SAD1024 clock noise help

[amptramp]

You may find a separate supply for the LFO plus the electronics it feeds going into the BBD with only a ground connection between the LFO/digital parts and the analog parts is one way to get rid of noise.  Using a low-power op amp like the TL061 for the op amp may alleviate some of the noise.

[four_corners]

I just had a thought, if the clock waveform on the scope changes frequency smoothly with the deepest LFO modulation, then there probably isn't anything wrong with that side of things.
I think the thought of others was that the clock for the BBD chip was stalling as the LFO waveform reached a peak of its swing so the BBD stops sampling the audio for long enough to introduce a "black spot" occurring at the same rate as the LFO speed.

The SAD1024 BBD chip can be clocked up to 1.5Mhz and I'm wondering about the numbers in the service information. That seems more like the setting for a chorus effect than a flanger to me. Flangers need to reach a fast clock speed at one end of the sweep in excess of 500kHz.

LFO ticking is most often a problem caused by the fast switch of the squarewave part of the LFO causing a glitch picked up by the audio circuits often via the power supply connections. In old equipment, this can happen when power rail smoothing capacitors have degraded, nearly always these are the electrolytic variety. Many equipment restorers replace them as a matter of course.

After a lot of fiddling last night, I was also leaning towards replacing some of the capacitors around IC11 and IC12 like you mention. They don't look bulging or anything, but the issue is also a subtle one, so it could be a non visibly bad cap, but one that still is effecting the LFO nonetheless. I'm guessing by your comment I should look for caps that are being fed power, not outputting signal?

On another note, after a few hours of research I found a few options about smoothing the edges of the square wave to alleviate the issue that sound promising.

First is from Stomboxology...



I also found this on an MXR Flanger post on the forum.
GibsonGM said this...

"there's only 1 other method I know of for silencing the 'tick'.  That is to find where the 9V power supply enters the board, and place a 100 ohm resistor in series with it.  Then, where the 9V and resistor meet, you place a 1000uF (approx) capacitor to ground.  Acts like a reservoir to supply the ckt when it calls for power, which may be causing the tick."


My problem is figuring out how to implement this correctly. I know enough about electronics to build a zillion pedals and fairly elaborate synths and debug them, but I'm not the best at modifying existing circuits...

[anotherjim]

There is surprisingly little in the way of power rail filtering. There already is 100R and 100uF for the +15 and -15 feeds and these are actually close to the LFO IC8. I would replace C43 & C44 with new caps of at least 25v rating. The 16v ones shown on the schematic are too close to 15v working for my liking. I would also (fit on the back of the board) a 100nF ceramic capacitor across the same pins as the electrolytic capacitors. This is standard practice, the electrolytic is good at filtering supply hum while a ceramic handles high-speed noise and glitches. Best practice places a ceramic to ground on every IC power pin but they haven't done that and to be fair, there isn't a lot of space on the board for "nice to have" extras.

There is a hack to slow the switching speed of the squarewave by adding a small capacitor on the output but I don't like the idea since it has to charge and discharge fast when the opamp switches and that only adds to the glitch in my opinion.

Yes, only worry about power rail capacitors for now. The ones elsewhere are usually under less stress and a bad one may or not be noticeable as a problem.

[four_corners]

There is surprisingly little in the way of power rail filtering. There already is 100R and 100uF for the +15 and -15 feeds and these are actually close to the LFO IC8. I would replace C43 & C44 with new caps of at least 25v rating. The 16v ones shown on the schematic are too close to 15v working for my liking. I would also (fit on the back of the board) a 100nF ceramic capacitor across the same pins as the electrolytic capacitors. This is standard practice, the electrolytic is good at filtering supply hum while a ceramic handles high-speed noise and glitches. Best practice places a ceramic to ground on every IC power pin but they haven't done that and to be fair, there isn't a lot of space on the board for "nice to have" extras.

There is a hack to slow the switching speed of the squarewave by adding a small capacitor on the output but I don't like the idea since it has to charge and discharge fast when the opamp switches and that only adds to the glitch in my opinion.

Yes, only worry about power rail capacitors for now. The ones elsewhere are usually under less stress and a bad one may or not be noticeable as a problem.

Well, put some 50V 100uF in c43 and c44, plus the 100nF across each of those electrolytic caps with no luck...

