Issue with Low Tide Mod build

[pcborpcp]

Hi there,

I got the Low Tide Modulator pcb along with some other projects. While I could finish the other effects just fine, I came across a problem with the Low Tide. I hope you can help me.

So the issue at hand is that the polarity protection diode (D3) gets fried whenever I power up the circuit.
After the first time, I removed all IC’s and transistors, because the board got really hot. But even without transistors and IC’s connected, the D3 gets fried immediately. I even tried to use a N4001 diode instead of the BAT48. The diode remains to get fried and it gets really hot.

I switched out the DC socket, double and triple checked the polarity of the wiring and the DC plug (center negative) and it’s all correct. I even let a friend look over the build, just to make sure I didn’t miss anything.

Do you have any idea, what the problem might be?
I would really appreciate your support on this.

Thanks

 

[pcborpcp]

I'd do the same at both pads of C39, 40, & 41 as that will help us identify if they are the problem. For electrolytic caps, the the square pad is the (+) side and the round pad is the (-) pad side. You should get a beep (or close to 0 ohms) on the (-) pads as they are connected to ground. On the (+) pads, you should get a resistance of less than ~2.2k ohms on C39/40, and a much higher resistance on C41 (possibly no connection).

Use your DMM to measure the resistance between the ground pad and the side of D3 that does not connect to the power. What do you get?

If it is no resistance or low resistance. you will need to isolate the unwanted connection to ground The most likely cause would be a solder bridge between two connections on the board, since you seem to have had the problem before you started pulling pieces off of the board. It is also possible though that the damage to the board from removing parts may have created a short to the ground plane in the board that might be harder to diagnose and repair. You can do a visual inspection to see if you have any stray solder bridges between pins that are close together.

Thanks, I'll measure everything tomorrow and will report back. Thanks for your help, I really appreciate it.

 

[bowanderror]

Thanks, I'll measure everything tomorrow and will report back. Thanks for your help, I really appreciate it.

No worries man, I know your pain! Those photos are a LOT better.

The red circles are potential short circuits, and the yellow ones are solder joints that may or may not be completely soldered (may need to touch them up).

I would also go back and clip any leads that are poking out, especially the stranded wires used for offboard stuff. They may not be the problem now, but can be an issue during boxing.

Here is a color-adjusted closeup of the cap section where the mask has come off:

Most of the square (+) pads look ok, and the solder creep on the (-) pads may not be an issue as long as it's just bleeding into the ground pour. Can't quite tell if there are any traces close to C40.

For anyone else helping with troubleshooting, here is the silkscreen layer with component names:

 

[bowanderror]

Just as a heads up, you may want to turn your soldering iron temp up a bit, that will help the solder flow more easily & give you smoother, more "domed" pad joints. There are lots of leads on the component (top) side of the board where the solder hasn't flowed through, which often means the temp was too low to fully wet the solder. If the joints look lumpy or dull (not shiny), it usually means either your tip wasn't sufficiently cleaned & carried oxidized solder onto the joint or that your iron is set too low for the type of solder/size of the pad. You can also add additional flux to help prevent cold solder joints.

It looks like the type of damage done around the capacitor pads was probably a result of pressing the tip too hard onto the pads rather than heat damage. Always try and make sure your tip is well-tinned, and that you're in contact with both the pad & the component lead. Pressing harder or twisting your tip won't help with heat transfer, so be as gentle as you can and flux/clean your tip for difficult joints.

 

[pcborpcp]

Just as a heads up, you may want to turn your soldering iron temp up a bit, that will help the solder flow more easily & give you smoother, more "domed" pad joints. There are lots of leads on the component (top) side of the board where the solder hasn't flowed through, which often means the temp was too low to fully wet the solder. If the joints look lumpy or dull (not shiny), it usually means either your tip wasn't sufficiently cleaned & carried oxidized solder onto the joint or that your iron is set too low for the type of solder/size of the pad. You can also add additional flux to help prevent cold solder joints.

It looks like the type of damage done around the capacitor pads was probably a result of pressing the tip too hard onto the pads rather than heat damage. Always try and make sure your tip is well-tinned, and that you're in contact with both the pad & the component lead. Pressing harder or twisting your tip won't help with heat transfer, so be as gentle as you can and flux/clean your tip for difficult joints.

