GGG Rat Clone. No sound, circuit board gets hot and might possibly catch on fire

[pedalbuilder1]

Hey guys!  The title is a pretty good description of my problem, but I'll give a little more detail.  I built a "Rodent" (Rat clone) from the General Guitar Gadgets site (http://www.generalguitargadgets.com/projects/17/92-rodent).  I used their PCB layout.  The only substitutions I made were to use a 1N4001 diode instead of a 1N4002 (D3), a 47 pF capacitor instead of 30 pF (C4), and a 2N5457 transistor instead of a 2N5458 (Q1).  When bypassed, the clean guitar signal works just fine.  When the pedal is engaged, there is no sound whatsoever, the LED does not light up, and the circuit board gets extremely hot very quickly.  The source of the heat seems to be around the 1N914 diodes (D1 & D2) and the 4.7 uF cap (C7).  I thought there might be a polarity issue, so I checked and rechecked the orientation of my electrolytics.  I replaced multiple components to try to weed out a faulty one, and after a while, I rebuilt the whole circuit board using all new components and new PCB (I etched two).  Still the same problem.

I have checked and rechecked all the wiring.  I did a clean soldering job both times.  Checked for stray solder bridges and and came up with nothing.  I'm stumped.

I did read the "Debugging: What to do when it doesn't work" thread, but I'm afraid if I start poking around taking readings while power is supplied to the pedal, the damn thing will catch on fire.  It gets REALLY hot in a matter of seconds.  

Any ideas on what might be the problem?  

[mth5044]

Sounds like you are connecting the power rails together somewhere. Check for solder bridges on the PCB or traces that may be messed up. Use the continuity tester on your multimeter to see if ground connects to +9v anywhere.

[ashcat_lt]

If C7 is (possibly failed) shorted, then D2 is trying to pull the 4.5V Vref down to 0.6V, and will take all the current that the opamp will give it in so doing.  How much current can a 308 source? 

Edit -
Course it's limited by that 1K...  Nevermind...  Sorry.

[pedalbuilder1]

Sounds like you are connecting the power rails together somewhere. Check for solder bridges on the PCB or traces that may be messed up. Use the continuity tester on your multimeter to see if ground connects to +9v anywhere.

I have found no solder bridges and I don't think my PCB trace is messed up.  I copied it directly from the GGG website.  I'm not a pro with a multimeter so I'm not exactly sure how to check every spot to make sure my +9v connects to ground.  I set it to the lowest Ohm setting and probed different spots on the the +9v and ground and got the infinity/1 reading.

[PRR]

> might possibly catch on fire

Please post video!!

> heat seems to be around the 1N914 diodes (D1 & D2) and the 4.7 uF cap (C7).

Heat in those parts would take multiple screw-ups.

BUT, for some reason, the layout put R14 in that area. R14 both filters and protects the power. If you've shorted +9V to ground anywhere on the board, R14 throws 1.7 Watts of heat. Since R14 is probably a 0.25W part, it sure will be unhappy.

Unfortunately such shorts are hard to find by meter.

Bright light and a magnifier. Also compare the drawing against the reality. Yes, again.

> see if ground connects to +9v anywhere.

If it connects "anywhere", it will appear to connect "everywhere".

One trick is a super low-range ohm-meter. Most conveniently, a "short squeaker". It makes a tone that rises with a few milli-Ohms of resistance. You move the probes around for the lowest pitch, there's your short. But I'm not finding one with Google, and mine is in the garage and it is too cold now to go dig it out.

[PRR]

> the lowest Ohm setting and probed different spots on the the +9v and ground and got the infinity/1 reading.

Reverse the ohm-meter leads. Or just re-re-check D3.

If you got D3 backward, it is a near-short to power polarity. Depending which way your ohm-meter is polarized, it may read infinity or very-low. Reversing the red/black will try things the other way.

Reversed chip will do the same thing. (What's worse, a hot '308 now that they are precious, or a hot 8088 CPU back in the day they were new? That's how I know a backward chip can act-short and go hot.)

[pedalbuilder1]

> might possibly catch on fire

Please post video!!

> heat seems to be around the 1N914 diodes (D1 & D2) and the 4.7 uF cap (C7).

Heat in those parts would take multiple screw-ups.

BUT, for some reason, the layout put R14 in that area. R14 both filters and protects the power. If you've shorted +9V to ground anywhere on the board, R14 throws 1.7 Watts of heat. Since R14 is probably a 0.25W part, it sure will be unhappy.

Unfortunately such shorts are hard to find by meter.

Bright light and a magnifier. Also compare the drawing against the reality. Yes, again.

> see if ground connects to +9v anywhere.

If it connects "anywhere", it will appear to connect "everywhere".

One trick is a super low-range ohm-meter. Most conveniently, a "short squeaker". It makes a tone that rises with a few milli-Ohms of resistance. You move the probes around for the lowest pitch, there's your short. But I'm not finding one with Google, and mine is in the garage and it is too cold now to go dig it out.

Thanks so much for the info, PRR.  I was coming on here to reply with more info when I saw this post.  My addition was going to be, "Although D1, D2, and C7 felt hot, I realized the heat is actually coming from R14 - the resistor right next to them."

So now I guess I need to find where my +9v shorts to ground.

EDIT:  That would also make sense why both the initial build and the 2nd build/rebuild were having the same problems.  I must have etched the circuit boards poorly (did them both at the same time with the same art transfer) and left some sort of copper bridge between +9v and ground somewhere. 

