Author |
Message |
yusynth
Joined: Nov 24, 2005 Posts: 1314 Location: France
|
Posted: Mon Dec 22, 2008 3:49 am Post subject:
|
|
|
OK 220K Then the voltage swing is weird it should stay below 0V and at least go down to -2.4V. Is some kind of voltage fed at the resonance CV input ?
Question : which version are you using the CA3096 version (U5) or the tranny version. _________________ Yves Last edited by yusynth on Mon Dec 22, 2008 9:01 am; edited 1 time in total |
|
Back to top
|
|
|
stijn
Joined: Jul 21, 2008 Posts: 39 Location: utrecht, netherlands
|
Posted: Mon Dec 22, 2008 8:42 am Post subject:
|
|
|
I have the descrete version. (bridechamber kit)
there is no voltage at the cv control for the resonance |
|
Back to top
|
|
|
stijn
Joined: Jul 21, 2008 Posts: 39 Location: utrecht, netherlands
|
Posted: Tue Dec 30, 2008 7:06 am Post subject:
|
|
|
Hi Ives, it turned out that the bridechamber kit has a PN4391 insted of a 2N3819 jfet, wich has a different pin out. So I desolderded it (wich was quite a hasle, are there any tricks fot this besides a solder sucker?) and flipped it but now I have no resonance at all, so I guess it's broken? |
|
Back to top
|
|
|
yusynth
Joined: Nov 24, 2005 Posts: 1314 Location: France
|
Posted: Tue Dec 30, 2008 7:48 am Post subject:
|
|
|
stijn wrote: | Hi Ives, it turned out that the bridechamber kit has a PN4391 insted of a 2N3819 jfet, wich has a different pin out. So I desolderded it (wich was quite a hasle, are there any tricks fot this besides a solder sucker?) and flipped it but now I have no resonance at all, so I guess it's broken? |
I see, you should let know this to Scott Deyo because he provided you with a component that is not pin to pin compatible with the 2N3819...
It may not be broken but have a different pinch voltage than expected. As I explained on my page the characteristics of the pinch voltage of a JFET brand are very dispersed. JFETs are not as reliable in terms of charateristics as can be BJT transistors and picking two JFETs from the same batch may give very different results. Here it's even more different because you have a different brand of JFET...
What you can do is to take a 47K resistor and connect it in parallel with R41, and see if there is an improvement of the resonance control. If you recover the resonance control then try using other values in order to find the one that gives you full range. Otherwise, that is if with the 47K resistor you don't recover the resonance control, then your JFET is dead and replace it (preferably with a 2N3819 ) _________________ Yves |
|
Back to top
|
|
|
chaorta
Joined: Jan 17, 2009 Posts: 10 Location: Chicago
|
Posted: Wed Jan 21, 2009 10:49 am Post subject:
|
|
|
Here's a dumb question from a new DIY'er...i'd like to purchase this and build it. I do have building experience, and can SORT of read a schematic. My question is more mechanical...is the PCB small enough to fit in a Doepfer A100 case and will it run at +/- 12v, and if so, how would i go about making this happen? _________________ Time is an illusion. Lunchtime doubly so. |
|
Back to top
|
|
|
yusynth
Joined: Nov 24, 2005 Posts: 1314 Location: France
|
|
Back to top
|
|
|
chaorta
Joined: Jan 17, 2009 Posts: 10 Location: Chicago
|
Posted: Wed Jan 21, 2009 1:47 pm Post subject:
|
|
|
Well that was easy, thanks _________________ Time is an illusion. Lunchtime doubly so. |
|
Back to top
|
|
|
softfin
Joined: Oct 11, 2006 Posts: 271 Location: Far in the north
Audio files: 1
|
Posted: Sat Feb 07, 2009 10:51 am Post subject:
|
|
|
yusynth wrote: | Hi Chaorta and welcome
Concerning the PCB I can't tell you because I have no Doepfer case to check it. But may be if give the size of the PCB you can make yourself an idea. The PCB is 10 cm per 7.5 cm and has four fixation holes.
Concerning the power supply, yes it will run at +12V/-12V, for this you just need to remove some components : remove R3 (82ohm) and bypass its pads with a wire, remove U2 (78L12/7812) bypass with an insulated wire the left and rigth pads (leave the middle pad unconnected) and that's it (see the diagram below : left before mod, right : after mod, bypass wires shown in red) |
Finished my pair of diode ladders. I put them in my +-12V case, so I did the suggested modifications, but found the -9V was only about -7.25V, so I exchanged R4 with a 120ohm resistor and also replaced R50 with 5k trimmer as a voltage divider. I didn't measure the resistance of the trimmer when I achieved precise -9V, but it is probably around 2k.
These long awaited filters have this excellent spiky but pleasing character. The only downside seems to be CV bleedthrough which is louder than on my other filters. Is there any simple way to reduce it?
Btw, six diodes are reversed on the overlay with the color coded resistors.