There are 2 more power caps, c47 and c48, so I could possibly replace those.

Would it make any difference to put larger caps in these 4 spots?

Thoughts on the softening of the triangle and square waves like the Stomboxology article mentions? The hack you mentioned that you don't like was different right (slowing the switching speed)?

Thanks again for all your help Jim, I really appreciate it.

[four_corners]

I went ahead and socketed IC12 just to see if that was the issue, but changing that opamp didnt do anything (I also replaced C51 since I was right there and had easy access, no help either).

I realized too that I did make a mistake earlier. I said I was getting the triangle and square wave off IC12, but it was actually IC11. IC11 pin 1 and 7 were giving me the triangle and square wave that I included images of. As far as IC12, I am getting a sine wave at the same frequency on pin 1,2, and 3. I guess R102 and C50 are filtering the LFO from IC11 into a sine wave.

Could this just be some sort of antenna-ish thing happening from specific ways the circuit is designed? It is just frustrating as I've heard numerous other Korg Tridents that are dead silent when the Flanger is engaged. It is just unfortunate since the Flanger makes up a huge part of the Trident sound, and it isn't a very common synth.

[anotherjim]

It could well be radiated but old keyboards generally had a lot of noise going on. Be careful with the various sound source levels in the synth versus the main volume control or the volume of whatever you're monitoring it with. If the synth sounds are set too quiet, any volume boost after that will also bring background noise to the fore.

[four_corners]

It could well be radiated but old keyboards generally had a lot of noise going on. Be careful with the various sound source levels in the synth versus the main volume control or the volume of whatever you're monitoring it with. If the synth sounds are set too quiet, any volume boost after that will also bring background noise to the fore.

Yeah this definitely isn't a noise floor thing, I can hear the noise floor in the background as a separate sound, this is specifically LFO ticking that only comes in with the Flanger.

Is it worth replacing the 4013 IC? I'll need to order one if so. At this point I'm grasping at straws here unfortunately. I'm also trying to wrap my head around that Stomboxology hack to see if I can figure out how to implement that, but not sure I know how yet.

[anotherjim]

I'd advise against possibly unnecessary replacement work on something like this. You don't want accidental damage to that SAD1024 - it's rare now. The 4013 is a logic chip and those generally either work or don't with no in-between.

[anotherjim]

BTW, it's IC11 that I would try changing for something else. A TL062 or TL022 or even an MC1458 is worth trying.
For the Stompboxolgy (actually Boss) "soft square" hack, I think you need to cut the ground to IC11 pin6 *. Then insert the 68k between pin 6 and ground then add the 33nF feedback cap between pins 6 & 7. Don't need to copy those exact values, 47k and 47nF probably work just as well and 33 cap values aren't carried by all stockists.
* Track cutting might be messy. An alternative is deadbug wiring. Bend pin 6 up away from the board and solder the resistor in between the pin and ground and add the cap over the topside of the pins.

[four_corners]

BTW, it's IC11 that I would try changing for something else. A TL062 or TL022 or even an MC1458 is worth trying.
For the Stompboxolgy (actually Boss) "soft square" hack, I think you need to cut the ground to IC11 pin6 *. Then insert the 68k between pin 6 and ground then add the 33nF feedback cap between pins 6 & 7. Don't need to copy those exact values, 47k and 47nF probably work just as well and 33 cap values aren't carried by all stockists.
* Track cutting might be messy. An alternative is deadbug wiring. Bend pin 6 up away from the board and solder the resistor in between the pin and ground and add the cap over the topside of the pins.

Well this makes me happy that I'm not totally dumb, because this morning I spent about an hour trying to do the soft square hack, and did exactly what you just suggested! I basically pulled pin 6 out of the socket so I wouldnt have to cut a trace. Regardless, it didn't work...

I ordered some TL022, so Ill find out in a week or so. I tried a LM358 but didn't notice anything different.

Thanks again for all your suggestions.

EDIT: actually i might have done it wrong, I connected the pin 6 resistor to pin 3 since it is connected to ground, but I just realized that is wrong as pin 3 has a resistor inbetween it and ground. I'll go try it out correctly.

EDIT 2: still didn't help...

