty cap across one of the lines  
cause swings in voltage?

Recently I shuffled some units around, pulled the 960 out of the  
system and put it back. I've noticed that now the + voltage takes an  
eternity to settle after powering up. It will swing up to 15.6v and  
then drift randomly between that and 14.8v for maybe 15 minutes  
before finally settling on 14.9, while the - voltage heads right to  
-14.9 and stays there, rock-solid. I've had a meter hooked up for the  
last few days trying to figure out the pattern.

I'm inclined to say that the problem isn't due to the PSU, which is a  
dotcom NOS by Power One, because my dotcom VCO doesn't drift in pitch  
at all but the two MOTM vcos do, together, and along with the +  
voltage. The 960 is the only thing between the PSU and the MOTM  
units, so it's the only thing that appears wonky.

Matthew

On 23-Nov-09, at 8:19 PM, Paul Schreiber wrote:

> a) technically, they are not 'bypass' caps, they are 'bulk' caps. A  
> 'bypass' cap is a small value cap (like 0.01uf) that is very close  
> to an IC's power pin. It is there to 'bypass' RF frequencies to  
> ground cause by the lead inductance of the part. The caps used in  
> the '960 are there to reduce voltage drop when wiring large  
> cabinets, where the '960 is say 20-30" away from the main supply  
> (or if you are wiring say 2-4 of them up in a big cabinet.
>
> b) the caps are across the +-15V lines, so every connector "sees"  
> the caps.
>
> Both bypass caps and bulk caps are needed in 'real systems' to  
> counter effects of the 'real world'.
>
> Paul S.
>

__________________________________________________
site: http://www.bodega-audio.com
label myspace: http://www.myspace.com/bodegaaudio
dj/prod. myspace: http://www.myspace.com/fateastcoast
podcast: http://www.bodega-audio.com/podcast/podcast.xml

Recently I shuffled some units around, pulled the 960 out of the system and 
put it back. I've noticed that now the + voltage takes an eternity to settle 
after powering up. It will swing up to 15.6v and then drift randomly between 
that and 14.8v for maybe 15 minutes before finally settling on 14.9, while 
the - voltage heads right to -14.9 and stays there, rock-solid. I've had a 
meter hooked up for the last few days trying to figure out the pattern.

I'm inclined to say that the problem isn't due to the PSU, which is a dotcom 
NOS by Power One, because my dotcom VCO doesn't drift in pitch at all but 
the two MOTM vcos do, together, and along with the + voltage. The 960 is the 
only thing between the PSU and the MOTM units, so it's the only thing that 
appears wonky.

-------------------------------------

This is very odd. It sounds like the Power One is unstable, the LM723N on 
the Power One is 'wandering'. They have a built-in Zener diode reference, 
maybe it is drifting around with temperature. You can try disconnecting the 
'960 from the Power One, and see if it still wanders around.

If it still does, you might get lucky and if it's in a socket you can get 
new ones from Mouser/Digikey for <$1.

If if does *not*, then:
a) are you exceeding 80% of the rated current? You might have an in-rush 
power issue. The +15V current draw is probably 2X the -15V draw. Use the 
meter to check the *current* on each supply line.

b) a new LM723N might still be the fix. If both of the parts are in sockets, 
swap them and see if that fixes it (there may even be 3 on there if there is 
also +5V).

c) you can see if it really is the '960 by keeping it connected to the Power 
One, but unplugging all the modules to it.

Paul S.

When I power it up and it reaches a stable state it tends to stay  
stable for a long time after I power it off again - maybe an hour.  
Since I'd been using it last night for long enough for it to settle  
down I decided to hold off until this morning, to let everything  
return to it's "drifty" state.

I disconnected the 960 and measured the voltages and, sure enough,  
the negative is solid but the positive drifts, from +15.5 slowly down  
to +14.9. This is with about 150ma worth of modules connected on a  
PSU rated for at least 1000ma.

I was thinking that it had to be the 960 because the pitch of my  
dotcom vco - connected directly to the PSU - didn't drift in pitch  
while my MOTM vcos - connected to the 960 - did. But now I remember  
that the dotcom gets +5v, so perhaps that is stable and that's what's  
keeping it tracking correctly.

In the next few days I'll try and extract the PSU and have a good  
look at it, see what's socketed.

Matthew

On 23-Nov-09, at 10:38 PM, Paul Schreiber wrote:

>
> Recently I shuffled some units around, pulled the 960 out of the  
> system and
> put it back. I've noticed that now the + voltage takes an eternity  
> to settle
> after powering up. It will swing up to 15.6v and then drift  
> randomly between
> that and 14.8v for maybe 15 minutes before finally settling on  
> 14.9, while
> the - voltage heads right to -14.9 and stays there, rock-solid.  
> I've had a
> meter hooked up for the last few days trying to figure out the  
> pattern.
>
> I'm inclined to say that the problem isn't due to the PSU, which is  
> a dotcom
> NOS by Power One, because my dotcom VCO doesn't drift in pitch at  
> all but
> the two MOTM vcos do, together, and along with the + voltage. The  
> 960 is the
> only thing between the PSU and the MOTM units, so it's the only  
> thing that
> appears wonky.
>
> -------------------------------------
>
> This is very odd. It sounds like the Power One is unstable, the  
> LM723N on
> the Power One is 'wandering'. They have a built-in Zener diode  
> reference,
> maybe it is drifting around with temperature. You can try  
> disconnecting the
> '960 from the Power One, and see if it still wanders around.
>
> If it still does, you might get lucky and if it's in a socket you  
> can get
> new ones from Mouser/Digikey for <$1.
>
> If if does *not*, then:
> a) are you exceeding 80% of the rated current? You might have an in- 
> rush
> power issue. The +15V current draw is probably 2X the -15V draw.  
> Use the
> meter to check the *current* on each supply line.
>
> b) a new LM723N might still be the fix. If both of the parts are in  
> sockets,
> swap them and see if that fixes it (there may even be 3 on there if  
> there is
> also +5V).
>
> c) you can see if it really is the '960 by keeping it connected to  
> the Power
> One, but unplugging all the modules to it.
>
> Paul S.
>
>
>
>
> ------------------------------------
>
> Yahoo! Groups Links
>
>
>

__________________________________________________
site: http://www.bodega-audio.com
label myspace: http://www.myspace.com/bodegaaudio
dj/prod. myspace: http://www.myspace.com/fateastcoast
podcast: http://www.bodega-audio.com/podcast/podcast.xml