Author |
Message |
frijitz
Joined: May 04, 2007 Posts: 1734 Location: NM USA
Audio files: 54
|
Posted: Thu Oct 06, 2011 5:29 am Post subject:
|
|
|
moellhoven wrote: | There is one thing i am not sure of i got it right. The kit came with an SSM2220 but without an TempCo.
From the documentation i figured it is missing and i added one of my spares to R14b. Is this OK or is this maybe the problem? |
R14a and R14b are alternative physical locations for the (single) tempco resistor. So as long as you didn't use both locations you should be OK (provided you used the correct resistance value (2k)).
Ian |
|
Back to top
|
|
|
moellhoven
Joined: Jun 18, 2011 Posts: 15 Location: Germany
|
Posted: Thu Oct 06, 2011 1:59 pm Post subject:
|
|
|
Than i got it right. Will search for the bug on saturday. |
|
Back to top
|
|
|
moellhoven
Joined: Jun 18, 2011 Posts: 15 Location: Germany
|
Posted: Sat Oct 08, 2011 4:41 am Post subject:
|
|
|
frijitz wrote: | I can't say much beyond that. If you want to measure the voltage at the A1a output (pin 1) as the coarse pot is varied from end to end, that might be helpful.
|
The voltage goes from 0,43V to -0.08V.
I've inspected the board and solder joints but i could not find a problem. |
|
Back to top
|
|
|
frijitz
Joined: May 04, 2007 Posts: 1734 Location: NM USA
Audio files: 54
|
Posted: Sat Oct 08, 2011 11:45 am Post subject:
|
|
|
moellhoven wrote: | frijitz wrote: | I can't say much beyond that. If you want to measure the voltage at the A1a output (pin 1) as the coarse pot is varied from end to end, that might be helpful.
|
The voltage goes from 0,43V to -0.08V. |
Perfect.
Next please look at the following as you vary the coarse pot:
1.) The voltage at the A1b output (pin 7)
2.) The voltage at the Q1 collector / R23 node
Ian |
|
Back to top
|
|
|
moellhoven
Joined: Jun 18, 2011 Posts: 15 Location: Germany
|
Posted: Sun Oct 09, 2011 2:16 am Post subject:
|
|
|
frijitz wrote: | moellhoven wrote: | frijitz wrote: | I can't say much beyond that. If you want to measure the voltage at the A1a output (pin 1) as the coarse pot is varied from end to end, that might be helpful.
|
The voltage goes from 0,43V to -0.08V. |
Perfect.
Next please look at the following as you vary the coarse pot:
1.) The voltage at the A1b output (pin 7) |
The voltage goes from 0,733V to -4,43V.
frijitz wrote: |
2.) The voltage at the Q1 collector / R23 node |
The voltage goes from -14,17V to -10,36V. |
|
Back to top
|
|
|
frijitz
Joined: May 04, 2007 Posts: 1734 Location: NM USA
Audio files: 54
|
Posted: Sun Oct 09, 2011 5:02 am Post subject:
|
|
|
moellhoven wrote: | frijitz wrote: | 1.) The voltage at the A1b output (pin 7) |
The voltage goes from 0,733V to -4,43V. |
I hope that's a typo. You meant +4.43V?
Quote: | frijitz wrote: | 2.) The voltage at the Q1 collector / R23 node |
The voltage goes from -14,17V to -10,36V. |
Those both look about right. Since the problem is at the low current end, it's a bit hard to tell exactly.
Can you connect a DVM directly across R23? This will be floating, so be very careful nothing gets connected to ground. You should see this voltage go to 0mV at the low frequency end, if the converter is working properly.
Thanks for sticking with this. It's hard troubleshooting long distance, but we should be able to figure it out eventually!
Ian |
|
Back to top
|
|
|
moellhoven
Joined: Jun 18, 2011 Posts: 15 Location: Germany
|
Posted: Sun Oct 09, 2011 5:25 am Post subject:
|
|
|
frijitz wrote: | I hope that's a typo. You meant +4.43V? |
Of course you are right.
frijitz wrote: | Can you connect a DVM directly across R23? This will be floating, so be very careful nothing gets connected to ground. You should see this voltage go to 0mV at the low frequency end, if the converter is working properly. |
Yes, it goes down to 0mV.
frijitz wrote: | Thanks for sticking with this. It's hard troubleshooting long distance, but we should be able to figure it out eventually! |
Thank you for sticking with me. Your help is highly appreciated! |
|
Back to top
|
|
|
frijitz
Joined: May 04, 2007 Posts: 1734 Location: NM USA
Audio files: 54
|
Posted: Sun Oct 09, 2011 11:09 am Post subject:
|
|
|
moellhoven wrote: | Yes, it goes down to 0mV. |
Hmmm... This is very unusual. Maybe I misunderstood the symptoms.
