First time building a pedal (LPB-1 clone), and having some issues with it

[Yoliste]

Hi everybody.

This is my first build and I'm having some issues with it. It's an LPB-1 clone. You can find a schematic here : https://hotbottles.files.wordpress.com/2012/02/lpb-schematic.png

Basically the bypass function is working fine (sound is ok when the pedal is off). When I turn it on, the LED lights up as expected, but instead of a clean boost I get a very quiet, weirdly distorted sound. By quiet I mean that even with the pot on the maximum setting the gain is way below unity.

Since I used a stripboard I hand drawn a complete schematic (with input, output and DC jacks, footswitch and LED). I uploaded this schematic here : http://imgur.com/a/I2eHO

Now here are some information that hopefully will help :

• I replaced both caps with 120nF since I did not have 100nF on hand.
• I initially wired the transistor in the wrong way, but unsoldering it and wiring it in the other way did not seem to change much.
• I added a wire between the back of the pot and ground (lug number 1).
• After some checks, I realized that the Connect pin of the DC jack seemed to short to ground, but not the Sleeve one, so I used the sleeve only.

Also here are some values that I measured when checking connections. Those measurement were done with the DC plugged in, and with the pedal switched on :

• DC sleeve to ground : 9.4V
• Collector pin of the transistor to ground : .43V
• Base pin to ground : .97V
• Emitter to ground : .34V
• Circuit input to ground : 15mV
• lug 2 or 3 of the pot, and both jack tips to ground : 0V.

Here are also some pictures of the build, although they are not of very good quality since I only had my phone to take them : http://imgur.com/a/YvAf1

So, does anyone have an idea of what can be wrong with it ?

[thermionix]

I don't see a 100k resistor on your board.

[GibsonGM]

The collector should be a darn sight higher than .43V!   Something is wrong between "+", the collector resistor, and your collector.

[Kipper4]

I see a switch wiring that won't work.

[Kipper4]

Google tonepad off board wiring 5.

[Yoliste]

Wow, that's a lot of replies, thank you very much.

I'm feeling very stupid right now. I must have checked that maybe 5 times but somehow it went over my head that I had forgotten a resistor... Anyway, it seems like the most obvious issue right now so I'm going to try it first. It also seems like I ordered the parts with an other schematic in mind that uses 430K and 43K resistors instead of 1M/100k so I'm gonna try this since I actually don't have any 100k resistor.

@Kipper4 can you expand on the issues with my switch wiring ? I followed this tutorial : http://www.madbeanpedals.com/tutorials/downloads/MBP_FootswitchWiring.pdf

Anyway, I'm gonna fix this resistor issue first and get back to you about the result. Again, thank you very much for your answers !

[GibsonGM]

What I do is build a pedal and omit the switch, so I have "In" and "Out" wires.   These get jumpered to their respective jacks. The ground portion of the jack just gets jumpered to board ground, or the battery "-".     So I can try the PCB out and assure that it works.

Then I wire in the switch, DC jack etc.  So I will know if there's a problem, most likely in the switch.

I'd recommend you do that here if you can!  Even if you have to temp. solder the jacks to the PCB it may be worth it.

For after, this is what Kipper wants you to look at.
http://www.tonepad.com/getfileinfo.asp?id=76

[thermionix]

The collector should be a darn sight higher than .43V!   Something is wrong between "+", the collector resistor, and your collector.

Question, for my own benefit since I'm still learning this stuff.  I know in common tube circuits, if you omit the grid-to-ground load, the tube will go into runaway and presumably burn up and die (never tried it, just read about it).  Can you get a similar result with a transistor that has no base-to-ground load?  Like in this case, maybe omitting the 100k is causing excessive current at the collector, and too much voltage drop across the 10k Rc?  Sorry if that's a dumb question, or doesn't make any sense.  Also sorry for thread hijack.  Just curious.

