JHS Angry Charlie - LED Delayed Light Up

[Twhjelmgren28]

Hey all,

I recently bought a JHS Angry Charlie with an LED that wouldn't light.  Circuit works fine other than the LED doesn't work.  I don't have an exact schematic of the circuit but here is a gut shot of what I'm looking at:




So I'm working mainly in the lower left corner of the above-posted image.  It seems to be some sort of millenium bypass-type setup since it uses a transistor (that's my best guess). 

I did message JHS and they recommended I relplace the following in this exact order:

1. Q1 Transistor - Part #:  MMBF170
2. D5 Diode - Part #: BAV70
3. R24 - 3k3
4. Stomp Switch

So I decided to replace the transistor and diode since the R24 resistance was pretty dead-on and the stomp switch had continuity across the poles as expected in both on/on settings.

After replacing Q1 and D5, the LED does light up but it's delayed.  The pedal continues to work flawlessly as soon as I step on the switch.  The only thing is this delayed light-up on the LED.  It's almost like a capacitor is charging up or something...I still haven't changed the LED or stomp switch as I don't want to go down a rabbit hole if it's related to something totally different. 

I'm wondering if maybe I didn't replace with a properly spec'd Q1 or D5?  Here are the mouser part #'s I replaced with:

Q1:  512-MMBF170

D5:  583-BAV70

Any help / thoughts are greatly appreciated - I've also got a message out to JHS to help point me in the right direction. 

Here is a very poor quality video showing the area I'm working in as well as the LED behavior:

Thanks for any and all help!

[antonis]

It looks like a very leaky power supply cap (C8 maybe..??)
Or "something" (like a wrongly placed resistor) forms an RC charge/discharge time constant..    

[Twhjelmgren28]

It looks like a very leaky power supply cap (C8 maybe..??)

I was wondering about C8 myself since it appeared to be coming right off the power supply.  Just to wrap my head around this, the way I understand it is:

C8 should be filtering AC to ground to smooth out / filter noise but should be letting DC do it's own thing (b/c DC shouldn't pass through a capacitor).  But a leaky capacitor means it's holding DC and discharging it to ground.  There is still enough capacitance that the capacitor is eventually "charged up" enough that the correct amount is getting to the LED.  Is that correct?

Or "something" (like a wrongly placed resistor) forms an RC charge/discharge time constant..  

Not sure if this was tongue-in-cheek.  I didn't mess with any resistors or capacitors and I'm guessing JHS didn't manufacture it this way...so, looks like C8 replacement should be my next move.  There is also C20 that appears to be coming off the same trace as C8...

[antonis]

C8 should be filtering AC to ground to smooth out / filter noise but should be letting DC do it's own thing (b/c DC shouldn't pass through a capacitor).  But a leaky capacitor means it's holding DC and discharging it to ground.  There is still enough capacitance that the capacitor is eventually "charged up" enough that the correct amount is getting to the LED.  Is that correct?

More or less, yes..!!

A fully charged capacitor exhibits an infinite resistance between its plates (no DC current flows through itself..)
A fully discharged capacitor is considered a short, hence it draws all the provided from power supply current..
(for an ideal supply source, this is a disaster..!!)
In practice, source's internal resistance forms an RC constant (neglecting cap's ESR) and after 4XRC time-lapse capacitor is concidered 98% charged..
RC time constant

[duck_arse]

I have seen a millenium type thing do that when I used a schottky on the gate to supply instead of the required 1N914/4148. but the named diode isn't schottky, is it. perhaps try remove the bav one and tack a 1N across to test.

[Twhjelmgren28]

I have seen a millenium type thing do that when I used a schottky on the gate to supply instead of the required 1N914/4148. but the named diode isn't schottky, is it. perhaps try remove the bav one and tack a 1N across to test.

Sorry for the delayed response and thanks for the help.  You were definitely on the right track duck_arse.  The problem is fixed.  JHS had initially recommended I use diode BAV70 - as mentioned above, that got the LED working but there was a delay in the LED turning on. 

After posting here, I reached back out to JHS and they indicated that I was perhaps not getting enough leakage in the millenium circuit - they recommended I swap BAV70 diode for MMBD914.  That fixed the problem. 

I do have a couple of questions related to this:

1. I assume the leakage that JHS was referring to was reverse current leakage?  The BAV70 that I was initally advised to use had a lower reverse current spec than the newer MMBD914 diode.  It seems the diodes are all over the place on that spec.  Just trying to wrap my head around what's going on.