I believe you said that with "coarse" all the way down, the 100Hz signal would still go through zero as "init" is varied. This is hard to understand -- if the expo source is putting zero current into the OTA then the OTA should output zero current independent of the"init" setting.
If this is really what you are seeing, then the problem must be a bad CA3280. But I've never seen this behavior before. (There's always a first time, of course. )
So could you please double check this for me before we go ahead? Thanks!
Ian |
|
Back to top
|
|
|
moellhoven
Joined: Jun 18, 2011 Posts: 15 Location: Germany
|
Posted: Sun Oct 09, 2011 12:18 pm Post subject:
|
|
|
frijitz wrote: | moellhoven wrote: | Yes, it goes down to 0mV. |
Hmmm... This is very unusual. Maybe I misunderstood the symptoms.
I believe you said that with "coarse" all the way down, the 100Hz signal would still go through zero as "init" is varied. This is hard to understand -- if the expo source is putting zero current into the OTA then the OTA should output zero current independent of the"init" setting.
If this is really what you are seeing, then the problem must be a bad CA3280. But I've never seen this behavior before. (There's always a first time, of course. )
So could you please double check this for me before we go ahead? |
This is what's happening here. The Init pot always affects the frequency regardless of the state of the coarse knob. Maybe i wired the Init pot wrong? It is connected to +5V Init -5V. |
|
Back to top
|
|
|
frijitz
Joined: May 04, 2007 Posts: 1734 Location: NM USA
Audio files: 54
|
Posted: Sun Oct 09, 2011 2:46 pm Post subject:
|
|
|
moellhoven wrote: | This is what's happening here. The Init pot always affects the frequency regardless of the state of the coarse knob. |
This means that U1a is putting out a bit of current, even when it has no current going into it via R23, ie when it should be cut off. This could be due to a bad U1 or to some leakage current into U1 pin 3.
At this point we'll need to do a bit of minor surgery. Please unsolder the end of R23 that connects to Q1. This is the side nearest the edge of the board. Use a length of insulated wire to connect this end of R23 to the -15V supply. This should ensure that U1a is turned off good and hard, independent of the expo converter. The freq should now be very low and "init" should have no effect.
If you still have a frequency over a few Hz, then make sure the board is clean around U1 pin 3 and the trace going to R23. A bit of leakage from U1 pin 4 to pin 3 could easily produce the effect you are seeing.
Ian |
|
Back to top
|
|
|
moellhoven
Joined: Jun 18, 2011 Posts: 15 Location: Germany
|
Posted: Mon Oct 10, 2011 12:01 pm Post subject:
|
|
|
frijitz wrote: | This means that U1a is putting out a bit of current, even when it has no current going into it via R23, ie when it should be cut off. This could be due to a bad U1 or to some leakage current into U1 pin 3.
At this point we'll need to do a bit of minor surgery. Please unsolder the end of R23 that connects to Q1. This is the side nearest the edge of the board. Use a length of insulated wire to connect this end of R23 to the -15V supply. This should ensure that U1a is turned off good and hard, independent of the expo converter. The freq should now be very low and "init" should have no effect.
|
OK, the frequency is now really low around 0.5Hz. The Coarse and Init knob have no effect. |
|
Back to top
|
|
|
frijitz
Joined: May 04, 2007 Posts: 1734 Location: NM USA
Audio files: 54
|
Posted: Mon Oct 10, 2011 2:13 pm Post subject:
|
|
|
moellhoven wrote: | OK, the frequency is now really low around 0.5Hz. The Coarse and Init knob have no effect. |
That's good news! It means that the OTA (U1a) is working OK as is the oscillator core.
Unfortunately it leaves us in a bit of a quandry, since we've tested everything separately and they all seem to be working.
Let's have a more sensitive look at the expo current source. Please solder a 100k resistor to the pad where you disconnected R23. (Leave R23 unconnected at that end). Carefully connect the other end of the 100k to the -15V supply (just use the same wire you had connected to R23). Then measure the voltage across the 100k at the low end of the "coarse" range. You will need a pretty sensitive meter for this -- it should measure down to below 1 mV.
