News:

Forum may be experiencing issues.

Main Menu

Any way to make up for volume difference of diodes in the Slow Loris?

Started by jcsul, October 31, 2014, 10:48:36 PM

Previous topic - Next topic

jcsul

Hey guys, first time posting, so sorry if a similar topic has been addressed before.
So anyways, built the Slow Loris a month or so back (first successful build, feltgood.jpg) and since I have been tweaking it to my liking and to grow my knowledge and skill. I dropped in a pair of germanium 1n34a's instead of the stock LED's for on of the  clipping Diode sets. They sound great, however the volume is significantly lower than expected. As in so low I have to crank the volume to about 3 o'clock to keep unity with my amp. Any one have any ideas for a work around in order to squeeze a little extra volume out of the 1n34s? Thanks for any help.
¯\_(ツ)_/¯

midwayfair

There's no make-up gain, so if you clip the signal to the level of Germanium diodes, that's as loud as it gets. If you're using it significantly above unity volume, then your only option without hacking the board is to use a boost after the pedal.

Scruffie

You could use 2 x 1N34A in series per side if it's just the germanium bit you're after and not the clipping threshold.
Works at Lectric-FX

jcsul

Midwayfair: currently I have a compressor after and have just been setting the compression to 0 and using that to boost the volume. I'm trying to refine my skills, so do you have any ideas off-hand for hacking board?

Scruffie: I'll add an extra diode to each side and see how the results turn out. As of now the tone sounds great with a single diode used, resulting in ~symmetrical clipping. Which is odd. To me asymmetrical clipping sounds more "sweet" to me. I'll try adding an extra pair, and hopefully it won't be "too much of a good thing" ha.
¯\_(ツ)_/¯

midwayfair

Quote from: jcsul on November 01, 2014, 08:21:08 AM
Midwayfair: currently I have a compressor after and have just been setting the compression to 0 and using that to boost the volume. I'm trying to refine my skills, so do you have any ideas off-hand for hacking board?

The absolute simplest thing I can think of is to replace the FET with a PNP transistor. Mind the pinout, you'll have to twist some pins. The emitter goes to Vc, and the collector goes where Q1's source is. You'll have to reduce R9 and move the side of R8 that's attached to ground to the junction of R9 and Q1 instead. On both of those resistors, the ground connection is to the south, so you're just desoldering the south side of R8 and turning it toward the north pad of R9.

I'm not sure exactly what to use for R9, but 2.2K is probably more than sufficient.

This might be noisy, however.

What might be easier is to take a little piece of perfboard or vero, two resistors, and a large capacitor (22uF is good enough), and make Q1 a basic JFET amplifier instead of a source follower. Build this on a daughterboard:

(pad for +9V connection)
        |
      2.7K
        |
    Q1 drain->Pad for output
        |
>>Q1 gate
        |
  Q1 source
        |
470R<->22uF
        |
(pad for ground)

Remove R9 from the PCB. The gate connection on the PCB gets connected to the gate on your daughter board. The +9V connection on the daughter board goes to the pad for the drain on the PCB. The pad for the output taken from the drain goes where R9 USED to connect to the source (i.e., the connection to C9, which is still your output cap). Ground goes to the open ground pad at R9.

Your output impedance is now a little high, so leave off R10.

To get a little more output, you could try a 3K or even 3.3k for the drain resistor, but a 2N5457 should settle around 6V with this setup, which will give you ~6dB of gain.

jcsul

Quote from: midwayfair on November 01, 2014, 04:28:41 PM
Quote from: jcsul on November 01, 2014, 08:21:08 AM
Midwayfair: currently I have a compressor after and have just been setting the compression to 0 and using that to boost the volume. I'm trying to refine my skills, so do you have any ideas off-hand for hacking board?

The absolute simplest thing I can think of is to replace the FET with a PNP transistor. Mind the pinout, you'll have to twist some pins. The emitter goes to Vc, and the collector goes where Q1's source is. You'll have to reduce R9 and move the side of R8 that's attached to ground to the junction of R9 and Q1 instead. On both of those resistors, the ground connection is to the south, so you're just desoldering the south side of R8 and turning it toward the north pad of R9.

I'm not sure exactly what to use for R9, but 2.2K is probably more than sufficient.

This might be noisy, however.