2.  Why use a millenium bypass with a 3pdt switch?  I thought the point of millenium was you could have an LED on a DPDT switch...

Thanks again guys!  You 2 have helped me a ton over the years.

[PRR]

newer MMBD914 diode.

(emphasis added)

1N914 is a VERY old diode spec, 1960s.

"MMBD" just means SMD surface mount instead of wire leads.

If this is doing what I think it is, I have no idea why they sold you a Schottky. "Better" is not always better. R.G. (and some microphone designers) use the "old old" 1N914 and 1N4148 diodes because their "flaw" (reverse leakage) fits their plans for pulling FET gates around.

Here we actually want a MINimum leakage but back in 1960s-1970s leakage was a curse to be reduced.

[Rob Strand]

It's the Millenium 2 bypass circuit but without the second GS diode,
http://www.geofex.com/Article_Folders/Millenium/millen.htm

The problem seems to be the original diode leakage is too low to turn on the MOSFET.  That could be because the PCB leakage or switch leakage is shunting the diode leakage current away.   You could clean the PCB but you have no control of what is happening inside the switch.

Schottky diodes have a *much* higher leakage.   That will make the circuit turn-on more readily.     However the Schottky goes against the main idea of the circuit, which is using a low current diode to reduce the switching pops.

A higher current silicon diode might be more of a middle ground but need to find a diode with an actual measured leakage higher than the actual measured leakage of the BAV70 which is also in the same package (that excludes power diodes).

What's not clear, and I haven't looked at other JHS circuits, is the circuit has a three contact switch *and* a Millenium circuit.   The main idea of the Millenium circuit was to use a two contact switch and low leakage is required to reduce pop.  If you have a three contact switch then you can switch the LED directly and there's no need to use a Millenium circuit at all!

[duck_arse]

The Millenium 1 is fascinating to watch - the LED fades in over about 1/2 second. I realized that I was watching the diode leakage current charging the parasitic capacitances of the JFET gate and the wires connecting to it. ......

.... I have found that some silicon diodes are extremely low leakage, in the picoamps, and that these will make the LED turn on even s-l-o-w-e-r. To combat that, simply use a type number 1N914 or 1N4148 diode. These are gold-doped signal diodes; the gold doping speeds them up, but also increases leakage. They are cheap, readily available and all have leakages in the 10 to 15 nA range, ideal for the Millenium 1.

from R.G. himself.

[Twhjelmgren28]

Thank you so much for all of the responses and apologies for the super delayed response.  I've got 2 young kids and always have my hands in a few projects. 

1N914 is a VERY old diode spec, 1960s.

When I referred to "new" vs "old" I meant how I chronologically replaced them in the JHS circuit - apologies for the confusion!

"MMBD" just means SMD surface mount instead of wire leads.

This was a duh moment for me - I can't believe I didn't put 2 and 2 together that the MMBD914 was a 1N914 but in SMD-form.  Makes total sense!

If this is doing what I think it is, I have no idea why they sold you a Schottky. "Better" is not always better. R.G. (and some microphone designers) use the "old old" 1N914 and 1N4148 diodes because their "flaw" (reverse leakage) fits their plans for pulling FET gates around.

Here we actually want a MINimum leakage but back in 1960s-1970s leakage was a curse to be reduced.

If I'm understanding you correctly here, the BAV70 is considered a Schottky - which I assume is also referred to as a switching diode (?); that seems to be what Mouser refers to schottkys as...however, the BAV70 doesn't have enough leakage for this circuit and the 914 was specifically chosen for its high leakage.  In this case we need to have a leakage of at least X in order for the circuit to work (x being whatever minimum value we require)?  Am I getting that? 

Schottky diodes have a *much* higher leakage.  That will make the circuit turn-on more readily.    However the Schottky goes against the main idea of the circuit, which is using a low current diode to reduce the switching pops.

I'm a little confused by this statement - what do you mean by much higher leakage?  The posts from duck_arse and PRR seem to indicate that I needed a diode that had a higher leakage...I may just be confused by wording.  Trying to understand the principle behind this stuff.

The Millenium 1 is fascinating to watch - the LED fades in over about 1/2 second. I realized that I was watching the diode leakage current charging the parasitic capacitances of the JFET gate and the wires connecting to it. ......