Thanks! Your previous post is super helpful and detailed.

My soldering iron temp is usually 572 °F/300 °C, is that too low?

The damage around the capacitor pads happened because I desoldered them and couldn't get the solder out of the holes. I had to push a needle through the hole on one side, while applying pressure with the soldering iron to the other side, that's how it happens. i tried to use a desoldering pump and wick with flux, but the solder wouldn't come out. I usually have no problem with getting holes clean with a pump, but here I tried it for an hour without success.

 

[bowanderror]

My soldering iron temp is usually 572 °F/300 °C, is that too low?

What type of solder are you using?

 

[pcborpcp]

What type of solder are you using?

lead based solder with 2,5% flux.

 

[bowanderror]

lead based solder with 2,5% flux.

What is the Sn/Pb percentage?

 

[pcborpcp]

Alright,

I re-soldered the pads that bowanderror highlighted (updated photos attached)

Then I measured the connections as zgrav suggested.

I put a fresh BAT48 in D3 and measured the resistance between the ground pad and all connections. I also measured all the other electrolytic caps and thought it might be helpful to check continuity of all connections as well. I made a chart with the data - see attachment below.

 

[pcborpcp]

What is the Sn/Pb percentage?

It's 60% Sn and 39% Pb.

 

[Stickman393]

You have a dead short between VCC and ground, judging by those measurements.

Might be a good idea to take a magnifier to the back of the board, try to find any lingering solder bridges. They can be *super* tiny.

 

[pcborpcp]

You have a dead short between VCC and ground, judging by those measurements.

Might be a good idea to take a magnifier to the back of the board, try to find any lingering solder bridges. They can be *super* tiny.

I can't find anything. This is driving me nuts!

 

[Stickman393]

Check there. You're basically looking for any spots where VCC and ground pads are in close proximity.

If you can follow the positive trace after the BAT48 on the board, it might lead you to the culprit.

Basically, when you plug in this thing your power supply drives the maximum amount of current that it is able to deliver through this short on the circuit board. It's no wonder that it's getting hot...

Happy hunting.

 

[pcborpcp]

Check there. You're basically looking for any spots where VCC and ground pads are in close proximity.

If you can follow the positive trace after the BAT48 on the board, it might lead you to the culprit.

Basically, when you plug in this thing your power supply drives the maximum amount of current that it is able to deliver through this short on the circuit board. It's no wonder that it's getting hot...

Happy hunting.

Thanks for chiming in on this, Stickman393.

I checked the part you highlighted, but couldn’t see anything problematic.
I then went on and measured resistance between every component and the positive DC pad. It was a bit of work, but I think I was able to isolate the problematic section. You can see it in the attached picture. Everything in the red section reads O.L. on the DMM. Every other contact has resistance (except some IC pins).

I inspected the section but couldn’t make anything out visually. My next step would be to pull out the caps C2, C3 and then C4 to check if they’re ok. But I thought I'll post this update first, to see peoples opinion on this.

 

[Stickman393]

O.L. on your multimeter means Over Limit. Basically, the resistance between 9v and those points in red is too high for your meter to measure. If your meter has a limit of say, 10 megaohms on the high end, all that means is that the resistance between those two points is more than 10 megaohms.

Given the part you've highlighted, I'd say reading over limit between those sections and 9v is no big deal. BUT, I think that the fact that you've decided to start poking around with your meter and the schematic is GREAT. It shows initiative and a desire to learn and understand the inner workings here: it's basically the same way I started out trying to understand a circuit. Not that I have access to any kind of wealth of knowledge...most of us are still learning here.

The measurements that you took that concerned me were for C22, C5, C41 and C32. Those four should read *far* more than 0 ohms to ground on one side of the component, and 0 on the other side of the component.

The reason for this is that each of these have one lead that connects to ground, while the other connects to a pad that has potential to ground (that is, voltage) once the circuit is powered.

This difference in voltage is what allows these capacitors to build and store a "charge". If resistance on *both* sides of the component to ground is 0, that means that the capacitor will never build a charge, sure, but it also means that the resistance between two points that are supposed to have a difference in potential is zero.

