troubleshooting the easy drive help...

[fuzzy645]

So I'm getting no sound on my easy drive build.  

It seems to me nothing is getting out of the emitter so wondering if that is my problem (please see image below with voltages).  My first thought was that I fried the tranny but I changed it out with the same result (I'm using a socket so it was an easy swap)_

Suggestions??

[Earthscum]

double and triple check the transistor pinout. If I remember correctly, 5089's have different pinouts depending on the manufacturer. It looks like the transistor is upside down, so acting like a reversed biased diode... not conducting.

[fuzzy645]

double and triple check the transistor pinout. If I remember correctly, 5089's have different pinouts depending on the manufacturer. It looks like the transistor is upside down, so acting like a reversed biased diode... not conducting.

Thanks...that seems to have done the trick. I got the 2N5089 from Small Bear but they don't tell you the manufacturer and they don't tell you the pin out.  I originallytried to determine which lead is E, C and B by looking up a data sheet for 2N5089 via Google, but I guess if I don't know the manufacturer that was an exercise in futility  

Anyway, I tried swapping the 2 outer leads and I now get sound and grit  

...now here are my voltages:

Collector Voltage:  8.65 V
Base Voltage .470 V
Emitter Voltage 10.5 mV

Now, I have 2 remaining problems:

1. The 1K gain pot seems kinda useless.  Yes, there is a change in gain, but not much change in tone.  It seems more like a bias control that would be set once via a trim pot rather than a knob a user will play around with.    I was expecting kind of a "distortion thickness" or "distortion gain" but that is not whats happening.

2.  It doesn't sound very good  Its OK, but I at this point I couldn't see using it at a gig, unless it can be tweaked/improved.  Granted I didn't add the volume and tone mod just yet, but still the overall tone does not sound very musical to my ears.

Any ideas?

I was thinking of playing around with the clipping diodes (maybe trying something asymmetrical).   Any other mods you might recommend?

Thanks in advance for your help!

[Enzo]

Really?  I have never seen a TO92 with an "American" part number - 2Nxxxx that was not EBC across the front.   Asian types - 2SAxxxx or 2SCxxxx - would all be ECB across the front.  The 2N and 2SC types are readily interchangeable, but the leg order must be considered when crossing the ocean, so to speak.

Ther is no agreement whatever on the leg order for JFETs though.

[Earthscum]

Ok, now your collector voltage is still high. That is why you aren't getting much of a difference. Try swapping in another transistor or two if you have them, making note of your collector voltage. You should have it sitting about 5V to get the maximum swing. If other transistors have similar voltages at C, try fiddling with the collector resistor. Try a 100k and see which way the voltage goes. Then try 47k. One way or the other will get you there. Remember that, as well (I always forget), that will come in handy when you come across explanations of biasing transistors. I'm more of a jfet guy, they are easier on my brain.

[fuzzy645]

Ok, now your collector voltage is still high. That is why you aren't getting much of a difference. Try swapping in another transistor or two if you have them, making note of your collector voltage. You should have it sitting about 5V to get the maximum swing. If other transistors have similar voltages at C, try fiddling with the collector resistor. Try a 100k and see which way the voltage goes. Then try 47k. One way or the other will get you there. Remember that, as well (I always forget), that will come in handy when you come across explanations of biasing transistors. I'm more of a jfet guy, they are easier on my brain.

Thank you!

I will try both of your recommendations (changing transistors and tweaking that is currently at 68K resistor right by the battery +).

Quick questions:

1.  is 5V a good general rule for collector voltage in most circuits?
2. what are typical voltages I should see at emitter & base too?  My base voltage is .470 V and emitter voltage 10.5 mV?
3. is it a good idea to bias the voltage at the base too?  If so, I guess I would do that by tweaking the 200K resistors in series?

Thanks

[Earthscum]

If you read through THIS THREAD, there are a couple top notch explanations on transistor biasing. If you are running a 9V battery, the ideal collector voltage would be 4.5V to give you the maximum swing without hitting the rail on either side. I use 5V, personally, because you have 1 diode drop and some resistance between ground and collector when it's fully saturated, but general rule is 1/2 VCC.

There are some circuits that just don't bias up like that, though. The Bazz Fuss, for example, spends most of it's time idling in saturation.

[fuzzy645]

If you read through THIS THREAD, there are a couple top notch explanations on transistor biasing. If you are running a 9V battery, the ideal collector voltage would be 4.5V to give you the maximum swing without hitting the rail on either side. I use 5V, personally, because you have 1 diode drop and some resistance between ground and collector when it's fully saturated, but general rule is 1/2 VCC.

