Juno-60 : LFO Always on Auto : Solved :D

[bobbynunn]

Hi,

I just picked up a Juno-60 in pretty decent condition I think. Everything seems to work apart from a couple of the key contacts need cleaning and a strange problem with the LFO.

As I understand it the LFO trigger is a common source of trouble on the 60 but most times it seems to be the trigger button does not work. In my case the LFO is always on automatic. Regardless of whether I set the toggle to automatic or manual the LFO responds as if in automatic mode.

I took a look at the trigger button - opened it up and blew out any dust and gave the contact tracks a good clean but I suspect that the problem is not with the trigger button but perhaps with the toggle switch which selects between auto and manual. I managed to unsolder the switch and again did my best to open it and clean it but still the problem persists. Could the switch be dead I wonder or could there be an issue elsewhere.

I would love to get to the bottom of this if anyone could shine some light! I have multimeter and can solder.

Any input greatly appreciated.

[bobbynunn]

Ok, I ordered an new toggle switch and replaced the old one but the problem persists.

Anyone got any idea at all what I could check for?

Everything else is working beautifully - the LFO itself is working without problems so it's just the being able to select manual mode for it which is not working.

[madtheory]

Check the schematic. There's probably some logic involved with the switch, so diagnose that circuit.

[bobbynunn]

Hi Madtheory,

I've been looking at the schematic searching for anything related to the LFO auto/manual switch.

It appears that PIN number 12 of IC9-40H273 is responsible for sending "1" for Manual and "0" for Auto.

I have been trying to trace the continuity from the man/auto toggle switch to the above mentioned IC but none of the pins is showing contact. There is however contact between the toggle switch across to the same board that IC9 is on. It does reach pins on IC11, IC12 and IC20. It also reaches one side of the battery too.

It doesn't seem to travel beyond the main panel boards via any of the interpanel connections.

I'm no expert so am glad for the input

[madtheory]

My quick guess is you've got the IC numbers wrong. Maybe. There doesn't appear to be a chip called 40H273 in existence. So that could explain why the signal is not reaching that IC! What are IC11, 12 and 20?

What happens when you do the LFO cal? Is the offset correct? You'll need an oscilloscope to do that properly.

[bobbynunn]

Hi Madtheory,

http://www.cyborgstudio.com/synthmp3s/r ... manual.pdf

On page 10 the IC9 40H273 is referred to in Table 4.

It is also mentioned in this thread, bottom post..

http://www.vintagesynth.com/forum/viewt ... =5&t=62440

So i'm not sure it is a mistake?

IC20 seems to perform summing of the VCF.

Page 21 of the SM has IC11 and IC12 relating to the LFO but what exactly they are i'm not sure.

Sadly, I don't have an oscilloscope.

[madtheory]

OK. If it was my synth I would definitely do the LFO cal before ripping out any chips! My guess was that if the offset is high, the LFO could act as if it's always on regardless of the switch. But I'm not sure because I don't understand the buffering here.

I found the data sheet for IC9 here. It's a flip-flop. So check the levels on the pins as per the sheet. If the signal is getting to Pin 1, then replace the chip. If the signal is not getting to pin 1, check continuity from the switch. Manual says LFO switch does not go via CPU but directly to LFO DCO, so you can avoid painful diagnosis with a Logic Analyser

Even if that is not a correct guess, without a scope it's going to be difficult to diagnose what the problem is. You're groping in the dark here. And it would be good, seeing as you're in there, to calibrate the entire synth.

Edit: why is the URL tag not working?

[bobbynunn]

Ok, thanks Madtheory - going to test pin #1 this evening - just hope it's a bad solder joint somewhere, i'll cross my fingers and hope since all the other signals from IC9 are working - or at least the functions associated to it are working correctly - I guess one terminal could always go bad though.

Interesting idea about the LFO offset - I might look into getting hold of an oscilloscope soon, seems like a useful tool in situations such as these.

Really appreciate you taking the time to help here!

[madtheory]

You're welcome:) I enjoy the chance to learn more about diagnostics. It's unlikely to be a bad solder joint by Roland! Their manufacturing is always top class. And even if it wasn't, the problem would've manifested within a few years of manufacture. If you're going down that route, check your own soldering :/

[bobbynunn]

If you're going down that route, check your own soldering :/

Yeah! Can't argue with that.

