Arp Quadra synth power supply upgrade

Thread Starter

kevin wiley

Joined Apr 3, 2021
15
Looking at modifying a power supply for an Arp Quadra synth.
I' m replacing caps and any ic that I find that's bad but am concerned about voltage spikes and think it needs some better protection. Any thots along this line? The schematics are on line and I am starting my upgrade to this old synth. Would love to hear from anyone.
Kevin
 

Attachments

Delta Prime

Joined Nov 15, 2019
1,147
Hello there :) welcome to AAC!
I' m replacing caps and any ic that I find that's bad but am concerned about voltage spikes and think it needs some better protection.
Referencing the repair service manual
Is always nice to have if you don't have it already. Could you possibly take a photo of the surgery you have already performed? :)
 

Attachments

dl324

Joined Mar 30, 2015
13,082
Welcome to AAC!
am concerned about voltage spikes and think it needs some better protection
Where do you think this protection needs to be? The regulator outputs won't be a source of problems. The inputs will handle "spikes" as long as they don't exceed their maximum input voltages.

Better copy of the schematic:
clipimage.jpg
 

Thread Starter

kevin wiley

Joined Apr 3, 2021
15
I have all the schematic diagrams and I purchased a capacitor replacement kit for all caps as well as 4075 filter mods and main cpu oscillator upgrade kit and tempco resistors for the vco's etc. I have yet to begin revamping everything, was looking for some advice about power spikes at power on/off, if anyone could see a need to improve the original ps.
Thank you for responding,
Kevin
 

Thread Starter

kevin wiley

Joined Apr 3, 2021
15
I don't really have the expertise to suggest an improvement, but I'm sure someone here does.
Thank you all for looking,
Kevin
 

dl324

Joined Mar 30, 2015
13,082
did not know the regs kept all the spikes at bay at startup
That's the function of voltage regulators. All three of the regulators are of the linear type. As long as the input voltages stay within the allowed range, the regulators will regulate their outputs.

That said, there could be cases where spikes on the input could perturb the outputs, but we'd need to know the problem(s) that was causing before an appropriate mitigation could be implemented.
 

jlm1948

Joined May 19, 2014
7
Actually, spikes happen at turn-on/turn-off. The regulators cannot avoid them, because they are temporarily out of their normal operation area.
Many of the circuits in synths are not balanced, so at start-up their output voltage goes from zero to their normal operating point, and that is unavoidable.
Now, there are solutions that involve a time delay, during which the output signal is muted, to be unmuted when the other circuitry has reached its permanent regime.
 

Thread Starter

kevin wiley

Joined Apr 3, 2021
15
Good point, was intrigued by the last link and wondered about spikes. Thank you all for responding, just doing my best to protect this old synth.
Kevin
 

3D09

Joined Oct 5, 2017
4
4000A Series CMOS digital logic had a 15V Absolute Maximum Rating, and 4000B Series CMOS digital logic had a 18V Absolute Maximum Rating. The 4000A devices would have a 5V to 10V power supply, and the 4000B devices would have a 5V to 15V power supply. The 4000A devices had numerous problems and were quickly replaced by the 4000B and 4000UB series. Usually a 4000A device would fail because of ESD damage. Analog synthesizers typically have patch bays that can direct ESD to an internal circuit. It was common to use 4016, 4051, 4052, 4053, or 4066 Analog Switches to route signals. Running the analog circuits at +13V/-13V versus +15V/-15V may cause some sonic differences, but maybe the Arp Qudras on classic recordings were running at +13V/-13V? I doubt you would notice a difference. When the 4000A devices in Arp Qudras started to fail, a design issue was probably suspected. If they couldn't reproduce the problem, they would have to make a guess. The resistor was changed to drop the supply voltage from 15V to 13V, to provide an Absolute Maximum Rating margin, for the 4000A devices. This was an obvious design flaw and had to be fixed. The 0.01uf capacitors would provide high frequency decoupling for the voltage requlators, so the voltage regulators may have been going unstable during power-up, or from an inductive load being switched on, like a motor. Look at the +13 and -13V lines with an oscilloscope; if you see a spike above 15V, when you switch things on and off, there's an issue. The SA15A Transient Voltage Suppressors (TVS) add overvoltage protection in case of a circuit failure. Some failures would be a dusty trimmer pot wiper or an intermittent Sense line, because of a corroded contact in the P1 connector. Note, the first schematic shows a P1 connector with Sense lines, the second one does not. I think you're on track. I would make the changes called out in Field Change Notice, except for the wrong resistor value. If there's any 4000A devices installed, you could get some 4000B devices to replace them, but it's not necessary. The service manual mentions a 4016 and 4053 in section 3.5.11. And if your playing the synth on a low humidity day, I would touch the chassis to discharge any ESD before I touched the patch bay wires or controls. If the TVSs get hot, turn it off immediately; this means the voltage is above 15V and the TVSs are clamping the voltage somewhere between 16.7V to 18.5V per the data sheet.
 