There are some circuits that just don't bias up like that, though. The Bazz Fuss, for example, spends most of it's time idling in saturation.

Thanks.  I will read that thread for sure.

FYI  - I swapped transistors (another 5089) and result was same.

I then added an extra 68K resistor in series with the 1st one right off the battery + (so in series so resulting resistance should be 132K)  but that had only a modest change on the voltages (surprisingly to me).  I will next try to swap out the resistor for a smaller value and see what that does.


I was also thinking of adding a voltage divider right off the battery + so the entire circuit would be supplied with only 4.5V.  Is that a good idea, or not??

I think I need to determine if this is either a) a soldering mistake on my part, or b) a design flaw.  I would assume (and hope) Mr. Davisson would have designed the circuit with appropriate collector voltage so I'm wondering if it is possible I have a soldering mistake somewhere. 

Thank you.

[fuzzy645]

OK, cutting R1 in half from 68K (by adding a parallel 68K resistor) had the effect of increasing voltage into the collector. In contrast, doubling it (by adding a series 68K resistor) slightly reduced voltage into the collector.

Is there a voltage divider in this circuit as you look at the schematic.  The spot where R2 and R3 meet certainly looks like one, but most people say it is NOT a voltage divider due to the cap in front of ground. 

Thoughts???

[fuzzy645]

UPDATE -

Great news. It must have been a bad solder joint in my perfboard wiring.  I finally buckled down and redid the entire circuit via a bread board.  Here were my initial new voltages via the breadboard:

C = 3.88 V
B = .6 V
E = 72 mV

Much Better.  I proceeded to plug it in and it sounded MUCH BETTER too

I then looked to get that collector voltage down by tweaking that 68K resistor down.  I ended up putting 2 of of the 68K resistors in parallel resulting in 32K and that brought the collector voltage down to a perfect 5 V!  New voltages:

C = 5 V
B =.654 V
E = 110 mV

I then proceeded to plug it in and it now sounds GREAT!!  The gain knob is even working much better too.

I'm so psyched I will now start swapping diodes and then monkeying around with some tone filters.  

Kid in a candy store.....

[Earthscum]

I waited to reply because I figured you'd pick up on something... not what I was expecting... awesome!

So, just so you know, lowering the resistance also lowered your gain with the pot all the way down, so it will give you a larger range to work with.

Now, try a smaller cap off the wiper, and you can boost just highs to clip to make a thinner sounding drive. With the 22uF and 1k, you are boosting everything from about 7 Hz up. If you used, say, a .22uF, when you turned up the gain, it would attempt to boost and clip everything above about 720Hz, kind of like a Tube Screamer. Everything below 720Hz would stay at the normal lowest gain setting. The frequency shifts, but this should give you a basic "visual" of what the gain knob is actually doing, and another easy mod point. If you want to "tighten up" the distortion, you could use a 2.2uF cap to keep the low-lows from trying to gain and clip. 

[fuzzy645]

Now, try a smaller cap off the wiper, and you can boost just highs to clip to make a thinner sounding drive.

Thanks for your reply.  I will experiment around with different cap values going into the wiper of the gain pot.

A few more questions:

1. Can a Tantalum be substituted for that cap going to the wiper of the gain pot? If so would it sound different from a standard radio shack electrolytic, even if at the same value?   Does that particular cap need to be an electrolytic cap as indicated in the Davisson schematic?

2.  What can be done to make the distortion "thicker" - in other words a bit more like a distortion unit (such as the RAT)  and less like an overdrive (or better yet to switch between the 2)?  I tried 3 different diode combos (all symmetrical) 1N4001's, 1N914's and red 5mm LEDs.   The 1N4001's sounded the best to my ears, but still I'm wondering if there is a way the gain pot can be tuned to "thicken" or "fatten" it up more.

3. Based on what you said, I am assuming the smaller the value of the cap that goes into the wiper of the gain pot, the higher the frequency range that will be boosted.  Is there a formula that can be employed in this case to predict the approximate frequency range?  

4.  Are there any other "tunable" areas of this circuit other than 3 mentioned so far (input resistor by 9V, the clipping diodes, and the 22 uf cap by the gain pot)?

Thank you again!

[Joe]

You can use a tanatlum, not sure if there would be much difference though. Nothing is critical, you can alter virtually ever part.

[jaapie]

The spot where R2 and R3 meet certainly looks like one, but most people say it is NOT a voltage divider due to the cap in front of ground. 

