DC leakage making pots scratchy?

[Randall]

Hey folks,

I posted a general question a couple of weeks ago regarding scratchy sounding new pots in A 65 DRRI rebuild. I gutted the pcb and re-wired an AB763 on new fiber board from Mojotone. When I inquired with the good folks at Mojo as to whether they knew of any defective CTS pots lately, they took the time to suggest look for dc leakage from bad caps.

I mentioned that I can measure voltage on the board itself, and was advised that this is static voltage from the B+, but if all my caps are good, it shouldn't make much, if any difference. I know the advice here has been to not use fiber in the first place, and I will scrap it out and re-do it if I have to. But I will hate myself all day if I re-do it and still have the same issue. So.... I would really like to find the source before have to rip it out, as much for my own enlightenment as anything.

I made some improvement by redressing a couple of wires inside, but it still scratches around 6-8 on both volumes more than I would like. I currently have 200mv across the Vibrato vol pot, and 160mV across the normal vol with both tone controls wide open. I can see the dc is coming from the 250pF treble cap and the .1uF bass cap, with the treble being the highest. It drops off to 4mV with both tone pots at zero. The .047 cap is passing no dc. This is the case for both channels. Would you think this is enough dc leakage to cause the noise in the pots?

FWIW, I am using the original power supply filter board, the one with the strange totem pole set up. But the amp did not have any issues at all before I re-wired it. It is working quite well, other than this issue.

http://vintagefenderamprepair.com/wp-co ... ematic.pdf

[Leo_Gnardo]

I posted a general question a couple of weeks ago regarding scratchy sounding new pots in A 65 DRRI rebuild. I gutted the pcb and re-wired an AB763 on new fiber board from Mojotone. When I inquired with the good folks at Mojo as to whether they knew of any defective CTS pots lately, they took the time to suggest look for dc leakage from bad caps.

I mentioned that I can measure voltage on the board itself, and was advised that this is static voltage from the B+, but if all my caps are good, it shouldn't make much, if any difference. I know the advice here has been to not use fiber in the first place, and I will scrap it out and re-do it if I have to. But I will hate myself all day if I re-do it and still have the same issue. So.... I would really like to find the source before have to rip it out, as much for my own enlightenment as anything.

I made some improvement by redressing a couple of wires inside, but it still scratches around 6-8 on both volumes more than I would like. I currently have 200mv across the Vibrato pot, and 160mV across the normal vol with both tone controls wide open. I can see the dc is coming from the 250pF treble cap and the .1uF bass cap, with the treble being the highest. It drops off to 4mV with both tone pots at zero. The .047 cap is passing no dc. This is the case for both channels. Would you think this is enough dc leakage to cause the noise in the pots?

Mighty nice of Mojo to point your attention to possibly leaky caps when the real trouble likely is that fershlugginer black fiber board. YES 160 mV is plenty enough to get your controls crackling, and even 4 mV is enough to be annoying (to me.) Here's a little experiment: remove the pot end of each of the tone caps from the board, and run the wire (or another wire if you don't want to remove wired buried under the board) directly to the caps from the T & B controls. THAT will identify the guilty party.

Enough about I'll hate myself - what you'll do is learn a valuable lesson. Avoid black fiberboard. And you'll be able to identify the problem in other amps. It's even turned out to be trouble for pickup builders. One guy on MEF measured only 6 or 7 KOhms between eyelets on black fiber pickup bases he had made. That's WITHOUT the pickup wires attached. Wonder where the signal went? I'm sure he's not the only one that happened to.

[tubeswell]

The slightest bit of DC on the pot will cause a scratch.

[Randall]

Could someone explain why this is?

[gingertube]

Randall I'll have a go.

Take a look at the schematic posted above.
The input triodes have +180V on the anode.
This DC is "blocked" by the 250pF to the top of the treble pot, 0.1uF to the bottom of the treble/top off the bass pot, 0.047uF to the fixed 6K8 MID set resistor.

Any leakage in these capacitors (or in the board itself across the capacitor leads) will result in some DC across the pots. As you change the setting of any of the treble , bass or volume pots then you tap off more or less of that DC which feeds (along with the signal) straight into the grid of the following triode.
The current in the following triode is determined by the tube bias which is the voltage between grid and cathode. When you change the pot setting (in the presence of DC leakage) you change the DC voltage you are providing to the grid of that following triode and hence the bias. That triode suddenly shifts its operating point and you get a step change in voltage at its anode. This couples through following capacitors to subsequent stages as a fast rise time edge which has a lot of high frequency component.
As the pot wiper moves along its resitance track you get some "Bounce". When you are just changing an AC signal level this bounce is not noticeable but when you are also changing a DC level the DC level bounces too. That means the grid voltage of the following triode bounces, its bias (grid to cathode volts) bounces and the anode voltage of that following stage bounces by that same amount MULTIPLIED by the gain of that following stage. The bounce manifests itself as a rapid series of sharp edge DC shifts (which can be considered as a "spiky" AC signal), lots of high frequency content and sounds "nasty".

Hope this makes sense.

Cheers,
Ian

[Randall]

Thank you Ian, that makes perfect sense.

Leo, I tried your experiment suggestion with some strange results. lifting the control ends of the tone caps and wiring directly to the pots did indeed remove almost all of the dc voltage to the pots, however the scratchiness continued. Here's where it gets weird. I could still measure up to 600mV on some eyelets with nothing but solder in them. I found in almost all cases that if I removed the solder with wick using a fair bit of heat, the voltages went away.