Ian |
|
Back to top
|
|
|
moellhoven
Joined: Jun 18, 2011 Posts: 15 Location: Germany
|
Posted: Mon Oct 10, 2011 3:23 pm Post subject:
|
|
|
frijitz wrote: | That's good news! It means that the OTA (U1a) is working OK as is the oscillator core. |
Great
frijitz wrote: |
Let's have a more sensitive look at the expo current source. Please solder a 100k resistor to the pad where you disconnected R23. (Leave R23 unconnected at that end). Carefully connect the other end of the 100k to the -15V supply (just use the same wire you had connected to R23). Then measure the voltage across the 100k at the low end of the "coarse" range. You will need a pretty sensitive meter for this -- it should measure down to below 1 mV. |
Hmm...now i get -14,66V. I measured at the pad where the 100k is connected. Is this the correct spot or did i misunderstand? |
|
Back to top
|
|
|
frijitz
Joined: May 04, 2007 Posts: 1734 Location: NM USA
Audio files: 54
|
Posted: Mon Oct 10, 2011 4:13 pm Post subject:
|
|
|
moellhoven wrote: | Hmm...now i get -14,66V. I measured at the pad where the 100k is connected. Is this the correct spot or did i misunderstand? |
You need to read across the resistor -- one probe on each end.
Ian |
|
Back to top
|
|
|
moellhoven
Joined: Jun 18, 2011 Posts: 15 Location: Germany
|
Posted: Mon Oct 10, 2011 11:51 pm Post subject:
|
|
|
frijitz wrote: |
Let's have a more sensitive look at the expo current source. Please solder a 100k resistor to the pad where you disconnected R23. (Leave R23 unconnected at that end). Carefully connect the other end of the 100k to the -15V supply (just use the same wire you had connected to R23). Then measure the voltage across the 100k at the low end of the "coarse" range. You will need a pretty sensitive meter for this -- it should measure down to below 1 mV. |
OK, it measures 0.4V across the resistor. |
|
Back to top
|
|
|
frijitz
Joined: May 04, 2007 Posts: 1734 Location: NM USA
Audio files: 54
|
Posted: Tue Oct 11, 2011 6:14 am Post subject:
|
|
|
moellhoven wrote: | frijitz wrote: |
Let's have a more sensitive look at the expo current source. Please solder a 100k resistor to the pad where you disconnected R23. (Leave R23 unconnected at that end). Carefully connect the other end of the 100k to the -15V supply (just use the same wire you had connected to R23). Then measure the voltage across the 100k at the low end of the "coarse" range. You will need a pretty sensitive meter for this -- it should measure down to below 1 mV. |
OK, it measures 0.4V across the resistor. |
Hmmm... I don't understand why you would have gotten 0mV across R23 before, when it was in place.
Anyway, you are seeing 4uA of current from the converter when you should see about a pA, ie about a million times less.
My best guess is that the converter transistor is damaged so that it has a large leakage current, without being totally burned out. This can happen fairly easily if it gets too hot during soldering.
So I suggest removing Q1/Q2 and temporarily replacing them with general-purpose pnp units (2N3906, etc.) This should get you the full frequency range back (I hope ).
Let us know if that works!
Ian |
|
Back to top
|
|
|
moellhoven
Joined: Jun 18, 2011 Posts: 15 Location: Germany
|
Posted: Tue Oct 11, 2011 1:29 pm Post subject:
|
|
|
frijitz wrote: | Hmmm... I don't understand why you would have gotten 0mV across R23 before, when it was in place.
Anyway, you are seeing 4uA of current from the converter when you should see about a pA, ie about a million times less.
My best guess is that the converter transistor is damaged so that it has a large leakage current, without being totally burned out. This can happen fairly easily if it gets too hot during soldering.
So I suggest removing Q1/Q2 and temporarily replacing them with general-purpose pnp units (2N3906, etc.) This should get you the full frequency range back (I hope ).
Let us know if that works! |
I'll let you know Unfortunately this probably needs to wait until next monday. Maybe i can quickly check it out in the morning I used a socket for the SSM2220 so it should not been damaged from soldering.
Thanks alot for your help! It is amazing to get this kind of support from the man itself!
BTW: Just before the Teezer i build a Threeler that worked beautifully right after first startup. Can't wait to get the panels and put them to use. |
|
Back to top
|
|
|
frijitz
Joined: May 04, 2007 Posts: 1734 Location: NM USA
Audio files: 54
|
Posted: Tue Oct 11, 2011 5:29 pm Post subject:
|
|
|
moellhoven wrote: | I'll let you know Unfortunately this probably needs to wait until next monday. Maybe i can quickly check it out in the morning I used a socket for the SSM2220 so it should not been damaged from soldering. |
No problem with next Monday -- I'm not going anywhere.
I like to see folks using sockets. I assumed you didn't, since you were using a bridechamber kit.
If the transistors are not damaged then there may be leakage from surface crud. Look especially at the pads for other transistor position -- the clearance there is quite small, and I don't remember whether there is solder mask between them.
I guess you could just look at what frequency you get without the transistors.