[puretube]

Did you own/have/play/hear that Synth before you started to work on it, and did it always tick?
Or are you working on it because that ticking (suddenly) appeared?
If it was quiet before, there should be no need in changing the whole topology with stompboxology remedies or trying different kinds of resistor-values or opamp types or other circuit-changes. In this case, it is probably rather a matter of a component that got old, rotten or otherwise misaligned.

The type of square/tri-LFO used here is notorious for "transmitting" spikes from any (longer)connection/wire in this case from pin7 of IC11 to the speed-pot (VR8).
Now: there is a long thin pcb-trace emanating from pin7 of IC12 closely passing between 2 pins of the speed-pot (VR8)on its way north to the input-circuitry of the BBD (R70+R72).
Although this trace is of low impedance origin: if there is dirt/dust/moisture or similar between the traces/solderpoints of the LFO speedcontrol-circuitry and this trace, noises/spikes could be picked up by it, transporting them to the BBD audio input. Be sure the area is clean.
Even inductive coupling via/from those vertical horizontal long jumper-wires could be possible, so these should be bent as indicated on the layout-print. (I can`t exactly see, where they connect in the schemo ...).
And, btw: have you tried slowly moving the BIAS ADJUST (VR6) left/right?

[four_corners]

Did you own/have/play/hear that Synth before you started to work on it, and did it always tick?
Or are you working on it because that ticking (suddenly) appeared?
If it was quiet before, there should be no need in changing the whole topology with stompboxology remedies or trying different kinds of resistor-values or opamp types or other circuit-changes. In this case, it is probably rather a matter of a component that got old, rotten or otherwise misaligned.

The type of square/tri-LFO used here is notorious for "transmitting" spikes from any (longer)connection/wire in this case from pin7 of IC11 to the speed-pot (VR8).
Now: there is a long thin pcb-trace emanating from pin7 of IC12 closely passing between 2 pins of the speed-pot (VR8)on its way north to the input-circuitry of the BBD (R70+R72).
Although this trace is of low impedance origin: if there is dirt/dust/moisture or similar between the traces/solderpoints of the LFO speedcontrol-circuitry and this trace, noises/spikes could be picked up by it, transporting them to the BBD audio input. Be sure the area is clean.
Even inductive coupling via/from those vertical long jumper-wires could be possible, so these should be bent as indicated on the layout-print. (I can`t exactly see, where they connect in the schemo ...).
And, btw: have you tried slowly moving the BIAS ADJUST (VR6) left/right?

I bought it earlier in the year and it has always had this clock ticking.

I'll check the cleanness of that trace as well as the jumpers.

So far, VR6 is the only thing that has actually changed the volume of the ticking, anything else has just changed the rate of ticking. As I turn VR6 to the left, it gets quieter, but fully left the flanger cuts out and just the synth sound is produced with no flanger. Right before it cuts out, it starts getting distorted, like a clipped signal.

I'm guessing VR6 is just changing the amount the flanger is effecting the synth signal, like a mixer sort of?

J-43 is carrying -15V to VR3 and VR4.

J-12 is Vdd for the CD4013 (IC7)

I wonder if using some sort of shielded cable for the jumpers?

[puretube]

VR6 sets the DC working point for all analog samplestages of the BBD. It`s like a center voltage on which the signal AC rides. (That is not neccessarily in the middle of the supply voltages).
The trimmer VR6 can externally shift this DC-level up or down and is set so that the maxima and the minima of the signal excursions just don`t get clipped at their top & their bottom. (Assumed that the signal is not higher than the highest "allowed" nondistorting level).

Because a BBD has the inherent problem that this bias-voltage internally rises with rising clock-frequencies, Korg helps to fight this rising DC-shift by adding some correcting voltage through R103 from the wiper of "Intensity"-pot VR10, which in fact is the user-control of the clock-frequency sweep-width.

[sorry - bedtime, here now ...]

[four_corners]

VR6 sets the DC working point for all analog samplestages of the BBD. It`s like a center voltage on which the signal AC rides. (That is not neccessarily in the middle of the supply voltages).
The trimmer VR6 can externally shift this DC-level up or down and is set so that the maxima and the minima of the signal excursions just don`t get clipped at their top & their bottom. (Assumed that the signal is not higher than the highest "allowed" nondistorting level).