Big thanks to Yves for this module! |
|
Back to top
|
|
|
yusynth
Joined: Nov 24, 2005 Posts: 1314 Location: France
|
Posted: Sat Feb 07, 2009 11:00 am Post subject:
|
|
|
softfin wrote: | These long awaited filters have this excellent spiky but pleasing character. The only downside seems to be CV bleedthrough which is louder than on my other filters. Is there any simple way to reduce it?
Btw, six diodes are reversed on the overlay with the color coded resistors.
Big thanks to Yves for this module! |
Did you match thorougly your diodes ? This is important to reduce the CV bleedthrough.
Concerning, the overlay error, this has been corrected on my site at least two months ago _________________ Yves |
|
Back to top
|
|
|
softfin
Joined: Oct 11, 2006 Posts: 271 Location: Far in the north
Audio files: 1
|
Posted: Sat Feb 07, 2009 1:48 pm Post subject:
|
|
|
yusynth wrote: |
Did you match thorougly your diodes ? This is important to reduce the CV bleedthrough.
Concerning, the overlay error, this has been corrected on my site at least two months ago |
I used the layouts I had archived some time ago, so I didn't notice the layout was already corrected.
Regarding the diodes, I took a strip of factory packed diodes and measured them simply by putting +9V across each one and measured the voltage drop. All diodes had almost exactly the same amount of voltage drop, the difference in drop between diodes being 0.01-0.02V. That was also pretty much as precise a measurement as possible on my multimeter.
Should I match them more carefully, and if so, is there a more precise method you could recommend? |
|
Back to top
|
|
|
yusynth
Joined: Nov 24, 2005 Posts: 1314 Location: France
|
Posted: Sat Feb 07, 2009 2:54 pm Post subject:
|
|
|
Yes you should... as a matter of fact they must be matched by pair within 2mV. The bleedthrough mostly comes for the unbalance between each side of the ladder. The principle of this kind of ladder is that the CV is canceled out by subtracting the left and right side of te ladder while the audio passes through because it is differentially transmitted by the input transistors. _________________ Yves |
|
Back to top
|
|
|
softfin
Joined: Oct 11, 2006 Posts: 271 Location: Far in the north
Audio files: 1
|
Posted: Sun Feb 08, 2009 9:06 am Post subject:
|
|
|
yusynth wrote: | Yes you should... as a matter of fact they must be matched by pair within 2mV. The bleedthrough mostly comes for the unbalance between each side of the ladder. The principle of this kind of ladder is that the CV is canceled out by subtracting the left and right side of te ladder while the audio passes through because it is differentially transmitted by the input transistors. |
Thanks for the explanation, it makes perfect sense.
Since I couldn't get enough precision using my 'better' multimeter, I tried matching the diodes again with a lower voltage (1.6V) and using my old, cheapish DVM which actually allowed me to measure the voltage drop more precisely in millivolts.
So I matched diodes to 1mv and replaced all diodes in the ladders of both of my units. It seems cv bleedthru decreased considerably, but it is still audible.
I suppose I should get a better quality multimeter, if I want to have more precise and stable readings. |
|
Back to top
|
|
|
yusynth
Joined: Nov 24, 2005 Posts: 1314 Location: France
|
Posted: Sun Feb 08, 2009 9:59 am Post subject:
|
|
|
softfin wrote: |
So I matched diodes to 1mv and replaced all diodes in the ladders of both of my units. It seems cv bleedthru decreased considerably, but it is still audible.
I suppose I should get a better quality multimeter, if I want to have more precise and stable readings. |
Well as a matter of fact it would be difficult to get better with the matching of diodes. Now the next points to take care of are :
1°) try using different CA3046 (the matching of the differential transistor pair may vary from a chip to another : by the way avoid Thailand labelled brand they are very bad)
2°) if you are using the discrete version check the match of the BC557 pair and of BC547 pair, if you are using the CA3096 try different ICs. _________________ Yves |
|
Back to top
|
|
|
softfin
Joined: Oct 11, 2006 Posts: 271 Location: Far in the north
Audio files: 1
|
|
Back to top
|
|
|
yusynth
Joined: Nov 24, 2005 Posts: 1314 Location: France
|
Posted: Sun Feb 08, 2009 4:04 pm Post subject:
|
|
|
from your sample I'd say that you won't get a better result. As a matter of fact you could add a 100pF to 1nf cap in parallel with R22 to create just enough slew to attenuate the bump on fast attacks (this trick was used in the Minimoog filter). _________________ Yves |
|
Back to top
|
|
|
softfin
Joined: Oct 11, 2006 Posts: 271 Location: Far in the north
Audio files: 1
|
Posted: Sun Feb 08, 2009 4:37 pm Post subject:
|
|
|
yusynth wrote: | from your sample I'd say that you won't get a better result. As a matter of fact you could add a 100pF to 1nf cap in parallel with R22 to create just enough slew to attenuate the bump on fast attacks (this trick was used in the Minimoog filter). |
Ok. I'll add the cap and perhaps a switch too, so I'll have the possibility to add just a bit of thump/click to the sound when I need it.