[Yoliste]

Okay, back from my second round of testing, and there has been some progress ! Now when I turn the knob the volume does get louder than unity, but the sound is still distorted. I measured again the voltages on the pins of the transistor :

C : 4.05V (much higher than before, as GibsonGM mentioned earlier)
B : .78V
E : .20V

As for the switch wiring, I did see the document that Kipper pointed me to, but I still fail to see an issue with the wiring I used :/ Can you enlighten me ?

Apart from that, wiring the pedal without the switch seems like a great idea, I'll definitely keep that in mind. Do you think it may be the problem here ? (especially now that I do get sound out of it)

[GibsonGM]

The collector should be a darn sight higher than .43V!   Something is wrong between "+", the collector resistor, and your collector.

Question, for my own benefit since I'm still learning this stuff.  I know in common tube circuits, if you omit the grid-to-ground load, the tube will go into runaway and presumably burn up and die (never tried it, just read about it).  Can you get a similar result with a transistor that has no base-to-ground load?  Like in this case, maybe omitting the 100k is causing excessive current at the collector, and too much voltage drop across the 10k Rc?  Sorry if that's a dumb question, or doesn't make any sense.  Also sorry for thread hijack.  Just curious.

In this case I don't think so. With tubes, the grid will begin to go positive....the grid stopper basically allows electrons to flow and replenish the supply, if you will.   I know of no transistor equivalent to this.     Now, the pulldown DOES prevent voltage from building up, and keeps things stable.

OK yoliste, you are getting closer!  If you have glitches (and you do) I would lose the switch for now, yes.....

Your voltages are nearly textbook for the LPB.   The collector seems a little low, but that probably is not too worrisome (would expect more like 5V).  So be SURE you have the right value collector resistor and that everything is OK up to 9V.

When you say 'wired the back of the pot to gnd'...is one lug of the pot actually connected to ground? (the pot's case is not connected to the lugs...)  In guitars they fold a lug over and solder it TO the pot back, if they do this.  Not really great for stompboxes, but not something we'd shoot you for, either
In any case, one lug DOES go to gnd, and must go to gnd for the pot to work....

[Yoliste]

The collector should be a darn sight higher than .43V!   Something is wrong between "+", the collector resistor, and your collector.

Question, for my own benefit since I'm still learning this stuff.  I know in common tube circuits, if you omit the grid-to-ground load, the tube will go into runaway and presumably burn up and die (never tried it, just read about it).  Can you get a similar result with a transistor that has no base-to-ground load?  Like in this case, maybe omitting the 100k is causing excessive current at the collector, and too much voltage drop across the 10k Rc?  Sorry if that's a dumb question, or doesn't make any sense.  Also sorry for thread hijack.  Just curious.

In this case I don't think so. With tubes, the grid will begin to go positive....the grid stopper basically allows electrons to flow and replenish the supply, if you will.   I know of no transistor equivalent to this.     Now, the pulldown DOES prevent voltage from building up, and keeps things stable.

OK yoliste, you are getting closer!  If you have glitches (and you do) I would lose the switch for now, yes.....

Your voltages are nearly textbook for the LPB.   The collector seems a little low, but that probably is not too worrisome (would expect more like 5V).  So be SURE you have the right value collector resistor and that everything is OK up to 9V.

When you say 'wired the back of the pot to gnd'...is one lug of the pot actually connected to ground? (the pot's case is not connected to the lugs...)  In guitars they fold a lug over and solder it TO the pot back, if they do this.  Not really great for stompboxes, but not something we'd shoot you for, either
In any case, one lug DOES go to gnd, and must go to gnd for the pot to work....

Yes, one of the lugs is connected to ground, and that same lug is connected to the back of the pot with another wire (a habit I indeed got when fiddling with guitar wiring). Although that's far from being the main concern here, do you know why it's uncommon to do that with pedals ?

