No that is not correct. I can hit the same key as many times as I want and it still plays. Once a key is assigned to a voice it stays with that voice until you hit the next key. i have tracked the output and you can see it move down the voices sequentially. I have an opamp that is getting hot on the channel that is not working. It is the upper left corner of schematic on voice diagram. checked all my resistances and compared to good channel and they all look good. I have replaced with several different types of opamps but no difference. I have replaced the 2044 and know that it works because when you get to that channel and it doiesnt sound you can push the SUB button on top panel turning it on and the key will sound through the channel. If you look at schematic it has a pathway directly to the 2044. One thing I noticed last night is that the ch 0-3 all have the correct looking sawtooth waveform at the summing point QA transistor and opamp. But looking at ch 4-7 none of them have a sawtooth waveform it looks just like the spike input described in the technical notes. Also the output of the 0-3 is in the 2 volt range where 4-7 are at 5.5 volt range. All except the hot opamp which is up around 9 vdc. Mind you channels 4,6&7 all sound normal just 5 that is not correct. Need a good Synth tech. It's all analog and spread out across a large board.
I understand. Looking at the schematics, I still think the problem could be the key switch matrix (Key scan and key data area). Since the missing note key output functions again, look at the LS138 area. The input logic to select key matrix zone on A0, A1,A2,A3 may be not shifting on one pulse OR one output on KS0 to KS7 is not there. Most likely one in KS0 to KS7 output.
Regards
Regards
Well it cant hurt to check it out. I do not believe I'm having a problem with a key because when I get to the key that is not working I can switch on the sub osc on/off and hit the same key now makes a tone turn it back off and the same key now makes no sound. At this point I need to start attaching more scope shots so you guys can see what i'm seeing. the fact that an overheating opamp is in the sawtooth gen circuit is not working and it is the same channel that is not sounding and goes directly to the VCF. Also it is the same channel every time that does not sound, CH 5.
LS138 is a demultiplexer. Decodes the 3 binary inputs to 8 decimal outputs. Inputs are run from the bus from a shift register. As more keys are pressed, the shift register keeps moving to detect new notes. If this section is defective, one output is not turned ON, then that key press won't be detected as a note. I see the bus is also shared with other selection switches. So when you turn on/off one of the other switches, the bus shifts. Regards.
LS138 (I17) and 4049 (I9) form a 8 x 8 matrix for the keyboard switches.
With A3 (pin 6 on LS138) high, the outputs sink in seq one at at a time following the bus signals A0, A1 and A2. All other outputs are high. If pin 6 go low, all outputs will go high.
4049 inputs are kept high with 8 x 3.3 k resistors. When a key is pressed, it's signal at 4049 input will go low when corresponding LS138 output goes low.
If there is a hard fault, the particular keys will go dead. Since the fault is shifting like every 8th note, fault is around this matrix area. Regards.
With A3 (pin 6 on LS138) high, the outputs sink in seq one at at a time following the bus signals A0, A1 and A2. All other outputs are high. If pin 6 go low, all outputs will go high.
4049 inputs are kept high with 8 x 3.3 k resistors. When a key is pressed, it's signal at 4049 input will go low when corresponding LS138 output goes low.
If there is a hard fault, the particular keys will go dead. Since the fault is shifting like every 8th note, fault is around this matrix area. Regards.
Hi Freecrowder,
Are all 4 sections of IC 'C' (a '4066) paralleled? My copy of the PDF is 'bleached out' or something in that area -- can't tell for sure. But I don't see any other sections of 'C' around the board. For the recent set of symptoms, that little bud ought to be getting a really close, thorough sniff, especially since you reported a 'stuck' 5,5V where the others had a sawtooth.
Also, I wouldn't expect an op-amp to stay cool with its inputs floated.
And forget about supply bypassing with 47pF -- that won't help a bit. Think in the range of 0,1 to 0,47uF. An LM4562 in a '4558 socket/circuit is probably asking for trouble, too. Try to sub with something older/slower/more stable.
Cheers
Are all 4 sections of IC 'C' (a '4066) paralleled? My copy of the PDF is 'bleached out' or something in that area -- can't tell for sure. But I don't see any other sections of 'C' around the board. For the recent set of symptoms, that little bud ought to be getting a really close, thorough sniff, especially since you reported a 'stuck' 5,5V where the others had a sawtooth.
Also, I wouldn't expect an op-amp to stay cool with its inputs floated.
And forget about supply bypassing with 47pF -- that won't help a bit. Think in the range of 0,1 to 0,47uF. An LM4562 in a '4558 socket/circuit is probably asking for trouble, too. Try to sub with something older/slower/more stable.Cheers
Ok great help guys will be looking at it closer today. I have already replaced that 4066, 4013 the main voice chip 2044 and of course the Sawtooth gen input opamp which gets hot. Right now I believe that has something to do with my problem BC it is the voice that doesn't work. I will look closer with Key issue but if i had a keying problem it would be there all the time. The key that had the issue would not go away by turning on the sub osc button on front. In 4 voice poly mode all keys work. again I need to start feeding some scope shots I can't expect much with out some data to share.
As far as electrolytics go the ones in the bad channel are good. It's unfortunate that the one area I'm having problems with are smudged out in the photocopy feeding the hot opamp. What I'm thinking about are lifting the input /output pins of the dual opamp and grounding them for each one individually and see which circuit is causing it to get hot. try to isolate to the circuit.
I don't think the heating could be due to output loading on the lower section of op-amp 'A' with the only load being a 240k resistor.