Which...ah...is a problem. Ohms law tells us that Current equals Voltage divided by Resistance, or I=V/R. If we substitute our knowns, that equation becomes I = 9/0.

Since division by zero is impossible, I (current) must be infinite. In reality, every material has some kind of resistance, so current won't necessarily be infinite, but it'll be A LOT.

I think I recall you saying that you were using a 1500ma 9v power supply. So...thats probably about what your upper limit would be. If we apply Watt's law (power=voltage x current), we can see another point of interest.

9v X 1.5A = 13.5 watts

Well...that's significantly higher than what your BAT48 is rated for. No wonder those things are burning up!

Though now that I take a second look at your readings, I'm curious if you measured those components on both sides, or if you swapped the polarity of your leads while measuring the same pad? If you did the latter, I'd recommend going back and measuring between ground and both pads on those four components and reporting back. Polarity will not matter for these measurements

 

[bowanderror]

Sn60/Pb40 solder melts around 361-370°F, so your iron may be set too high for that particular solder. That said, the soldering iron set point for a particular solder is usually higher than the melting point anyway, and is highly dependent on the power rating (in Watts) of your soldering iron and the size of the tip you're using. If you're having trouble getting the solder to fully melt at 570°F, then it's probably either a tip cleaning issue, flux issue, or the tip is shot. A setting of ~400-450°F is more reasonable for the solder you're using. What type of flux are you using?

Stickman is right that your Vcc is shorted to ground somewhere on the board. I'd remove the sockets you have for the +5V regulator and see if that solves it, you can solder that part directly to the board anyways. I know it sucks, but removing components attached to Vcc is a pretty foolproof way to find the short. Like Stickman mentioned, I'd focus on parts that have Vcc & GND pads close together, especially those that may have shorts on the top side of the board (like electrolytic caps).

 

[zgrav]

FYI -- easiest way to remove the 3 pin socket without damaging the board would be to cut the socket into three pieces and then remove it one pin at a time.

 

[Stickman393]

^ I would disagree RE: soldering iron temps there.

Soldering iron temps have a lot more to do with heat transfer than they do with the ultimate temperature of the joint. That's because the majority of the heat applied goes towards a latent process - the non-sensible heat required to turn a solid into a liquid.

Same deal as when you boil a pot of water: the temperature of the water rises to 212⁰F at sea level, but then it starts boiling.

The temperature of the water doesn't rise beyond 212⁰, but heat is still being applied to the water. So where does the heat go?

Behold...the latent heat of vaporization. The heat energy causes a rapid expansion in the volume of the water molecules, at which point individual molecules break loose of the bonds of liquid water and become water vapor. The energy applied is reflected in molecular activity, but not in temperature.

This hidden heat of state change is difficult to overstate: for water, you need one BTU per pound of water to raise the temperature of that pound of water by one degree F. But to evaporate that pound of water? That'll take a total of 970 BTUs. It takes more than five times the amount of heat energy required to bring a pound of water from it's freezing temperature to it's boiling temperature (180 BTUs) in order actually boil the whole pound.

The latent heat of fusion for 60/40 solder is about 395 times the amount required to heat it by one degree. Transferring that much heat efficiently requires either a large tip or a high temperature delta.

Hell, it's impossible to set my own iron any lower than 662⁰f...cause I'm using an induction iron, and the temperatures are determined by selecting a tip rather than a setting on the screen. 350⁰C is as low as they go.

But...anywho...thanks for sticking with me on all that. I work in thermodynamics (*cough*HVAC*cough*) so I like to spread my knowledge of the super awesome world of heat transfer.

I'm a big fan of the first law of thermodynamics. You know, the law of the conservation of energy. That's a good one. I'm lazy as fuck.

 

[pcborpcp]

O.L. on your multimeter means Over Limit. Basically, the resistance between 9v and those points in red is too high for your meter to measure. If your meter has a limit of say, 10 megaohms on the high end, all that means is that the resistance between those two points is more than 10 megaohms.

Given the part you've highlighted, I'd say reading over limit between those sections and 9v is no big deal. BUT, I think that the fact that you've decided to start poking around with your meter and the schematic is GREAT. It shows initiative and a desire to learn and understand the inner workings here: it's basically the same way I started out trying to understand a circuit. Not that I have access to any kind of wealth of knowledge...most of us are still learning here.