Capacitors and inductors can be part of a voltage divider, too! If you replace one of the resistors with a cap,
the amount of voltage dropped by the cap increases as the frequency drops, until ALL the voltage is dropped at f=0 (or DC). That has the effect of either rolling off high or low frequencies depending on whether the cap or resistor is going to ground-- and that's how you make an RC filter

[Earthscum]

A formula to go by (gives you a rough estimate of where the frequency response is happening) is just your R-C filtering.

Here's a new tweak point for you: The feedback biasing. Basically, at this point, you have a 220k (.01u) 220k. T-filters are calculated similar to normal R-C filters with some considerations. I'd recommend going through Geofex. It is chock full of info, but you will want to look at the sections on Graphic EQ's, Tech of the Wah, etc. I get lost on tangents on that site (R.G.'s site, BTW).

But, for now, the feedback only biases the base because everything above about 144Hz gets dropped to ground through the cap. Here is where you should be able to thicken up things a bit. The feedback is a low-pass filter, so only signals below 144Hz are making it back to the base and getting limited. The rest of the signal will get forced through the clippers. Try doubling (.022), quadrupling (.047) or going a decade bigger (.1u). You will cause the feedback to pass frequencies back to the base at about 66Hz, 30Hz, and 14Hz. If you go smaller by half (.0047), 1/4 (.0022) and a decade (.001), you will start to get to frequencies 300Hz, 650Hz, and 1440Hz.

And, finally, play around with the .1uF at the clipper section. For making it easy to visualize, imagine that while the diodes are conducting, that cap acts like a high-pass in conjunction with the 440k feedback resistance. As is, it appears to pass EVERYTHING, with a corner of about 3.6Hz. There are a whole lot of complex reactances going on with this that I wouldn't even know how to explain, though. The clue here is that you could use an additional cap across the (440k) to shave some of the highs off. Say, something like 150pF would shave off highs at about 2500Hz. 100pF would give more "prescence" at about 3600Hz.

ETA: BTW, Joe... since you still pop in every now and then, just gotta say Thanks a BUNCH for the circuits! The diode compression trick is awesome!

[fuzzy645]

You can use a tanatlum, not sure if there would be much difference though. Nothing is critical, you can alter virtually ever part.

Thanks...will compare in the breadboard for the heck of it just to see and I'll report back.

Capacitors and inductors can be part of a voltage divider, too! If you replace one of the resistors with a cap,
the amount of voltage dropped by the cap increases as the frequency drops, until ALL the voltage is dropped at f=0 (or DC). That has the effect of either rolling off high or low frequencies depending on whether the cap or resistor is going to ground-- and that's how you make an RC filter

Thank you, that makes sense.

A formula to go by (gives you a rough estimate of where the frequency response is happening) is just your R-C filtering.

Here's a new tweak point for you: The feedback biasing. Basically, at this point, you have a 220k (.01u) 220k. T-filters are calculated similar to normal R-C filters with some considerations. I'd recommend going through Geofex. It is chock full of info, but you will want to look at the sections on Graphic EQ's, Tech of the Wah, etc. I get lost on tangents on that site (R.G.'s site, BTW).

But, for now, the feedback only biases the base because everything above about 144Hz gets dropped to ground through the cap. Here is where you should be able to thicken up things a bit. The feedback is a low-pass filter, so only signals below 144Hz are making it back to the base and getting limited. The rest of the signal will get forced through the clippers. Try doubling (.022), quadrupling (.047) or going a decade bigger (.1u). You will cause the feedback to pass frequencies back to the base at about 66Hz, 30Hz, and 14Hz. If you go smaller by half (.0047), 1/4 (.0022) and a decade (.001), you will start to get to frequencies 300Hz, 650Hz, and 1440Hz.

And, finally, play around with the .1uF at the clipper section. For making it easy to visualize, imagine that while the diodes are conducting, that cap acts like a high-pass in conjunction with the 440k feedback resistance. As is, it appears to pass EVERYTHING, with a corner of about 3.6Hz. There are a whole lot of complex reactances going on with this that I wouldn't even know how to explain, though. The clue here is that you could use an additional cap across the (440k) to shave some of the highs off. Say, something like 150pF would shave off highs at about 2500Hz. 100pF would give more "prescence" at about 3600Hz.

Cool.  THat makes a lot of sense. I appreciate you taking the time to explain all this.   I will read up on Geofex about EQs.

...heading to Radio Shack to get some more cap values and I will try plugging into the bread board. I will report back.