Thanks for your help. |
|
Back to top
|
|
|
numbertalk
Joined: May 05, 2008 Posts: 992 Location: Austin, TX
Audio files: 5
|
Posted: Mon Feb 09, 2009 8:54 am Post subject:
|
|
|
What a great sounding filter! This has been in my build queue for a while but think I need to bump up its priority a bit |
|
Back to top
|
|
|
stijn
Joined: Jul 21, 2008 Posts: 39 Location: utrecht, netherlands
|
Posted: Wed Apr 08, 2009 1:32 am Post subject:
|
|
|
ok Yves, I finally got the right jfet , I put it in but in all my exitement
hooked up the power in the wrong way
so something got fried. can u tell me wich parts I should replace to make it work again? |
|
Back to top
|
|
|
yusynth
Joined: Nov 24, 2005 Posts: 1314 Location: France
|
Posted: Wed Apr 08, 2009 10:22 am Post subject:
|
|
|
stijn wrote: | ok Yves, I finally got the right jfet , I put it in but in all my exitement
hooked up the power in the wrong way
so something got fried. can u tell me wich parts I should replace to make it work again? |
Ouch... This is a great classic and it also happened to me a couple of time...
The following ICs and trannies may have fried (ordered from most certain death to may have survived) :
the 7812/78L12 regulator has probably been fried
the BC327 is also probably dead
the other ICs (TL074) and trannies have probably survived because the regulator died first, but I would replace the TL074 which doesn't like reverse power !
Did you see a fume ? _________________ Yves |
|
Back to top
|
|
|
stijn
Joined: Jul 21, 2008 Posts: 39 Location: utrecht, netherlands
|
Posted: Thu Apr 09, 2009 12:04 am Post subject:
|
|
|
tnx for the fast reply Yves.
ooh and yes there was a fume! |
|
Back to top
|
|
|
numbertalk
Joined: May 05, 2008 Posts: 992 Location: Austin, TX
Audio files: 5
|
Posted: Thu Oct 08, 2009 7:30 pm Post subject:
|
|
|
yusynth wrote: |
Here is a tip : for those of you who want to obtain a genuine EMS sound while using DOTCOM or YUSYNTH VCOs it is recommended to change the value of R8 to 27K. The output levels of these VCOs are much higher than those of an AKS and this makes a big difference in sound. |
Hi Yves,
I notice on the schematic you also flag R9 as being changed in addition to R8. What value would you suggest changing R9 to? I don't have dotcom or yusynth vcos but my vcos are 10Vp-p, so I'm guessing I'd want to change these 2 resistors as well.
Thanks! |
|
Back to top
|
|
|
yusynth
Joined: Nov 24, 2005 Posts: 1314 Location: France
|
Posted: Thu Oct 08, 2009 11:18 pm Post subject:
|
|
|
numbertalk wrote: | yusynth wrote: |
Here is a tip : for those of you who want to obtain a genuine EMS sound while using DOTCOM or YUSYNTH VCOs it is recommended to change the value of R8 to 27K. The output levels of these VCOs are much higher than those of an AKS and this makes a big difference in sound. |
Hi Yves,
I notice on the schematic you also flag R9 as being changed in addition to R8. What value would you suggest changing R9 to? I don't have dotcom or yusynth vcos but my vcos are 10Vp-p, so I'm guessing I'd want to change these 2 resistors as well.
Thanks! |
Hi if you have 10Vpp VCOs then it's the same as for YUSYNTH or DOTCOM VCOS, use 27K for R8 and leave R9 unchanged.
What is shown on the schematic is that one may modify either R8 or R9, or both to achieve the desired attenuation of the signal.
In your case you only need to change R8 as mentioned in the quote above. _________________ Yves |
|
Back to top
|
|
|
ndkent
Joined: Jan 03, 2006 Posts: 66 Location: new york
|
Posted: Fri Oct 09, 2009 2:20 am Post subject:
|
|
|
yusynth wrote: |
Here is a tip : for those of you who want to obtain a genuine EMS sound while using DOTCOM or YUSYNTH VCOs it is recommended to change the value of R8 to 27K. The output levels of these VCOs are much higher than those of an AKS and this makes a big difference in sound. |
I could see doing that if for practical purposes all your VCOs had every waveform ranged +-10v, but considering the original EMS has level pots out of the VCO I'd think an variable attenuator would be more ideal than adding more fixed attenuation. |
|
Back to top
|
|
|
yusynth
Joined: Nov 24, 2005 Posts: 1314 Location: France
|
|
Back to top
|
|
|
Scot Solida
Joined: Oct 24, 2009 Posts: 100 Location: Hutchinson Kansas
|
Posted: Thu Oct 29, 2009 6:34 pm Post subject:
|
|
|
I just finished building this one tonight. At first, it didn't work properly, but using the 220 resistor as described above did the trick. It sounds terrific! I am using it in my Dotcom system, but I left R8 alone. for now. The input control knob should be enough for my needs. It won't do any harm will it?
In any event, I still need to mount the PCB to the panel. After that, I will do some A/B comparisons with my VCS3. I'll post the results in another thread. |
|
Back to top
|
|
|
|