Ian |
|
Back to top
|
|
|
moellhoven
Joined: Jun 18, 2011 Posts: 15 Location: Germany
|
Posted: Wed Oct 19, 2011 12:23 am Post subject:
|
|
|
frijitz wrote: | So I suggest removing Q1/Q2 and temporarily replacing them with general-purpose pnp units (2N3906, etc.) This should get you the full frequency range back (I hope ). |
I am back from my short trip and still eager to find this bug
When i replace the SSM with two 2N3906 the behaviour is pretty much the same as before.
Without the transistors the Coarse knob has almost no effect. The Init knob controls the range then. It goes to zero in the middle and to about 220 Hz in the extreme positions. |
|
Back to top
|
|
|
frijitz
Joined: May 04, 2007 Posts: 1734 Location: NM USA
Audio files: 54
|
Posted: Thu Oct 20, 2011 11:49 am Post subject:
|
|
|
moellhoven wrote: | Without the transistors the Coarse knob has almost no effect. The Init knob controls the range then. It goes to zero in the middle and to about 220 Hz in the extreme positions. |
So without the transistors the OTA is still being fed current! We've looked at this carefully enough to see that all the components and subsystems seem to be operating properly. All that's left that I can think of is that your board is dirty.
You need to clean carefully all around the Q1/Q2 pads. Especially the five pads for the inline pair, marked a798. Remember, to clean, do NOT flood the area with solvent. This will just spread the gunk around. Use a Q-tip that is moistened but not saturated and scrub gently at the suspected problem areas. This will disolve the gunk and pick it up without spreading it around. It's difficult to clean under a socket, so if cleaning around it doesn't work, you may have to remove it.
Ian |
|
Back to top
|
|
|
moellhoven
Joined: Jun 18, 2011 Posts: 15 Location: Germany
|
Posted: Thu Oct 27, 2011 12:44 am Post subject:
|
|
|
frijitz wrote: | All that's left that I can think of is that your board is dirty.
You need to clean carefully all around the Q1/Q2 pads. Especially the five pads for the inline pair, marked a798. Remember, to clean, do NOT flood the area with solvent. This will just spread the gunk around. Use a Q-tip that is moistened but not saturated and scrub gently at the suspected problem areas. This will disolve the gunk and pick it up without spreading it around. It's difficult to clean under a socket, so if cleaning around it doesn't work, you may have to remove it. |
Just a quick update. After a first quick cleaning session the frequency range improved to (16 Hz - 30 kHz). On the weekend i will take a bit more time cleaning the whole board. Thank you very much for all the help!
Another lesson learned |
|
Back to top
|
|
|
frijitz
Joined: May 04, 2007 Posts: 1734 Location: NM USA
Audio files: 54
|
Posted: Thu Oct 27, 2011 3:32 am Post subject:
|
|
|
moellhoven wrote: | Just a quick update. After a first quick cleaning session the frequency range improved to (16 Hz - 30 kHz). |
Whew, sounds like we finally got it! Hopefully you can get it working all the way down in frequency ... it may take some patience, but just keep at it.
Ian |
|
Back to top
|
|
|
ATOM
Joined: Sep 20, 2004 Posts: 48 Location: Norway
|
Posted: Sat Nov 05, 2011 2:38 am Post subject:
|
|
|
WOW. Got a new sound palette.
Got the Teezer up and running behind a muck-up panel,
a ten-turn pot for Init was really helpful for dialing in some of that magic.
Still unsure if I go for a ten-turn pot for tune or the course/fine combo.
Tried a switch for sync but I was missing some mojo.
Thank you Fritz for this great VCO.
ATOM |
|
Back to top
|
|
|
sonicwarrior
Joined: Dec 22, 2005 Posts: 266 Location: Cologne, Germany
|
Posted: Tue Nov 08, 2011 1:38 pm Post subject:
|
|
|
I've got a TZFM PCB + Panel from Bridechamber today and have some questions:
If I want to use a 1K Tempco (because I have them) I have to change R6, R7 and R9 from 100K to 50K, right?
Do I have to change R11 and R13 too then?
I'm also not quite sure about R5 (56K) in this case as I'm not really able to read schematics and understand everything that is going on. |
|
Back to top
|
|
|
frijitz
Joined: May 04, 2007 Posts: 1734 Location: NM USA
Audio files: 54
|
Posted: Tue Nov 08, 2011 2:04 pm Post subject:
|
|
|
sonicwarrior wrote: | If I want to use a 1K Tempco (because I have them) I have to change R6, R7 and R9 from 100K to 50K, right?
Do I have to change R11 and R13 too then?
I'm also not quite sure about R5 (56K) in this case as I'm not really able to read schematics and understand everything that is going on. |
You need to halve the values of all the other resistors connected directly to the (-) input of A1a. That is: R2, 4, 5, 6, 7, 9, 19. No other changes.
Ian |
|
Back to top
|
|
|
|