If its not a bad terminal on the IC then what else could it be most likely - cracked PCB or damaged trace? I guess it could be any individual component going bad or damaged in the pathway....oof

Groping in the dark indeed...

[madtheory]

Well the service manual is one of the best ever because the firmware has diagnostics onboard , anything after that is basic digital electronics and a bit of old school stuff (which was still current 20 year ago) such as flip-flops, Boolean logic serial interfaces, memory registers, hexadecimal code etc. etc. Also know the failure modes of passive and active components. Understand the data sheet for each IC.

Cracked traces and such like are simply continuity. Easy to test assuming you are careful and methodical. The voodoo in diagnostics is to be strictly scientific. Make no assumptions. There is a reason for everything, and the reason will have hard evidence of its existence. So it's no use "hoping" what the problem is. You have to PROVE what the problem is. Maybe a €1500 vintage synth is not the thing to be learning on? Even if it is well documented?

[bobbynunn]

. Maybe a €1500 vintage synth is not the thing to be learning on? Even if it is well documented?

Perhaps not, I shall do as you advise and refrain from "hoping". I realize it's easy to do more harm than good when poking around without any clue. I was simply trying to be optimistic that the fault could be found and fixed without major surgery. I have been lucky so far with other older bits and bobs when it has mostly been a case of replacing a cracked resistor or two, or the cleaning of a fader...The only basis for my optimism is or was that this auto/manual issue is the only single issue the synth exhibits.

I'm currently doing an engineering bachelor but have only had one semester of digital systems so while I understand how a flip-flop works in theory I have not had any experience with them in practice. I will be very careful and gentle and if I am unsure how to proceed I will stop and seek assistance - wouldn't be a good feeling to mess things up further.

[Werecow]

Ok, thanks Madtheory - going to test pin #1 this evening - just hope it's a bad solder joint somewhere, i'll cross my fingers and hope since all the other signals from IC9 are working - or at least the functions associated to it are working correctly - I guess one terminal could always go bad though.

Interesting idea about the LFO offset - I might look into getting hold of an oscilloscope soon, seems like a useful tool in situations such as these.

Really appreciate you taking the time to help here!

A clarification: I don't think madtheory was talking about pin 1 of IC9 specifically (it's not really interesting, since it's tied to ground), just as a general troubleshooting guide. In this case, the point of interest is pin 12, the Q output for the auto/manual latch. That should toggle between high & low logic when you flip the front panel switch between manual and auto LFO. Since it's a latched output, you only need a voltmeter at that point -- no need to dig any deeper with a scope (yet ). Once you confirm that, you'll know which way to go in the circuit. If the logic is switching, you move on to panel board A; if it's not, you have to start looking at waveforms before the latch, which will require a scope, at least a dual channel.

[bobbynunn]

A clarification: I don't think madtheory was talking about pin 1 of IC9 specifically (it's not really interesting, since it's tied to ground), just as a general troubleshooting guide. In this case, the point of interest is pin 12, the Q output for the auto/manual latch. That should toggle between high & low logic when you flip the front panel switch between manual and auto LFO. Since it's a latched output, you only a voltmeter at that point -- no need to dig any deeper with a scope (yet ). Once you confirm that, you'll know which way to go in the circuit. If the logic is switching, you move on to panel board A; if it's not, you have to start looking at waveforms before the latch, which will require a scope, at least a dual channel.

Ah, thank you for that clarification ! So I should check the voltage on pin 12 when flipping the toggle - should there be continuity from the switch to that pin too or not? I remember testing that pin for continuity and there was none.

[madtheory]

That's pretty cool you're doing a BA! Maybe they'll let you borrow an old 15MHz dual channel analogue scope? That will be all you need. They'll probably have logic analyzers lying around too. IME the techs like to hang on to the old stuff, if only to demonstrate how well made the gear is! Also they'll be pretty amazed that anyone still cares about 8 bit data lines and eighties microprocessors... or they might think you're some kind of wierdo lol!

And actually Werecow is being kind and diplomatic. I did mean (incorrectly) pin 1 oops!