What might be easier is to take a little piece of perfboard or vero, two resistors, and a large capacitor (22uF is good enough), and make Q1 a basic JFET amplifier instead of a source follower. Build this on a daughterboard:

(pad for +9V connection)
        |
      2.7K
        |
    Q1 drain->Pad for output
        |
>>Q1 gate
        |
  Q1 source
        |
470R<->22uF
        |
(pad for ground)

Remove R9 from the PCB. The gate connection on the PCB gets connected to the gate on your daughter board. The +9V connection on the daughter board goes to the pad for the drain on the PCB. The pad for the output taken from the drain goes where R9 USED to connect to the source (i.e., the connection to C9, which is still your output cap). Ground goes to the open ground pad at R9.

Your output impedance is now a little high, so leave off R10.

To get a little more output, you could try a 3K or even 3.3k for the drain resistor, but a 2N5457 should settle around 6V with this setup, which will give you ~6dB of gain.

Doesn't sound super hard, just a little tricky making room for the daughter board. The noise could be an issue if I go with the PNP option though, or it could basically result in the same amount of noise. as of right now I have to boost the volume pretty high to keep unity, hence bringing out more of the pedals internal noise. Having to set the volume lower even with increased internal noise may basically just keep the overall amount of noise constant. Thanks for the detailed advice. Neither of this will be super difficult to wire in. I'm excited to see how thins turn out.
¯\_(ツ)_/¯

mgwhit

Quote from: jcsul on November 03, 2014, 05:18:49 PM
Having to set the volume lower even with increased internal noise may basically just keep the overall amount of noise constant.

Or make it noisier.  The Volume control in the Slow Loris is totally passive, so turning it up can't increase the noise in your signal -- it just allows you to hear the noise that is already there.  Introducing any active signal boosting is probably just going to (a.) amplify existing noise, and (b.) potentially add more noise.




jcsul

Quote from: mgwhit on November 03, 2014, 07:27:23 PM
Quote from: jcsul on November 03, 2014, 05:18:49 PM
Having to set the volume lower even with increased internal noise may basically just keep the overall amount of noise constant.

Or make it noisier.  The Volume control in the Slow Loris is totally passive, so turning it up can't increase the noise in your signal -- it just allows you to hear the noise that is already there.  Introducing any active signal boosting is probably just going to (a.) amplify existing noise, and (b.) potentially add more noise.

Hmmm I see. Hadn't considered that at all. At this point, much more noise at all is going make it sound like garbage honestly. I could run the pedal into a noise gate, but I've never met a gate I liked. Any advice on a solution mgwhite? I suppose I could always try a different set of germanium diodes. Like I said, I'm still learning, but a set of diodes with a lower voltage drop would result in more volume and less distortion right?
¯\_(ツ)_/¯

mgwhit

I've never built a RAT with Germanium clipping diodes, so I've never run into the same problem.  I'd recommend going with two Germanium diodes in series or a Germanium diode with a higher forward voltage (OA126 anyone? ;) ).  You won't get the same level of dirt, but you might retain the germanium flavor.  Midwayfair did the definitive thread on clipping diodes here.

One last suggestion might be to try a less-noisy op amp.  Heresy, I know, but if you're really looking to make the Germanium diodes work, that might be the simplest solution since the LM308 is probably the source of the noise in your circuit.  I know I've seen other suggestions, but I can't remember them off the top of my head.  Google around, and if I remember I'll update this.

Update: The NE5534 is the drop in replacement op amp that a lot of people like with the RAT, and it is a lower noise part.  The TL071 is lower noise, too, but if you use it you should remove the compensating capacitor.

jcsul

Apparently those little OA126 bastards are fetching a hefty sum these days. Cheapest I could find on the net was about 7 USD a piece. But I mean, if anyone stumbles across this post and has some they'd like to get rid of on the cheap, hit your boy up ;) haha.