Thread Starter

kevin wiley

Joined Apr 3, 2021
15
A big thank you to all who read and responded here on the Quadra ps question. Everything is up and running on the synth as it should, and it sounds great! My concerns about upgrading the ps are likely unfounded, it tested out ok after all the caps were updated on every circuit board and running in normal use. There are over a hundred old tantalums in there, plus resistor corrections in the 4075 filters in each section, and a new clock circuit for the main cpu, designed and created thanks to Synthchaser.com. I will continue to study the circuits, much for me to learn, and a number of ics were bad due to solder bridging (my fault) as I was trying to fix some else's work. There is a question about the op amp bridge rectifier circuit in the phaser section that I don't understand and proceeds to affect the volume control part of the circuit, a 4016 cmos that limits the volume in and out of the transistor ladder circuit that distorts something awful when the two opposed diodes are in circuit. They are presently disconnected to be civil. Notice the compressor circuit at the beginning of the signal flow, which controls the output volume in the 4016. I've replaced most ics in this circuit but cannot understand why the distortion is present with the power diodes CR1 and 2.
 

Attachments

3D09

Joined Oct 5, 2017
4
I assume your refering to diodes CR11 and CR12. Opamp Z118 (1458) is configured as a unity gain inverter, with R47 as the input resistor, and R48 as the feedback resistor. The waveform on RESONANCE FEEDBACK will be an inverted version of the waveform on RES REC (resonance receive), when the 4016 switch is closed. There is also a clamp in the feedback loop. It is comprised of CR11, CR12, R58, and R59. CR11 and CR12 are 1N34A germanium diodes with a 1V voltage drop. R58 and R59 form a voltage divider, with a divide ratio of R59/(R59+R58) = 220/(220+10000) = 1/46.5 The diodes will begin to clamp when the output of the voltage divider reaches 1V. At that time the input of the voltage divider will be at 46.5V. The input of the voltage divider is also the output of the opamp, or the RESONANCE FEEDBACK signal, so it's being limited to +/-46.5V. The 1458 opamp has an absolute maximum rating of 18V. So why clamp at 46.5V? Maybe there's an incorrect resistor value on the schematic? Maybe someone thought it would be cool to go into distortion when the feedback signal got very high, but they sized R59 incorrectly, so that feature never really worked? Without R59 the circuit will clamp at 1V, and therefore it will be a typical distortion effect, so I suspect R59 is open circuited. Or, R58 is short circuited.
 

Thread Starter

kevin wiley

Joined Apr 3, 2021
15
I assume your refering to diodes CR11 and CR12. Opamp Z118 (1458) is configured as a unity gain inverter, with R47 as the input resistor, and R48 as the feedback resistor. The waveform on RESONANCE FEEDBACK will be an inverted version of the waveform on RES REC (resonance receive), when the 4016 switch is closed. There is also a clamp in the feedback loop. It is comprised of CR11, CR12, R58, and R59. CR11 and CR12 are 1N34A germanium diodes with a 1V voltage drop. R58 and R59 form a voltage divider, with a divide ratio of R59/(R59+R58) = 220/(220+10000) = 1/46.5 The diodes will begin to clamp when the output of the voltage divider reaches 1V. At that time the input of the voltage divider will be at 46.5V. The input of the voltage divider is also the output of the opamp, or the RESONANCE FEEDBACK signal, so it's being limited to +/-46.5V. The 1458 opamp has an absolute maximum rating of 18V. So why clamp at 46.5V? Maybe there's an incorrect resistor value on the schematic? Maybe someone thought it would be cool to go into distortion when the feedback signal got very high, but they sized R59 incorrectly, so that feature never really worked? Without R59 the circuit will clamp at 1V, and therefore it will be a typical distortion effect, so I suspect R59 is open circuited. Or, R58 is short circuited.
I am refering to diode CR1 and 2 across the 4558 labeled "Z 12B"... they need to be disconnected or major amplification and distortion are created at Z 14 expander outputs. It all worked before I replaced all the capacitors. I double checked the caps for proper orientation and size match so I don't think any resistors are to blame either. I am the original owner of the keyboard and minimum work has needed to be done on it over the years. so I'm short on answers.
 

Thread Starter

kevin wiley

Joined Apr 3, 2021
15
I guess my real question would be, is it ok to remove CR1 and 2 from the circuit and leave it since it seems to work properly, levels are all good, no distortion or anything. I wish I understood the circuit fully. I appreciate what 3D09 said, but I'm not able to understand too much of it. Highschool electronics and hobby interest since is all I got. But I love analog electronics and synths.
 
Top