I also just double-checked every resistor (and checked the collector one once more) just to make sure I had not anything mixed up and I'm quite confident that all values are the right ones. Only difference is that the base resistors are now 430k and 43k instead of 1M and 100k. I guess the next step is to try without the switch just to be sure.

Also, do you think that having spent some time with the transistor wired backwards at the beginning, and then some more time with the missing resistor could have blown it or something ?

Again, thank you very much to all of you for your time and your answers !

[Yoliste]

Okay, so I bypassed the switch entirely, and there is no difference ! The upside of this is that at least the switch seems to be correctly wired\o/

Another thing that I noticed : if I lower the volume on my guitar to about 60% then the pedal doesn't distort (and still acts as a clean boost as intended). So it seems that the transistor is indeed saturating. Reading from here : http://www.coda-effects.com/p/lpb1-circuit-analysis.html it seems like I could decrease the gain of the transistor with either reducing the collector resistor, or increasing the emitter one. Do you think it may be a good idea ?

[PRR]

After tedious analysis, I concur you lost a 100K resistor. Sometimes the eye is quicker than the brain.

> seems that the transistor is indeed saturating

It has a gain near 20. Max output is under 3V. 3V/20 is 0.15V maximum input. This is a large but not huge guitar level. It WILL distort a hot guitar. I think it always did. The "Linear" in the name is not to be taken real literally. And its clear intent is to slam the snot out of your amplifier. If it also distorts itself, that's the frosting on the cake, no?

Turn down (at the guitar) for clean. Turn-UP for LOUD.

> reducing the collector resistor, or increasing the emitter one.

Either one screws with the happy DC bias you recently found.

You generally select plate/collector resistor for the external load. 10K is a fine value here. Larger Re will reduce gain, true. But Re with base voltage also sets the collector-emitter current and thus the the DC voltage at the collector. Make Re large, current drops, collector voltage soars too close to 9V. Raise the 100K until Vc comes down. You can go round-and-round a few times. But because this plan is barely-stabilized, it may drift out of whack over time.

Ah. I don't think I have seen this approach, and have not tried it. But it will cut gain and for small changes will not be too badly biased?

https://s24.postimg.org/veju7bgn9/LPB_clone_mod.gif

[Yoliste]

After tedious analysis, I concur you lost a 100K resistor. Sometimes the eye is quicker than the brain.

> seems that the transistor is indeed saturating

It has a gain near 20. Max output is under 3V. 3V/20 is 0.15V maximum input. This is a large but not huge guitar level. It WILL distort a hot guitar. I think it always did. The "Linear" in the name is not to be taken real literally. And its clear intent is to slam the snot out of your amplifier. If it also distorts itself, that's the frosting on the cake, no?

Turn down (at the guitar) for clean. Turn-UP for LOUD.

> reducing the collector resistor, or increasing the emitter one.

Either one screws with the happy DC bias you recently found.

You generally select plate/collector resistor for the external load. 10K is a fine value here. Larger Re will reduce gain, true. But Re with base voltage also sets the collector-emitter current and thus the the DC voltage at the collector. Make Re large, current drops, collector voltage soars too close to 9V. Raise the 100K until Vc comes down. You can go round-and-round a few times. But because this plan is barely-stabilized, it may drift out of whack over time.

Ah. I don't think I have seen this approach, and have not tried it. But it will cut gain and for small changes will not be too badly biased?

https://s24.postimg.org/veju7bgn9/LPB_clone_mod.gif

Hi, and thank you for your answer. As you can see in the other posts, I did correct the missing resistor mistake and it now works much better. Also thank you for the additionnal information about this circuit, I didn't know that it was known to distort with higher gain guitars. Mine has a DiMarzio Tone Zone which is indeed quite hot, hence the distortion I guess.

Since the last replies I tried adding a 43k resistor in parallel to the 10k one (that should be equivalent to a 8k one) and it distorts less. Can you explain me what is wrong with modifying the value of the collector resistor ?
Edit : just measured the voltages again and the collector voltage is now around 5V, I think I understood

I think I'm also going to try the mod that you linked, thank you !

