DIY Fender Blender 'noise gate' effect problem :(

[auwoodstock86]

Well, just that. I've build a fair number of stompboxes (even a Meatball, if it matters for any reason) and never came across such an elusive problem...

The effect sounds acceptable when the settings are high and play power chords energetically, and the note would keep sounding a fair period, although eventually it fades out noise gate-fashion. But if i play a single string, even with Sustain pot maxed, after some five seconds the note just fades abruptly like a crappy noise gate... even more, if i lower the Sustain level to minimum, the note fades in just like a second or two -in the same way. If i try to play dead notes (i.e. heavy palm muting), the noise is even more present than the note itself... i mean, the attack sounds like pure noise, like when some volumen pot is dirty and scratches while being adjusted.

I also note that the octave effect is not quite apparent in my builds.. maybe as a result of being using a humble Yamaha Pacifica for the testing, but it could also have to do something with the diodes, right?

I tried first using 2n5088, then BC547, then 2N5089 (less noise, for all i know), and nothing. Switched from germanium 1N34As to 1n914s just in case... and nothing. Triple checked resistor values according to FuzzCentral schematic, and checked every individual capacitor with no discrepance, and still nothing. I even checked the diodes using a DCA55 to make sure they were working properly, and thou i found them to be good, the problem persists. I own a CNC machine in which i made the three pcbs i've populated so far for this device, all of which have the same damned problem, and am pretty sure they have no shorts nor whatever of the sort (if any, maybe the tracks of the tone boost circuit are like very long and go around the plate...).

Well, i'd really appreciate any help or hint. Im desperate. Im not a particularly patient person and after several hours of measuring, desolder, resolder, rebuilding and coursing im just about smashing my entire workbech XDDD

Seriously, guys, let me know any idea on this particular. Have a great weekend!

[R.G.]

Unwanted gated sound is nearly always a bias problem.

Have you ever read "what to do when it doesn't work?"

I suggest you start there instead of just swapping parts out.

[dwmorrin]

The effect sounds acceptable when the settings are high and play power chords energetically, and the note would keep sounding a fair period, although eventually it fades out noise gate-fashion. But if i play a single string, even with Sustain pot maxed, after some five seconds the note just fades abruptly like a crappy noise gate...

RG beat me to it, but I would agree that one or more of your transistors is biased incorrectly.
There are 5 transistors, and really any of them working incorrectly will introduce weirdness.
Measure the 3 voltages to ground for each transistor, and if you can't figure it out from there, post all 15 voltages here, and maybe we'll spot something.  Make sure to clearly label each voltage.

I also note that the octave effect is not quite apparent in my builds.. maybe as a result of being using a humble Yamaha Pacifica for the testing, but it could also have to do something with the diodes, right?

The octave effect needs a good balanced output from Q3, and your input needs to resemble "as close" to a sine wave as possible.  For most guitars and guitarists, this happens with neck pickup, somewhere just beyond the 12th fret.
Could have something to do with the diodes, but I would check Q3 first.

[R.G.]

More to the point, if you're subbing in parts and hoping it'll get better, go read and follow ALL of "what to do when it doesn't work."

[auwoodstock86]

First of all, thanks for the time taken to reply. I did read the Debugging thread, but thought i may as well at least introduce my issue before puting up any tech data... but as i started writting, i just got kinda anxious and started to shout my frustration out, hehe. Really sorry u.u

Well, here goes the substance:

1. What does it do, not do, and sound like? = Don't have really anything to add to my last post.
2. Name of the circuit = Fender Blender
3. Source of the circuit (URL of schematic or project) = http://fuzzcentral.ssguitar.com/blender.php and http://www.geofex.com/FX_images/blender.gif
4. Any modifications to the circuit? = But for the true bypass, none that i know of
5. Any parts substitutions? = Yes. I didn't use ceramic caps as suggested by FuzzCentral, but poliester film ones. And at this point i'm using 914 diodes instead of the germanium 34s, which you may remember from my previous post i already tried.

Voltages:

Power supply voltage [no load] = 9.35 [V]  <= Switching, regulated supply
Power supply voltage [plugged in] = 9.34 [V]

Q1
C = 4.730
B = 0.698
E = 0.139
Q2
C = 9.340
B = 4.730
E = 4.140
Q3
C = 6.310
B = 3.592
E = 3.026
Q4
C = 0.791
B = 1.340
E = 0.718
Q5
C = 3.712
B = 1.066
E = 0.463

D1 [one of the two attached to Q4 base]
A = 1.340
K = 0.888
D2 [one of the two attached to Q4 base]
A = 1.340
K = 0.892
D3 [one of the two between C9&C10]
A = 0.001
K = 0.100 [and going up]
D4 [one of the two between C9&C10]
A = 0.123 [and going up]
K = 0.001

Now, i see that the voltage drop between Base and Emitter is some 0.56 Volts in all cases, but the Collector figures vary, though is consistently higher than that of the Base EXCEPT in Q4... guess thats some hint...

[R.G.]

Now, i see that the voltage drop between Base and Emitter is some 0.56 Volts in all cases, but the Collector figures vary, though is consistently higher than that of the Base EXCEPT in Q4... guess thats some hint...

Yep, that's a hint. And this is one of the powers of the "what to do" thread - it forces you to look in places where you might not otherwise think to go. You're on the edge of solving your own problem. Looking at Q4 is what I would advise.

[Mark Hammer]

There's an awful lot of 10uf caps in there.  Any chance one or more may have their polarity reversed?

