Power Supply Quieting

[guitarkitbuilder]

In my music schools we use a number of Line 6 Spider III 150 watt amplifiers, and the controls on them have always been a bit screwy.  The controls, such as a six position effects selector, use potentiometers which I assume go to some sort of A/D converter with triggers for when each switch "position" has been reached.  There are small LEDs around the perimeter to indicate which effect is selected.  However sometimes these LEDs will dance around a bit (change on their own, without the knob being touched), and clearly the amp is confused about which effect is selected.  This is at best disconcerting and at worst prevents you from selecting something.  And worst of all, it's intermittent - comes and goes, and we've never been able to correlate it with any other activity or location. 

My assumption further has been that this is due to some sort of power line noise or poor power supply regulation or noise filtering.  We have six of these amps and they all do it from time to time at all locations.  While I haven't asked them in a couple of years since we bought them, the retailer and manufacturer have both answered along the lines of "check your power," ignoring the fact that this happens in all three commercial buildings we lease, as well as in my home, as well as in several different clubs where we've used them to gig.  Their answer might as well be "upgrade the mid-atlantic power grid" as far as I'm concerned, but I digress.

I've got one of these amps on my bench right now to fix an input jack.  I thought I might also try to address the noise problem.  I'm soliciting the forums suggestions for 1) your agreement or alternative theory on what the cause might be and 2) if you think it's a power noise issue, where would you attempt to improve the power conditioning and how?  For example, I might try a ferrite core on the power cord, or something in the filter section, or perhaps something at the potentiometers?  I'm really looking for practical experience here on suppressing transients that might confuse switching logic.

thanks

[J M Fahey]

In a very general way, these pots furnish a certain voltage, say, from 0 yo 5 volts, the digital part scans those values (one per pot) and decides what to do.
Say, from 0... to 0.2 volts it's reverb, from 0.2 to 0.4 it's delay, and so on.
I guess that if it's accidentally set very close to some turnover point, any small voltage variation, noise, interference, or plain old dirt/grime in the contact (which in a regular analog pot might show as scratch) will make that voltage to be ether above or below the threshold point, causing it so select either one or the other effect, seemingly at random.
Don't see a home made cure for it.
Maybe Line 6 will furnish better pots next time, or filter PSU better, or shield better, but that's beyond anything you or us can do.
At least that's what I see.

[joecool85]

In a very general way, these pots furnish a certain voltage, say, from 0 yo 5 volts, the digital part scans those values (one per pot) and decides what to do.
Say, from 0... to 0.2 volts it's reverb, from 0.2 to 0.4 it's delay, and so on.
I guess that if it's accidentally set very close to some turnover point, any small voltage variation, noise, interference, or plain old dirt/grime in the contact (which in a regular analog pot might show as scratch) will make that voltage to be ether above or below the threshold point, causing it so select either one or the other effect, seemingly at random.
Don't see a home made cure for it.
Maybe Line 6 will furnish better pots next time, or filter PSU better, or shield better, but that's beyond anything you or us can do.
At least that's what I see.

I agree.  The only way you could get it more reliable is if you could feed that part of the circuit with a regulated voltage (via LM317 etc).  Not sure if that's even possible.  Do you have a schematic for this amp?

[gbono]

They use a potentiometer in that application? Why not and RPG or optical encoder - opps I forgot keep it cheap and cheerful. What about the Spider IV's? Same issue?

[guitarkitbuilder]

They use a potentiometer in that application? Why not and RPG or optical encoder - opps I forgot keep it cheap and cheerful. What about the Spider IV's? Same issue?

I can see the merits of a potentiometer approach, in that it allows you to load amp presets electronically, without having to have a servo or similar mechanism on the switches.  I just think they didn't engineer enough noise rejection into the control logic.  I don't know about the Spider IVs.

Do you have a schematic for this amp?

No.  Line 6 is notoriously unhelpful about providing schematics, and due to all the control switching, LSI chips, etc., it's not an amp that many would attempt to repair. 

[J M Fahey]

I just think they didn't engineer enough noise rejection into the control logic.