.... I have found that some silicon diodes are extremely low leakage, in the picoamps, and that these will make the LED turn on even s-l-o-w-e-r. To combat that, simply use a type number 1N914 or 1N4148 diode. These are gold-doped signal diodes; the gold doping speeds them up, but also increases leakage. They are cheap, readily available and all have leakages in the 10 to 15 nA range, ideal for the Millenium 1.

from R.G. himself.

Super helpful - thanks duck_arse!

[Rob Strand]

I'm a little confused by this statement - what do you mean by much higher leakage?  The posts from duck_arse and PRR seem to indicate that I needed a diode that had a higher leakage...I may just be confused by wording.  Trying to understand the principle behind this stuff.

In "normal" circuits we use resistors to turn devices on, or hold devices off.   The peculiarity of this circuit is it uses diodes in place of resistors.  The diode leakage allows a very small current to flow, like a very large, perhaps impractically large, resistor.

In the original Millenium/Millenium 2 circuit from geofex the currents on the input of the LED circuit are intermingled with the audio signal paths.  The whole motivation of the circuit is to use simpler bypass switches but the side effect is the intermingling of the LED and audio signals.  By keeping the currents low the DC voltages produced in the audio signal path are kept low and that means the LED circuit will not cause click/pops when switching.   So that's where the *need* for low currents comes in.   

Next is getting the right balance of currents so the circuit can turn on and off.  The problem you have is the currents are so low you are getting slow switching times so you need to raise the currents.

One solution is to try to find a silicon diode with a bit more leakage.  That can be a little finicky.   A more drastic way to up the current is a different type of diode.  The next ratchet up on the leakage current scale is a Schottky diode.   Schottky's have higher leakage than Silicon signal diode by nature.

There is a method behind all this.   The main hurdle for DIY'ers they need to understand a bit more stuff they usually don't need to know about.

I think RG explains a lot of stuff in the link I posted earlier,
http://www.geofex.com/Article_Folders/Millenium/millen.htm

[Twhjelmgren28]

I'm a little confused by this statement - what do you mean by much higher leakage?  The posts from duck_arse and PRR seem to indicate that I needed a diode that had a higher leakage...I may just be confused by wording.  Trying to understand the principle behind this stuff.

In "normal" circuits we use resistors to turn devices on, or hold devices off.   The peculiarity of this circuit is it uses diodes in place of resistors.  The diode leakage allows a very small current to flow, like a very large, perhaps impractically large, resistor.

In the original Millenium/Millenium 2 circuit from geofex the currents on the input of the LED circuit are intermingled with the audio signal paths.  The whole motivation of the circuit is to use simpler bypass switches but the side effect is the intermingling of the LED and audio signals.  By keeping the currents low the DC voltages produced in the audio signal path are kept low and that means the LED circuit will not cause click/pops when switching.   So that's where the *need* for low currents comes in.   

Next is getting the right balance of currents so the circuit can turn on and off.  The problem you have is the currents are so low you are getting slow switching times so you need to raise the currents.

One solution is to try to find a silicon diode with a bit more leakage.  That can be a little finicky.   A more drastic way to up the current is a different type of diode.  The next ratchet up on the leakage current scale is a Schottky diode.   Schottky's have higher leakage than Silicon signal diode by nature.

There is a method behind all this.   The main hurdle for DIY'ers they need to understand a bit more stuff they usually don't need to know about.

I think RG explains a lot of stuff in the link I posted earlier,
http://www.geofex.com/Article_Folders/Millenium/millen.htm

Thanks for the detailed response! I get 90% of this. I think my confusion stems from PRR mentioning that i was initially advised to use a Schottky, the BAV70 (or at least that's how i interpreted his post).  But there wasn't high enough leakage (as you also mentioned in your first post).  Which is why i was then advised to use the 914. 

But, if I'm understanding correctly whether or not the BAV70 is a Schottky, it's leakage wasnt high enough.  My hang up was, if the BAV70 is Schottky and schottkys have much higher leakage, then went did the BAV70 not have enough leakage. 

Your second response clarified additional info but i was just confused by my paragraph directly above this one.

[Rob Strand]

But, if I'm understanding correctly whether or not the BAV70 is a Schottky, it's leakage wasnt high enough.  My hang up was, if the BAV70 is Schottky and schottkys have much higher leakage, then went did the BAV70 not have enough leakage. 

I believe the BAV70 is *not* a Schottky.
https://www.diodes.com/assets/Datasheets/BAV70.pdf

The tabulated data could throw you off thinking it's Schottky but the leakage plot (fig 3) shows low leakages in the order of silicon diodes.