[stallik]

@PRR any chance of more info on the short squeaker? Never come across one before and searches lead me to circuits with a constant tone. Sounds like a very useful tool

[greaser_au]

Reversed chip will do the same thing. (What's worse, a hot '308 now that they are precious, or a hot 8088 CPU back in the day they were new? That's how I know a backward chip can act-short and go hot.)

Part of a test spec from 'back in the day' (from before OH&S requirements):
Suspected failure:  overheating device.
Test method: Place finger on top of device momentarily.
Test result: Pain sensed, fingerprint pattern attached to device, or reversed manufacturers logo embossed into fingertip.
Conclusion: Device definitely overheating.

david

alright.... you found me out... I made that one up...  

[PRR]

> more info on the short squeaker?

This is it:

http://www.productiondevices.com/Circuit%20probe.htm

The company has been in business for decades. They say 1965, I know them since the 1970s.

Another source: http://www.probemaster.com/product_info.php?cPath=7&products_id=60

[pedalbuilder1]

So I checked and re-checked, and re-re-checked everything and no luck.  I etched a new circuit board and spent a ton of time making sure that everything was especially clean.  No wild solder bridges or close PCB traces.  I rebuilt the whole thing using all new parts on the PCB (again), and it still doesn't work!  So that leaves me with doing something I probably should have done before rebuilding it the first time.  Re-wire it with new pots and new switch.  I don't know how the 3PDT or the pots could be dumping V+ to GND, but it's the only thing left.  I'm building from a verified schem, PCB and parts layout (GGG) and there's no reason why it shouldn't work.  I'll update tonight once I swap out those parts.

[PRR]

> new pots and new switch.  I don't know how the 3PDT or the pots could be dumping V+ to GND, but it's the only thing left.

Divide and confuse. You do NOT need the 3PDT switch for basic testing. You need the pots, but they don't have to be in a metal box. (It will buzz, but buzzy signal beats silent smoke.)

Then as you add frills *one-by-one*, smoke-check each change.

[pedalbuilder1]

> new pots and new switch.  I don't know how the 3PDT or the pots could be dumping V+ to GND, but it's the only thing left.

Divide and confuse. You do NOT need the 3PDT switch for basic testing. You need the pots, but they don't have to be in a metal box. (It will buzz, but buzzy signal beats silent smoke.)

Then as you add frills *one-by-one*, smoke-check each change.

Yeah that is true.  I already eliminated the DC jack for that very reason.  Not necessary for testing.  I'll forgo any switching until I rule out pots.  Thanks for the advice/reminder! 

[amz-fx]

"Although D1, D2, and C7 felt hot, I realized the heat is actually coming from R14 - the resistor right next to them."

If R14 is getting really hot, it is because you are drawing a lot of current through it. I have seen this before in Rat pedals when the power supply was connected backwards...  in which case a lot of current is flowing through R14 and D3. This is also suspected because the LED will not light up either with the reversed power.

Set your meter to measure DC volts and put the black probe on ground anywhere in the box. Power up the pedal and quickly use the red probe to measure either end of R14. If the meter reads -9v (or close) then you have the power connected backwards. If it reads +9v then the problem is something else.

regards, Jack

[duck_arse]

have we seen any photos of your build yet? and are you sure the footswitch is/was wired "east-west" lugs-wise? if you measure the battery voltage as you work the bypass switch, does it drop to zero with one click, then back up with another click?

[pedalbuilder1]

have we seen any photos of your build yet? and are you sure the footswitch is/was wired "east-west" lugs-wise? if you measure the battery voltage as you work the bypass switch, does it drop to zero with one click, then back up with another click?

I have not posted photos yet, but if swapping the pots/jacks doesn't fix it, I will.  I bypassed the 3PDT and still have the same issue, so I've ruled that out.  And yeah, I made sure to wire it with the lugs running horizontally rather than vertically.

[pedalbuilder1]

"Although D1, D2, and C7 felt hot, I realized the heat is actually coming from R14 - the resistor right next to them."

If R14 is getting really hot, it is because you are drawing a lot of current through it. I have seen this before in Rat pedals when the power supply was connected backwards...  in which case a lot of current is flowing through R14 and D3. This is also suspected because the LED will not light up either with the reversed power.

Set your meter to measure DC volts and put the black probe on ground anywhere in the box. Power up the pedal and quickly use the red probe to measure either end of R14. If the meter reads -9v (or close) then you have the power connected backwards. If it reads +9v then the problem is something else.

regards, Jack

Hey Jack, I connected the power and got a reading of +8.6 on the battery side of R14 and 1.1 on the other.  I have replaced the pots, jacks, and LED, as well as rebuild the circuit board three times now from new PCBs and all new components.  I have bypassed the 3PDT as well as the DC jack.  The only thing I have not replaced is the actual wire from the board to the pots.  I know it's unlikely, but do you suppose there is an issue with the PCB layout from GGG?  I'm sure it would have been discovered prior to my project, but I really can't think of anything else.  Here are some photos of my build.  It looks a little messy because it is out of the box.  Also, on the first photo it looks like the GND from the LED is connected to the output jack.  That is simply an illusion caused by the angle of the shot.  Also, the + battery wire and the GND LED wire each have 2 wires soldered together.  That's just left over from when I removed the 3PDT and DC jack. 

I really hope it's something dumb like a backward electrolytic that I messed up three times in a row!  

[Tony Forestiere]

If the jack on the right is the output jack, don't think the LED is supposed to be there. (But I am not that smart). 

[pedalbuilder1]

If the jack on the right is the output jack, don't think the LED is supposed to be there. (But I am not that smart). 

It's not.  It's just an illusion from the camera angle.

[Tony Forestiere]

Sorry. Carry on.