I fully agree.
They really don't need to spend a f******g cent on better pots , supplies, etc.; just rewrite some lines of code so the processor stays at the last known value, integrates to filter the noise, etc. Oh well.
I've recently seen a Fender 25W bass amp schematic:
it had only 1/2 TL072 preamplifying the bass signal, and then fed it into a DSP board which "did *everything*".
All pots (volume, treble, bass, presence, etc.) were 10K lin,  fed 5V; and the cursor voltage went to the 8 inputs of a CD4051 (1 of 8 digitally controlled switch); the single output went into the mysterious DSP of which no schematic was shown.
Obviously the pot voltages were scanned in order, and the DSP decided what to do.
Is such amp repairable?
Only in a very limited way: replace damaged switches, pots, speaker or jacks (the "mechanicals"); the amp itself is in the Fender "do not repair ... replace only" list.
One good thing about these amps: when broken, the user (or the shop, under Fender orders) discards it.
You can grab it, junk the PCB and use the cabinet/chassis/speaker/transformer to build your own "SSGuitar approved"  amplifier of your choice for a *very* low cost.
Cabinet, chassis, etc. are the most expensive or troublesome parts for the home builder.

[phatt]

I just think they didn't engineer enough noise rejection into the control logic.

I fully agree.
They really don't need to spend a f******g cent on better pots , supplies, etc.; just rewrite some lines of code so the processor stays at the last known value, integrates to filter the noise, etc. Oh well.
I've recently seen a Fender 25W bass amp schematic:
it had only 1/2 TL072 preamplifying the bass signal, and then fed it into a DSP board which "did *everything*".
All pots (volume, treble, bass, presence, etc.) were 10K lin,  fed 5V; and the cursor voltage went to the 8 inputs of a CD4051 (1 of 8 digitally controlled switch); the single output went into the mysterious DSP of which no schematic was shown.
Obviously the pot voltages were scanned in order, and the DSP decided what to do.
Is such amp repairable?
Only in a very limited way: replace damaged switches, pots, speaker or jacks (the "mechanicals"); the amp itself is in the Fender "do not repair ... replace only" list.
One good thing about these amps: when broken, the user (or the shop, under Fender orders) discards it.
You can grab it, junk the PCB and use the cabinet/chassis/speaker/transformer to build your own "SSGuitar approved"  amplifier of your choice for a *very* low cost.
Cabinet, chassis, etc. are the most expensive or troublesome parts for the home builder.

I just wish to say that was a very good overview of amplifiers and a hint at the simple truth.
That being they use DSP and didgi (or is that Dodgy) control because it's cheap and easy to do but impossible for the home fix it guy to fix if it goes belly up.

@ Guitar kit builder: If you live in a very humid climate this can cause issues. Moisture can actually alter the resistance of the tiny pads an switches causing all kinds of frustrations.
Though it sounds more like a design issue from what you say.
Phil.

[gbono]

just rewrite some lines of code so the processor stays at the last known value, integrates to filter the noise, etc. Oh well.

Ah but if were that simple why not just provide new SW/FW to fix the problem? The solution would be simple to "integrate" if that did solve the problem - without adding latency or timing issues.

Well either way DSP, and smaller fine pitched SMT components, are here to stay. Though it is not beyond the resourceful person to rework and even reprogram these designs - assuming the value of your labor isn't an issue

[J M Fahey]

Ah but if were that simple why not just provide new SW/FW to fix the problem?

And they probably will, as soon as enough customers complain about that.

[joecool85]

Ah but if were that simple why not just provide new SW/FW to fix the problem?

And they probably will, as soon as enough customers complain about that.

Exactly.  If they get no complaints, or even a handful, they won't do anything.  But if they get hundreds of complaints they probably will bother with a fix.

[J M Fahey]

To be more precise, they will need to add a software equivalent to hysteresis, which is one of the electronic tricks to "clean" dirty signals and resolve transducer ambiguity.

[gbono]

Well its water under the bridge as far as getting Line6 to support this isssue - the Spider III is no longer in production. Still think a $.50 encoder would help the problem.

[J M Fahey]

$ 0.50 for an encoder?

Ow my God !!! I'm going to faint !!! xP
Wait until the head acountant hears about this !!! :trouble :trouble

Well, I think that's about the reaction when somebody suggests that outlandish expense.  :lmao:

To put things in perspective, the official minimum wage for a regular factory worker in Guangdong Province is 0.39 U$ an hour.
Specialized workers and supervisors get a mind boggling 0.69U$ an hour.

[gbono]

To put things in perspective, the official minimum wage for a regular factory worker in Guangdong Province is 0.39 U$ an hour.
Specialized workers and supervisors get a mind boggling 0.69U$ an hour

As they say junk in junk out.

Yes, it a great environment for low cost and untrained labor. If there is a skill set there you bet some western influence paid for it and it's more expensive to keep around then you let on. Remember the labor law (LCL) in 08'? Right now the Chinese government is desperately trying to double those slave wages because inflation is coming along with a bimodal wealth distribution. Oh, but really, you should already be in Vietnam by now.