BTW, if you know of any article that explains the concept of the feedback I would appreciate that.

Thank you!! 

[Earthscum]

Feedback is fairly easy concept. What we use is NEGATIVE feedback, in op amps, and in transistor circuits. In op amp circuits, the output feeds back into the negative input. What it is doing is feeding a (set by designer) amount of signal back to keep the amount of gain in check.

In transistor circuits, the "self bias" network creates the feedback, so we just use it to our advantage. In a "fixed bias" arrangement, where the resistor goes to the positive rail rather than from the output of the circuit, there is no feedback. In Joe's circuit, the "feedback" mechanism is essentially disabled by not allowing any signal above that 144Hz to feed back into the base, so it is amplified by a larger amount than the portion of the signal that is allowed to "mellow out" the input.

Hopefully that makes sense to ya. Feedback just keeps the amount of gain in check, and you can form different filters to tailor the amount of gain.

[fuzzy645]

Just wanted to update you that I tried some of your suggested mods and they sound killer!!!

I went to radio shack and unfortunately I couldn't find a .22 uf electrolytic.  I was able to find a 10 uf electrolytic and a 2.2 uf  but I didn't realize until I got home that the 2.2 uf was nonpolarized. I was afraid to try the non-polarized 2.2 cap, but I did try 10 uf and it did improve things quite a bit.  

I also swapped out the .01 cap at the feedback biasing section in favor of a .047 uf and boy did this fatten things up nicely!

The one mod you said I didn't understand is what to do at the diode section. I assume you are talking about the .1 uf labeled C3 in the diagram in this thread. Should I replace that cap with a 150 pf, or should I add a cap in parallel with this existing cap?

..oh, and I tweaked the resistor again by the 9V+.  I wanted to bring the collector voltage down a shade, so I ended up taking my 2 series 68k resistors (which is 34k) and adding a 4.7K in series with it resulting in 38.7k.  This changed voltages slightly but I do hear an improvement.

New voltages are:

C =4.92 V
B = .652 V
E = 106.3 mV

I noticed the emitter voltage went down a hair.  Does this mean there will be less play on the gain pot?  Either way the sound is quite good.

Oh, and thanks for the info on feedback.

Thanks!

[Earthscum]

Here's the Quick&Dirty on caps... any cap will work. If you use a polarized cap, you just need to make sure it is plugged in the correct way. Generally 1uF and larger are going to be Elec's, <1uf-1nF are poly, etc. and <1nF-lowest pico's are generally ceramic. A ceramic 2.2uF is huge (now they have some pretty amazing processes that can miniaturize these). A .47 poly is a rolly-polly jelly bean. a 1uF poly cap is a monster.

Don't fret too much, just remember that polarized caps like to be in DC situations, or a situation where  you won't reverse the polarity. Tantalum caps apparently VERY much prefer to have an established polarization and will not tolerate a reverse polarity.

[fuzzy645]

Here's the Quick&Dirty on caps... any cap will work. If you use a polarized cap, you just need to make sure it is plugged in the correct way. Generally 1uF and larger are going to be Elec's, <1uf-1nF are poly, etc. and <1nF-lowest pico's are generally ceramic. A ceramic 2.2uF is huge (now they have some pretty amazing processes that can miniaturize these). A .47 poly is a rolly-polly jelly bean. a 1uF poly cap is a monster.

Don't fret too much, just remember that polarized caps like to be in DC situations, or a situation where  you won't reverse the polarity. Tantalum caps apparently VERY much prefer to have an established polarization and will not tolerate a reverse polarity.

Ahhh, so it doesn't really need to be a polarized cap, and doesn't even need to be an electrolytic at all.  Thank you.  3 more questions:

1.  Is the type of design that Joe has here similar to any known or classic models  (such as Fuzz Face, Tube Screamer, etc...).  I think if is, I would like to look at the corresponding schematic to get some ideas, compare things, and help solidify the concepts I have been learning.

2.  Can I easily swap out the 2N5089 for a germanium if HFE is similar? I see a whole bunch of NOS russian germanium NPNs on ebay.  I assume the current gain is lower, but wondering if something can be tweaked in the circuit to compensate if I decide to experiment in that direction.

3.  I hope I don't cause a "stir" by my next question, but would any of those "fancy" audiophile caps have any affect on tone? I'm not talking cap value now, but rather the matieral (such as paper in oil etc..).  If so, where in the circuit would you get the most "bang for your buck" if trying this out.

Thanks again...you have been most helpful!!  I am learning a lot.