Now that all leads are back in place I am down to 3-4 mV on each pot. That's a lot of difference from simply replacing and resoldering a few parts. Now, I think of myself as a good solderer with the proper tools, so I really don't think we are talking about a stray blob. My theory is that when I applied heat to the eyelets, enough to see some sparkling around the edges on the fiberboard, I dried it out enough to stop it from conducting.

I still don't know why I am getting scratchiness, but I am pretty sure this fiberboard is evil, and might explain why I had to go 100uF on my bias cap (fiber board) and the pop the standby switch makes as well.

[Leo_Gnardo]

I could still measure up to 600mV on some eyelets with nothing but solder in them. I found in almost all cases that if I removed the solder with wick using a fair bit of heat, the voltages went away.

Yes that's strange all right.

Theory 1: the eyelet board is OK but the lower board is conductive. (There is a lower board like Fender does, right?) If solder or wires contact the lower board, that's where the fun begins.

Theory B: (I'll call this one Randall's theory.) Moisture contained in the fiber board increases conduction. After heating an eyelet, the area nearby is dried up and less conductive. Over time the board absorbs water from the air and resumes its conductivity.

Now that all leads are back in place I am down to 3-4 mV on each pot. That's a lot of difference from simply replacing and resoldering a few parts. Now, I think of myself as a good solderer with the proper tools, so I really don't think we are talking about a stray blob. My theory is that when I applied heat to the eyelets, enough to see some sparkling around the edges on the fiberboard, I dried it out enough to stop it from conducting.

supports Theory B.

I still don't know why I am getting scratchiness, but I am pretty sure this fiberboard is evil, and might explain why I had to go 100uF on my bias cap (fiber board) and the pop the standby switch makes as well.

Even a couple of mV will leave your pots sounding scratchy. Fiber board apparently is evil - not all - just the type where carbon black or other conductive pigments are used. So far - that's just the black fiberboard. I've never found a similar problem on old Traynors that use a similar fiberboard that's grey or tan.

I use 100uF bias filters all the time. But I'd like to know what problem you encountered that was fixed by increasing the bias filter. For me, some amps lose a background hum with an increased bias filter value.

Pop noise when switching from standby to go mode isn't unusual. There have been several solutions proposed. In one particularly troublesome amp I left the preamp on all the time, and used the standby switch to cut B+ to the power tubes only. That one went from miserable to "where's the noise?"

[Randall]

Theory 1: the eyelet board is OK but the lower board is conductive. (There is a lower board like Fender does, right?) If solder or wires contact the lower board, that's where the fun begins.

Yes, there is a lower board a la Fender.

I use 100uF bias filters all the time. But I'd like to know what problem you encountered that was fixed by increasing the bias filter. For me, some amps lose a background hum with an increased bias filter value.

That was my issue, background hum with stock value cap

Pop noise when switching from standby to go mode isn't unusual.

My problem is the opposite, loud pop going from play mode to standby.

[martin manning]

Theory B: (I'll call this one Randall's theory.) Moisture contained in the fiber board increases conduction. After heating an eyelet, the area nearby is dried up and less conductive. Over time the board absorbs water from the air and resumes its conductivity.

I like this theory.

[Randall]

Plus, I remember now that a couple of weeks ago I was chasing down a stray 16V on another section of the board the same thing happened. After unsoldering/soldering an eyelet a couple of times the voltage just disappeared. I didn't understand why at the time, but I think I do now.

[Leo_Gnardo]

Theory B: (I'll call this one Randall's theory.) Moisture contained in the fiber board increases conduction. After heating an eyelet, the area nearby is dried up and less conductive. Over time the board absorbs water from the air and resumes its conductivity.

I like this theory.

Every summer when it gets damp, I get a load of old Fenders coming through, with the leaky-board syndrome.* July-Aug is the peak. Wet as it's been this year, I'm expecting a bumper crop.

Not just scratchy pots, but squeeks squeals chattering pops and general "Gas Music from Jupiter." Some of it is amusing, but it's not what the guitarist wants to hear.

I rebuild the errant audio circuits on tie strips above the fiber board. So far nobody wants to pay to completely replace the board with better stuff.

[Randall]

Sigh... I guess I have to take this project to G10 land. I sooo did not want to have to do that. What a waste of time and money.

[gingertube]

Randall,
In the day job I do electronics for airborne use. I routinely coat the finished PCB assemblies with a conformal coating. This stuff can be purchased in spray cans. If you can dry the board then spraying them with the conformalm coating will then keep the moisture and hence leakage out. It also helps high voltage insulation.

In a previous job I used to do Medical Electronics amoungst which were pH meters with 100MOhm input impedance amplifiers. With those I had to bake the boards and re-coat them about every 12 months.

With a guitar Amp Board you should only need to dry and coat once.

Use a fan heater for an hour or just a desk lamp over the unit for say 24 hours.

This is the sort of stuff I'm talking about.
http://au.mouser.com/ProductDetail/Tech ... 4L9ApRFxFY

These coatings can be "soldered through" for repairs.
Cheers,
Ian

[Structo]

It seems as if the old fish paper recipe has not been improved over the years.

Note to self: If I ever build a Fender amp, use G10.

[Leo_Gnardo]

It seems as if the old fish paper recipe has not been improved over the years.

Note to self: If I ever build a Fender amp, use G10.

What I get from Antique/CE, sold as "fishpaper" is a good deal thinner than the Fender style circus board. Also it's grey and I've never had a problem with it going conductive. Makes a good insulator when you need that. They also sell "cambric" rated to 5000V, but I don't have any experience with that.

Yes, note to all our selves, friends & everyone else besides - avoid that black fiber board. Maybe some is good, but there's dodgy stuff out there that will just cause headaches & frustration. Some kits include it. Avoid them or substitute another build method, glass/epoxy or phenolic or tie strip.