I had a few ideas last night while thinking over the problem that I'll probably give a shot. 1: I bought the diodes in a pack of ten, so if I can test the forward voltage on them I can find which one has the largest forward voltage. I read a write up on Ge diodes last night of a guy saying he measured from 190mv-400mv on his Ge diodes. So I may be able to find some within the higher end of that range, drop them in, and be happy. They have plenty of distortion to spare at the moment, so losing a bit of it won't be an issue for me. Plus, I can always drive my amp a little harder to make up. I dig my amps distortion, so having a touch of that blended in sounds fine to me.
2: what if I were to run a pair in series on either side, or just on one said (haven't tried them in asymmetrical, so I'm already curious to try that.) Again, noob here so correct me if my thinking is off. Wouldn't running them in series result in basically twice the voltage, therefore roughly an increase in volume? Again, if I lose some distortion I'll live.

Pretty interested in trying a new OP amp as well. I'm still a little nervous when dealing with chips and their tiny weak little legs. So changing some around would be good experience, and hearing the subtle tone and response changes would be educational as well.
Thanks to every one who has responded so far,  I appreciate the insight and it's helping me learn.
¯\_(ツ)_/¯

mgwhit

Quote from: jcsul on November 05, 2014, 05:36:09 PM
Apparently those little OA126 bastards are fetching a hefty sum these days. Cheapest I could find on the net was about 7 USD a piece. But I mean, if anyone stumbles across this post and has some they'd like to get rid of on the cheap, hit your boy up ;) haha.

I was joking (a bit) about those OA126s just because there's been such a run hullabaloo about them lately.  I will admit to owning a few and they do have high forward voltages -- almost twice other Germaniums.

Quote from: jcsul on November 05, 2014, 05:36:09 PM
2: what if I were to run a pair in series on either side, or just on one said (haven't tried them in asymmetrical, so I'm already curious to try that.) Again, noob here so correct me if my thinking is off. Wouldn't running them in series result in basically twice the voltage, therefore roughly an increase in volume? Again, if I lose some distortion I'll live.

Yep.  You'll have more headroom before diode clipping starts and, as a result, more volume.

Quote from: jcsul on November 05, 2014, 05:36:09 PM
Pretty interested in trying a new OP amp as well. I'm still a little nervous when dealing with chips and their tiny weak little legs. So changing some around would be good experience, and hearing the subtle tone and response changes would be educational as well.

People go back and forth about whether or not you can hear op amp differences in this (and other) circuit(s).  There was actually a good discussion here a few weeks back about Rats and LM308s.  I leave that for others to argue, but you should definitely get lower noise out of the NE5534.  Be careful swapping ICs out of sockets -- you need to apply upward pressure on both ends simultaneously or you can bend and break pins.  Radio Shack carries a cheap IC extractor tool (basically just bent tweezers) that makes it a no-brainer.

I found the original article I was looking for.  Jack Orman also suggests lowering the input impedance, which you can do by lowering the values of R1 and R3.  (Not sure what he does, but 470K each is as low as I'd go -- and I wouldn't be happy about it.)  The "bias filtering" Jack mentions has already been increased on the Slow Loris (C11).

jcsul

From what I've read those OA126 are very highly regarded. I'd love to try some out and see what kind of sounds they put out. Bummer they've become so pricey. I'm gonna try adding an extra diode or two, see what sort of results that gets me. I'll report back (probably later tonight when I'm off) with the results off that.

I'll have to invest in on of those little tools. Sounds rather helpful. I've had problems in the past when I put an IC into the socket the wrong direction, and broke a leg off halfway up. Managed to salvage the project, but I nearly lost my mind when I snapped half of that leg off during extraction ha. Seems to me than OP amp change is kind of more of a "feel" difference than a "hear" difference. Different slew rates and various other factors seem like they would produce more of a difference in how the pedal responds then how it sounds, it that makes sense. I'm ordering the board and parts for a moodring soon, so I'll add on a 5534 so I can experiment with changing it out. Also, I'll check out the article when I'm off work. Probably has some cool info in it. Thanks again.
¯\_(ツ)_/¯

jcsul

Just an update in case anyone comes across this wondering, adding an extra diode (1n34) in series to one side made some nice changes. A little bit less distortion, and a nice little boost in volume. Not a huge amount, but I'm pretty satisfied with it. The asymmetrical clipping sounds a little nicer too. Just sounds a little more harmonic, if that makes sense. The Ge diodes are more aggressive IMO than the LEDs that the build recommends. Would definitely recommend trying a trio of Ge diodes if you're wanting to make the Slow Loris a little more ballsy, not that it already isn't.
¯\_(ツ)_/¯

mgwhit