[thermionix]

The collector should be a darn sight higher than .43V!   Something is wrong between "+", the collector resistor, and your collector.

Question, for my own benefit since I'm still learning this stuff.  I know in common tube circuits, if you omit the grid-to-ground load, the tube will go into runaway and presumably burn up and die (never tried it, just read about it).  Can you get a similar result with a transistor that has no base-to-ground load?  Like in this case, maybe omitting the 100k is causing excessive current at the collector, and too much voltage drop across the 10k Rc?  Sorry if that's a dumb question, or doesn't make any sense.  Also sorry for thread hijack.  Just curious.

In this case I don't think so. With tubes, the grid will begin to go positive....the grid stopper basically allows electrons to flow and replenish the supply, if you will.   I know of no transistor equivalent to this.     Now, the pulldown DOES prevent voltage from building up, and keeps things stable.

I'm a bit lost on how the pulldown has any effect on DC here, being on the other side of the input cap, but that's maybe a different lesson.  In OP's situation, adding that 100k from base to ground did bring his collector voltage up to normal level, right?  Or was there another change that I missed?  I'm not trying to argue, if it seems that way, just trying to understand.

[Kipper4]

Ok I concede I missread your diagram the switching may work. sorry.

[GibsonGM]

I'm a bit lost on how the pulldown has any effect on DC here, being on the other side of the input cap, but that's maybe a different lesson.  In OP's situation, adding that 100k from base to ground did bring his collector voltage up to normal level, right?  Or was there another change that I missed?  I'm not trying to argue, if it seems that way, just trying to understand.

I'm working from the LPB-2 schematic by JD Sleep from years back.  It has a 4.7M pulldown on the input side of the input cap (R1 on the schem. OP linked to), which is where my thoughts were with that.  That would of course only deal with DC coming in from a faulty effect ahead of the LPB...since he had the 'wrong' value in the divider (provided by the outside schematic), I allowed myself to slide over to the other side of the cap, mentally      At that stage of the 'investigation' I was not sure if C1 could be leaky.     Missing that 100k was much more important tho....

Ok, you likely know the 100k is part of the bias stabilization network.  Simply a voltage divider to set up the base of the transistor for [somewhat] linear operation.   The original was NOT 100k, as Yoliste discovered and changed back to 43k with 430k on top.   If this resistor is missing, the voltage divider is not complete and the base doesn't get proper bias, that's all.  So when he added it, it worked.  Then he brought values back to original. 

That network also stabilizes the transistor, providing feedback to prevent  thermal runaway ("Compensation")....It's not actually a pulldown as we'd define that, it's just part of that network. 

I confused the 2 things and should have hit ENTER 2x before my last sentence about the pulldown, ha ha! 

[Yoliste]

Ok I concede I missread your diagram the switching may work. sorry.

No worries, I'm aware that my hand drawn schematics may not be the clearest and may not follow the standards as I initially just made them for myself !

As for my circuit, I'm now trying out different resistor values. Right now I have a 1.1k emittor resistor (2 2.2K in parallel as that was all I had xD), and a 53K base -> ground resistor (R4 on the original schematic). I think I'm not too far from having the right values (It has a very very slight distortion, and the base voltage is at about 5.7V which if I understood well is a bit high).

Again, big thanks to all of you for your replies and your time, this forum is awesome !

[GibsonGM]

Hi Yoliste, enjoy...5.7V on the COLLECTOR?   Yes, a little high, but depends what you are after.   If you center-bias, you get the most 'swing' in either direction. 

But as it was pointed out, LPB is meant to be a little dirty anyway, so asymmetric bias is ok.   Feel free to mess around with values, it's YOUR project!

Do some reading about load lines some time - used for tubes and transistors (applies to opamps too).  It will give you a better understanding of what is going on with those resistors, bias points, and how the transistor works.