Because a BBD has the inherent problem that this bias-voltage internally rises with rising clock-frequencies, Korg helps to fight this rising DC-shift by adding some correcting voltage through R103 from the wiper of "Intensity"-pot VR10, which in fact is the user-control of the clock-frequency sweep-width.

[sorry - bedtime, here now ...]

Thanks so much for the explanation, I know you didn't have to go out of your way to explain this, but I feel really grateful that people still exist in this world that are willing to educate novices like myself just for the sake of spreading knowledge.

Anyway!

Something interesting I just noticed is that the flanger sounds the most like a "correct" flanger sound when VR3 and VR4 (Clock center and clock width respectively) are turned to max resistance. If they are set anywhere in the middle, it starts sounding like a doubled up flanger, like 2 sine waves where one is running a little slower than the other so you get like a bad ensemble flange sound. Maybe I should change these trimmers to higher values?

 

[puretube]

BTW: in reply #4 you mentioned VR5 & VR7 doing nothing.
While I haven`t found VR7 in the schemo yet,
trimmer VR5 is labeled "MOD.LEVEL", and presets the max. intensity/width/depth with which the Joystick can "sweep" whatever function it is assigned to. With the users "JOYSTICK INTENSITY"-pot turned to zero, it changes nothing at all. So it does not concern our problem.

Gimme some time to find out about trimmer VR7 ...
[edit:] found it: it says "DELAY TIME" and belongs to a "DELAY VIBRATO"-section, which imho is of no concern for the Flanger-section. (And has no influence at all, when the "DELAY VIBRATO INTENSITY"-pot is turned to zero). Not our problem.

On page 22, point 8. of the service-manual:
https://www.vintagesynthparts.com/wp-content/uploads/2018/01/Korg_Trident_MK1_Service_Manual.pdf
there`s the adjustment procedure (you posted before) for the Flanger:


BUT: it contains a typo IMHO!
It says the idle clock-frequency at the lowest "MANUAL"-setting (VR11) should be roughly 100kHz (90~110kHz).
Okay!
It further says the clock-frequency at the highest "MANUAL"-setting (VR11) should be roughly 1kHz (0.9~1.1kHz).
FAIL! (by a factor of thousand).
This correctly should read roughly 1MHz (0.9MHz~1.1MHz i.o.w.: 900kHz~1100kHz).
During the measuring, "INTENSITY" (VR10) needs to be turned waydown to zero!

Do you BTW still hear the ticks, when "INTENSITY" is ZERO?

You can flange manually with the "MANUAL"-pot, without the LFO being involved - only the control-voltage swept by "MANUAL" gets you through the range from 100kHz to 1MHz, when VR3 & VR4 are set appropiately. These interactively influence the clock frequency and of course depend on the given max/min-setting of the "MANUAL"-pot, so you gotta fiddle with the two trimmers while sweeping the "MANUAL" to its extreme settings. Start with just setting VR3 in the middle and try to adjust VR4, so the clock doesn`t go higher than 1.1MHz when "MANUAL" is at max ...

When this unmodulated clock-frequency-range is set according to the procedure,
you can concentrate on the LFO again (by turning up the "INTENSITY"-pot), to hear if the clicking situation (if it was absent with intensity down) has improved.
If neccessary, the "BIAS ADJ" (VR6) by now might need some realignment (by ear).

[four_corners]

BTW: in reply #4 you mentioned VR5 & VR7 doing nothing.
While I haven`t found VR7 in the schemo yet,
trimmer VR5 is labeled "MOD.LEVEL", and presets the max. intensity/width/depth with which the Joystick can "sweep" whatever function it is assigned to. With the users "JOYSTICK INTENSITY"-pot turned to zero, it changes nothing at all. So it does not concern our problem.

Gimme some time to find out about trimmer VR7 ...
[edit:] found it: it says "DELAY TIME" and belongs to a "DELAY VIBRATO"-section, which imho is of no concern for the Flanger-section. (And has no influence at all, when the "DELAY VIBRATO INTENSITY"-pot is turned to zero). Not our problem.