[auwoodstock86]

There's an awful lot of 10uf caps in there.  Any chance one or more may have their polarity reversed?

There sure is! And I had a fine time straigtening the leads of all 30 i've used so far, since i have a batch some 1000 10uF caps i inherited from an old technician friend to my father, which have odd zig-zag leads with a .3" spacing :s

Well, i made sure the anode was more positive than the cathode in all of them, and are correctly polarized.

Now, for the life of mine, cant figure out what's wrong with my pcb, regarding Q4. Redid all solders and measured continuity and resistance on both the leads and pads (and the connected pads) just to make sure i had no cold solders, and still nothing. R18 is actually 10k as it should, and R17 and R19 have also correct values and are correctly soldered, and also are the two 27k resistors that go to ground from the cathode of the two diodes conected to the base of Q4. Tried as well brushing the whole thing with paint thinner just to wipe out any remaining paste, but the voltage is the same.

Is it valid to just change R18's value to raise the Collector voltage? I've heard of many succesful builds of this circuit with no modification, so it has to be a schematic-to-pcb issue, i guess :/

Gosh, i realy hate to seem (and be!) helpless, but really could use (even) more help D':

[pinkjimiphoton]

see if you have bad solder or maybe a reversed electro somewhere around q4, it's way off on the bias.

look at all your other transistor voltages.

i'm not familiar with this circuit, try replacing the c resistor with a pot and find a value that works well, take the pot out and measure the resistance and put whatever the nearest standard value is, or make a composite to get ya what you need if you need an oddball value.

these things probably had 10% tolerances or so, so say you get it working at best at 17k... you could use a 22k or a 10k and a 6.8k in series or whatever.

[cortezthekiller]

I found this thread which reports voltages from a working build:

http://www.diystompboxes.com/smfforum/index.php?topic=66767.0

As RG mentioned, the collector of Q4 should be higher ( at about 1.75 volts).
The other difference of significance I've noticed is that Q5's Collector should be around 4.75 volts, whereas yours is at 3.7.

Hope this helps.

[dwmorrin]

Simulating this problem, I found that Q4 will saturate with 0.600V drop rectifier diodes, and Q4 will behave nicely with 0.300V or less drop diodes.
So Ge diodes must be used.  Measure the diode drop with your meter before installing.
The diodes work with the 100k to make up Q4's bias.

[auwoodstock86]

Reprise:

First of all, i apologize for having disappeared for half a year without even say thank you. I went abroad on vacation and then got just too involved in college and this problem was too obscure for me, haha. But it was indeed thanks to you all that i found some hot points which i'd like to share with the community.

First, the diodes....

Simulating this problem, I found that Q4 will saturate with 0.600V drop rectifier diodes, and Q4 will behave nicely with 0.300V or less drop diodes.
So Ge diodes must be used.  Measure the diode drop with your meter before installing.
The diodes work with the 100k to make up Q4's bias.

Thanks, dw, i totally owe this one to you. I'd never have thought about the voltage drop of the diodes as i was sure i had bought 1n34as. I don't know if i was scammed or if this is even possible, but the batch of diodes i bought from Polyda2008 at ebay (http://www.ebay.com/usr/polida2008?_trksid=p2047675.l2559) proved to have a forward drop of some 0.64 V, and it wont change heating the diodes up nor with fingers nor with breath... and they do look like germanium ones! (http://www.ebay.com/itm/100PCS-Diode-PANASONIC-ST-DO-7-1N34A-1N34-/260964809138?pt=LH_DefaultDomain_0&hash=item3cc2b6f9b2).

When i replaced them with leaky ac128 transistors (i didn't have any germanium diode on hand) with a B-E drop of 0.34 V, the sound was just amazingly different. All the "gated" cr*p was gone and it was almost usable. The only thing i still felt wierd was some fizzle and treble harshness when at hig levels, so i went on and continued to measure bias and try to find other inconsistencies, and i did found the apparently infamous R5 discrepancy.

According to the schematic i followed (the Fuzz Central one, as found at http://fuzzcentral.ssguitar.com/blender.php), Q1 Emitter resistor is 1k5, whereas in RG's drawings it is 15k. The voltages in each case are as follows (both using proper 1N34As bought the day after from a local dealer, with a drop of 0.32 V):

R5=1k5 Q1 C: 4.480 B: 0.680 E: 0.140 Q2 C: 9.020 B: 4.480 E: 3.900 Q3 C: 6.020 B: 3.560 E: 3.010 Q4 C: 1.760 B: 1.190 E: 0.600 Q5 C: 4.280 B: 0.980 E: 0.390

-

R5=15k Q1 C: 6.970 B: 1.140 E: 0.620 Q2 C: 9.010 B: 6.970 E: 6.390 Q3 C: 6.010 B: 3.550 E: 3.010 Q4 C: 1.550 B: 1.200 E: 0.610 Q5 C: 4.270 B: 0.980 E: 0.390

As for the sound, i think i like it more with the 15k resistor, although it may sound a little bit darker than the other, which at high levels sounds more trebbly and noisy. Do you think Q1's Collector should be necessarily around half of the working voltage?

I tried to tame down the noise by adding ceramic caps in diferente places (i.e., across B-E and C-E in various transistors), but the result was a much darker sound. I may be too fussy, but would like to be able to reach full levels without sounding like a fizzle festival. If can spot or recommend anything, i'll try it ASAP.

Even thou i get no further help with this, i really hope this readings would help someone trying to put together this particular circuit. After all, this is what this forums are all about

Augusto G.