Author |
Message |
mangros
Joined: Jan 21, 2011 Posts: 2 Location: UK
|
Posted: Sat Apr 09, 2011 4:50 pm Post subject:
Sample and Hold won't hold |
|
|
I've built the YuSynth EMS and ARP filters and think they are great. But I'm having trouble with the S&H/Noise. I've used the Bridechamber PCB, and lots of it is working: all the noise functions, the S&H clock, internal and external gates, slew.
With the rate turned up fast and connected to other gear, it would appear to be working with the normalled white noise input; the usual random effect occurs.
But when the rate is lowered it's clear something is wrong. The S&H output is always draining very quickly (0.2 secs) to -10.5V. (it's always negative, but does jump around)
THings that might make a difference: I'm using ferrite beads instead of the 10R resistors on the power rails, and a BC549C instead of a BC547C. I'm using a BF245B JFET - from reading some of the threads on here it seems it should be ideally be a BF245C, but surely it would still work? I've checked all the component values, soldered the line/exp jumper (to exp) and tried a different TL074 but it behaves just the same.
Any help would be most appreciated! |
|
Back to top
|
|
|
yusynth
Joined: Nov 24, 2005 Posts: 1314 Location: France
|
Posted: Mon Apr 11, 2011 5:06 am Post subject:
|
|
|
If you have a scope check the output level of the white noise. If you see that there are peaks to more than +5V or les than -5V then your problem may be due to a high nois level. The solution then is to change the noise source transitor to a less noisy one or reduce the value of R34.
If this not the case try changing the capacitor C15 (use a non-polarized type, don't use a tantalum there) to a new one.
Also since you are using a BF245B instead of a BF245B you may have to change the value of R14. _________________ Yves |
|
Back to top
|
|
|
mangros
Joined: Jan 21, 2011 Posts: 2 Location: UK
|
Posted: Mon Apr 11, 2011 2:18 pm Post subject:
|
|
|
It seemed I had a very loud noise signal... but as the transistor was giving a nice white spectrum I swapped the 470K for a 150K and it dropped the level enough - the sample and hold now holds!
It seems fine now - but what would you recommend for R14 if I stick with a BF245B?
Many thanks! |
|
Back to top
|
|
|
yusynth
Joined: Nov 24, 2005 Posts: 1314 Location: France
|
Posted: Mon Apr 11, 2011 11:56 pm Post subject:
|
|
|
mangros wrote: | It seemed I had a very loud noise signal... but as the transistor was giving a nice white spectrum I swapped the 470K for a 150K and it dropped the level enough - the sample and hold now holds!
It seems fine now - but what would you recommend for R14 if I stick with a BF245B?
Many thanks! |
Yes it seems that you were in the quite rare case where the noise signal becomes to strong and the peaks go far abroad the polarisation scheme of the JFET. As a matter of fact , in general the modern transistor are good and it's difficult to find one that provide a sufficient noise level therefore the S&Hcircuit is calculated bearing this in mind. But in the case of a really very noisy transistor the S&H circuit is no longer optimal.
For R14, if it works now don't bother with trying another value. But I guess that one would have to increase it's value (3.3M).
Regards
Yves _________________ Yves |
|
Back to top
|
|
|
|