What is the pin 5,6,7 section of op-amp 'I' doing? Looks to me like it's supposed to clamp the positive excursion of the sawtooth at ~7,4V. That diode leaking badly would alter timbre and warm up both IC's.
Much as I'm looking forward to those 'scope shots, the 'socket' voltages for op-amp 'A' (with it out of the board) may tell more of the story. And don't forget about the input protection diodes' ability to add nominally to the dissipation.
I don't recommend floating the op-amp inputs -- it is unlikely to give the reliable indication you hope for. Depending on the device, it may simply stick to one rail or the other, or oscillate rail-to-rail ultrasonically. IMHO the possibility of either result (or probably others I'm not smart enough to imagine) defeats any useful troubleshooting indication.
Cheers
What is the pin 5,6,7 section of op-amp 'I' doing? Looks to me like it's supposed to clamp the positive excursion of the sawtooth at ~7,4V. That diode leaking badly would alter timbre and warm up both IC's.
Much as I'm looking forward to those 'scope shots, the 'socket' voltages for op-amp 'A' (with it out of the board) may tell more of the story. And don't forget about the input protection diodes' ability to add nominally to the dissipation.
I don't recommend floating the op-amp inputs -- it is unlikely to give the reliable indication you hope for. Depending on the device, it may simply stick to one rail or the other, or oscillate rail-to-rail ultrasonically. IMHO the possibility of either result (or probably others I'm not smart enough to imagine
) defeats any useful troubleshooting indication.Cheers
I do agree the half circuit with 240K is not likely and I have verified that it is 240k.
Yes IC I is to clamp and protect the analog switch from damage from high excursion when frequency command changes. I have tested the diode with just a diode check in circuit and it tests good. Is that a good enough test you think? Would be just as easy to replace it if in question? BTW I have attached a service manual and adobe page 30 describes everything going on around that IC A that is getting hot for reference. It is in earlier post.
I will unsocket IC A and take voltage readings at each pin and report back. Also try to add some scope shots. One issue that complicates things is that 1/2 A being a voltage to current converter means that when I probe with voltage probe I see no signal at the inverting input. I'm assuming BC with + input grounded so is the - input is virtual ground also. How do you get around this with scope?
Yes IC I is to clamp and protect the analog switch from damage from high excursion when frequency command changes. I have tested the diode with just a diode check in circuit and it tests good. Is that a good enough test you think? Would be just as easy to replace it if in question? BTW I have attached a service manual and adobe page 30 describes everything going on around that IC A that is getting hot for reference. It is in earlier post.
I will unsocket IC A and take voltage readings at each pin and report back. Also try to add some scope shots. One issue that complicates things is that 1/2 A being a voltage to current converter means that when I probe with voltage probe I see no signal at the inverting input. I'm assuming BC with + input grounded so is the - input is virtual ground also. How do you get around this with scope?
Typically you'd have to measure the other side of whatever resistor seems to be providing most of the signal -- in this case it looks like the 39k coming from off-board.
We also need to know the voltage at IC 'I' pin 7, as well as what's coming from the ULN2003 (bank of Darlingtons with freewheeling diodes, other board), DC00 through DC07, SYN-42.
The diode check functions probably vary from instrument to instrument; the only ones I've used are older and only assess forward voltage, not leakage or knee sharpness.
That service manual PDF is loaded with goodies, but I haven't gotten to page 30 yet! Thanks for including it.
Regards
We also need to know the voltage at IC 'I' pin 7, as well as what's coming from the ULN2003 (bank of Darlingtons with freewheeling diodes, other board), DC00 through DC07, SYN-42.
The diode check functions probably vary from instrument to instrument; the only ones I've used are older and only assess forward voltage, not leakage or knee sharpness.
That service manual PDF is loaded with goodies, but I haven't gotten to page 30 yet!
Thanks for including it.Regards
So I have taken readings and will attach. First let me explain what I have done to make valid comparison. I took the Opamp A from a correctly working channel and added a socket so that I could make it removable. Now I can compare measurements from the bad channel and the good channel. Looking at DC levels from TP to gnd I see only one that looks significantly different that is the inverting input pin 2. The good channel works correctly with a varying voltage level based on frequency commanded. You can watch this level vary as you go accross the keys and the frequency is changed every 8 key. That is correct according to the explanation from page 30 Sawtooth gen section. That whole page is pertinent.
Now looking at the Bad channel the same pin 2 is always at a constant 2.6 vdc level. It never changes as the good channel does as you go across the keys. That is not correct. It also doesn't ouput a sawtooth signal like the good channel. The input signal to pin 2 on all 8 voices is from 4 opamps and a multiplexor. You can see that pg.13 upper left I11 SAWT. AMP 0-7. I am looking at that point later today. If you look at pg.12 you'll see the chip layout for ID. SAWT. AMP section. I have done continuity checks from each opamp out through the 39K resistors to the inputs of opamp A and they all check good. I think this is the significant section right now. I'll post pics in next post
Now looking at the Bad channel the same pin 2 is always at a constant 2.6 vdc level. It never changes as the good channel does as you go across the keys. That is not correct. It also doesn't ouput a sawtooth signal like the good channel. The input signal to pin 2 on all 8 voices is from 4 opamps and a multiplexor. You can see that pg.13 upper left I11 SAWT. AMP 0-7. I am looking at that point later today. If you look at pg.12 you'll see the chip layout for ID. SAWT. AMP section. I have done continuity checks from each opamp out through the 39K resistors to the inputs of opamp A and they all check good. I think this is the significant section right now. I'll post pics in next post