Oh man, this is such a bummer. I can’t believe I went over the whole circuit for nothing. But I admire your positive spin on this embarrassing situation

Though now that I take a second look at your readings, I'm curious if you measured those components on both sides, or if you swapped the polarity of your leads while measuring the same pad? If you did the latter, I'd recommend going back and measuring between ground and both pads on those four components and reporting back. Polarity will not matter for these measurements

No, I didn't swap the leads. I put the black lead on the negative DC pad and measured both/all pins of the caps with the red lead.
I just did a measurement again (just to make sure) and it still reads 0,00 Ohm on both pins (+/-) on C22, C5, C41, C32.

Now I did what bowanderror suggested:

I'd remove the sockets you have for the +5V regulator and see if that solves it

And now I get the following readings:
C22: 21 Ohm (+) / 21,8 Ohm (-)
C5: 70 Ohm (+) / 70 Ohm (-)
C41: 54,5 Ohm (+) / 63, 3 Ohm (-)
C32: 64,8 Ohm (+) / 58, 8 Ohm (-)

So it seems Stickman393 was right all along with narrowing it down to IC1/LM78L05! But I guess the resistance of the electrolytic caps is still too low?

 

[Stickman393]

'Atta pepper. No need for embarassment. Its similar to what I tell my apprentices: if you're not breaking anything, you're not learning.

Are you referencing ground or +9v on those measurements?

Measuring resistance across caps is tricky. In actuality, a capacitor contains an insulator, and thus it's resistance is infinite. This is why we use them as "coupling" devices: if you examine most guitar pedal schematics, you can isolate each "building block" of the circuit by using the caps in the signal path as a start and end point.

Basically...these caps are meant to block DC from passing, so that each circuit block can operate independently and predictably when power is applied.

But...caps are not simply an insulator. They are two plates that are separated by an insulator.

These plates operate something like how a magnet would. If a positive charge is applied to one side, it will *attract* (or induce) a negative charge on the other. This property allows alternating current to "pass through" a capacitor.

Alternating current is electrons and holes constantly moving forwards and backwards in opposite directions, constantly changing the force (or voltage) at which they do so, and often changing the direction in which they move.

A capacitor will store a charge across it's plates, but as the electrical pressure is released on a plate the effect is like deflating a balloon. This constant charge and discharge cycle on one plate induces a constant charge and discharge cycle on the other plate, which gives the appearance of AC passing across the insulator of the capacitor. Its, like, basically magic.

DC, however, is different. Electrons travel in one direction, holes in the other. The electrical pressure, or voltage, does not change. Therefore, as voltage is applied across a capacitor, electrons rush into one plate, and holes rush into the other. They (mostly) do not travel from plate to plate, because there is an insulator between the two plates. The voltage never changes, and a capacitor only discharges when there is a reduction in applied voltage.

But remember, electrons and holes can only leave the way they came into the cap. They do not go in one lead and leave through the other. For the most part.

Sooo...what was the point of all that? Ahhhh, erm, well to explain that it is that exact property that tends to confuse ohm meters. Ohm meters operate by applying a voltage...typically around 1-2Vdc, and using the measured current between the leads to determine resistance.

Ohms law. V/I=R. That's all it's doing: taking two knowns: the voltage the meter applies, and the current that it measures, to determine the unknown R Variable.

BUT...you're doing this with a very low voltage that is capable of generating...eh...not a lot of current. Because of this, when you place an ohm meter across a capacitor the meter will see electrons moving...but that current will slow as the cap in builds a charge. Because of this, you will see the resistance measurement increase slowly over time.

If you read all that...uhhh....well, Imma level with you: part of how I learn is by restating what I've learned. That's why a lot of my posts tend to be fuckkkin essays. So thank you for bearing with me.

And second, if that's all gibberish, that's OK. Like I said, nobody should listen to me. Or at least, people should be entirely skeptical of what I have to say and I appreciate when folks pointing out when I get shit wrong. Even if my lizard brain gets all rage and murder for a couple mins.

See? We're all sane here.

 

[fig]

I rather enjoyed the read.