On page 22, point 8. of the service-manual:


BUT: it contains a typo IMHO!
It says the idle clock-frequency at the lowest "MANUAL"-setting (VR11) should be roughly 100kHz (90~110kHz).
Okay!
It further says the clock-frequency at the highest "MANUAL"-setting (VR11) should be roughly 1kHz (0.9~1.1kHz).
FAIL! (by a factor of thousand).
This correctly should read roughly 1MHz (0.9MHz~1.1MHz i.o.w.: 900kHz~1100kHz).
During the measuring, "INTENSITY" (VR10) needs to be turned waydown to zero!

Do you BTW still hear the ticks, when "INTENSITY" is ZERO?

You can flange manually with the "MANUAL"-pot, without the LFO being involved - only the control-voltage swept by "MANUAL" gets you through the range from 100kHz to 1MHz, when VR3 & VR4 are set appropiately. These interactively influence the clock frequency and of course depend on the given max/min-setting of the "MANUAL"-pot, so you gotta fiddle with the two trimmers while sweeping the "MANUAL" to its extreme settings. Start with just setting VR3 in the middle and try to adjust VR4, so the clock doesn`t go higher than 1.1MHz when "MANUAL" is at max ...

When this unmodulated clock-frequency-range is set according to the procedure,
you can concentrate on the LFO again (by turning up the "INTENSITY"-pot), to hear if the clicking situation (if it was absent with intensity down) has improved.
If neccessary, the "BIAS ADJ" (VR6) by now might need some realignment (by ear).

I figured out yesterday actually that VR5 and VR7 are both dealing with the joystick/modwheel stuff, they just have both circuits on that same board (as you've noticed).

Yay, I'm not crazy!!! I've always been getting around 500-900 kHz for the Manual pot after trying to calibrate the clock frequency which seemed to be 10x off. Though, at zero, I haven't been able to get down to 100 kHz, the lowest is roughly 500 kHz.

When intensity is at zero, there is no clicking, but if I manually flange with the Manual pot, I get clicking as the pot turns. There is one exception to this, sometimes when the Manual pot is fully CW or CCW, I'll get some alien aliasing/snake hissing sort of noises for a few seconds, but it sounds like a different noise than the clock ticking. If I just stop a tiny tiny bit short of fully CW or CCW on the Manual pot, this noise is never heard, so I'm not super worried about this as it is avoidable.

Is it good that the clicks are not heard with the Intensity potentiometer at zero?

It seems like it might be worth probing around between the output pin 1 of IC12 and Q19 (that are right by VR4 and VR3), as if the output of pin 1 looks clean, then I know the clicking is coming in post-LFO generation. Should I be able to see some sort of spike/clipping on the LFO on my scope if I find it?

[puretube]

Should I be able to see some sort of spike/clipping on the LFO on my scope if I find it?

That`s what I tried/hoped to see at pin7 of IC12. (from there creeping into the audio-path).

Is it good that the clicks are not heard with the Intensity potentiometer at zero?

It leaves a chance, that ticking is nevertheless induced in that long thin vertical trace to pin6 of IC12.

[puretube]

If you (momentarily) hook up a cap (100nF) from the wiper of the Manual-pot (VR11) to ground, does this help the alien noises or the ticking?
Try same test (wiper-to-ground-cap) at the Intensity-pot (VR10), too.
Both with Intensity max & min.

[puretube]

The proper setting of VR 3 & VR4 is neccessary to find out if the ticking is induced spikes from the LFO into the audio-path,
or if it comes from the clock-generation "bouncing" at its extreme excursions or otherwise influencing the BBD sample-clocking at the rate of the LFO.

[four_corners]

If you (momentarily) hook up a cap (100nF) from the wiper of the Manual-pot (VR11) to ground, does this help the alien noises or the ticking?
Try same test (wiper-to-ground-cap) at the Intensity-pot (VR10), too.
Both with Intensity max & min.

We are getting somewhere!!! So the cap from ground to Manual pot wiper didn't fix anything... BUT, I tried hooking the cap to one side of the Intensity pot, and the clicking totally turns from clicking, to a smoothed static sound!!! With Intensity at 10, there is a new sort of high resonance heart beating sound, but I tried a green 470nF polyester cap I had sitting next to me and the heart beating sound dropped down a ton. This lug on the Intensity pot is connected to Pin 1 of IC12.

If I want to try out a higher value electrolytic, which direction would I place it? Positive lead to pin 1 and then negative to ground?

(EDITED per next page correction)