Andromeda (Nobels ODR-1) question about hard clipping diodes

[ryanjdenaro]

A few weeks ago I completed an Andromeda build and recently I wanted to see if I would like it better running two hard clipping diodes in series opposing another set of two in series instead of the standard design, which is just a single set of opposing diodes.  I like the idea of the more open overdrive sound as opposed to the more compressed sound. I wanted to test it first so I removed one side of each diode from the circuit and then wired them to a switch to where I could test either configuration.  I also wired an LED on the third pole of the switch as an indicator. I checked all of the connections for continuity. The funny thing is when I tried switching between the original configuration and the additional diodes I noticed absolutely no difference in sound. Then I just decided to totally disconnect one side of the diodes from the circuit.... Again there was no change in sound, even with the diodes disconnected (still sounded like an overdrive, but exactly the same as before, same volume and everything)!  So I'm guessing it is only doing soft clipping at this point, otherwise I would have heard some difference toggling or removing the hard clipping diodes, right?  I feel like I should have heard least a change in volume when adding/removing the diodes from the circuit, right?

So anyway I'm just a bit puzzled,  and not sure why the hard clipping diodes don't seem to be working.   Anyone have any idea as to what is happening or things I should check? Voltage checkpoints? Etc?  I did buy an SMD JFET (MMBF5457) from Tayda, could that be the problem?  I have heard lots of things about fake JFETs and such, but it's not really clear to me what types of problems might result from that, or if the problem I'm experiencing would be at all related.  As a whole, it does seem to work as an overdrive pedal, but yeah, I don't know what's going on with those diodes.  I also don't have the authentic pedal to compare it to either, so I'm not really sure what it should sound like either.

[PRR]

For reference:
https://aionfx.com/project/andromeda-natural-overdrive/

[ryanjdenaro]

I used the older "Legacy" PCB
Link to the documentation: https://aionfx.com/app/files/docs/andromeda_legacy_documentation.pdf

• I omitted the 2nd 47uF capacitor in the power supply circuit "C24"
• I used the 3.3uF electrolytic capacitor on the output as marked in the BOM list, not the 1uF as marked on the schematic.
• I used a 220uf electrolytic capacitor for the main filter capacitor, per the original circuit design for "C22".  The build doc calls for 100uF.
• I included the Bass knob

Some voltage readings :
power supply (before connecting to pedal): 9.47
Ground to power jack lug after power supply is connected to pedal (pedal off): 9.47

IC1
1. 4.71
2. 4.71
3. 4.64
4. 0
5. 4.68
6. 4.71
7. 4.72
8. 9.47

IC2 (JRC 4588D):
1. 4.71
2. 4.72
3. 4.70
4. 0
5. 4.72
6. 4.72
7. 4.72
8. 9.47

Q1 (MMBF5457, from Tayda, supposedly a "Fairchild/ON Semiconductor" https://www.taydaelectronics.com/mmbf5457-jfet-n-channel-amplifier-transistor-sot-23.html )
Gate: 0
Source: 1.094
Drain: 9.47

Q2 (used 2N5088, although I noticed updated build doc for the 125B enclosures size specifies 2N3904)
E: 9.47
B: 4.183
c: 3.605

D1 (I used 1N4007, build doc calls for 1N4002 in the BOM list and 1N4001 in the schematic)
Cathode: 9.47
Anode: 0

D2
Cathode: 4.71
Anode: 4.71

D3
Cathode:4.71
Anode: 4.71

D4
Cathode: 4.71
Anode: 4.66

D5
Cathode: 4.66
Anode: 4.71

I verified the polarity on the electrolytic capacitors.

[ryanjdenaro]

Another thing I noticed in the newer build doc (https://aionfx.com/app/files/docs/andromeda_documentation.pdf) is that the first paragraph about the clipping diodes (page 6) mentions that the hard clipping diodes are "(signal to ground)"; however, in the schematic it doesn't look like the hard clipping diodes are going signal to ground.  They're going signal to "VA" (or "VB", if you're looking in the newer build doc), which does not measure 0V, when probed in the circuit.  I'm getting 4.71V at VA., or specifically at resistor "27", the pad right next to pin 1 of IC2.

[Vivek]

There are few aspects of the Andromeda Schematic that appear odd to me

For 1, if the output of the soft clipping IC is below one diode drop, the hard clipping diodes that follow won't do much.


Maybe D2 and D3 have to be LED ?


Another interesting point, it appears that the jumper decides on 1 or 2 hard clip diodes in series, but it appears that it's not mentioned in the documentation. A search for "jumper" found only one occurrence

[FlyingWild]

Nobles ODR1 schematic: https://www.experimentalistsanonymous.com/diy/Schematics/Distortion%20Boost%20and%20Overdrive/Nobels%20ODR-1.pdf

The hard clipping diodes do not go to ground they go to the half voltage supply.

Interestingly if you look at the ODR-S version:

https://www.experimentalistsanonymous.com/diy/Schematics/Distortion%20Boost%20and%20Overdrive/Nobels%20ODR-S.pdf

The silicon diodes have been swapped for germanium AA112 which have a lower forward voltage than the 4148. It's been discussed here a bit: https://forum.pedalpcb.com/threads/odr-1.3007/page-2#post-47949

[antonis]

Another thing I noticed in the newer build doc (https://aionfx.com/app/files/docs/andromeda_documentation.pdf) is that the first paragraph about the clipping diodes (page 6) mentions that the hard clipping diodes are "(signal to ground)"; however, in the schematic it doesn't look like the hard clipping diodes are going signal to ground.  They're going signal to "VA" (or "VB", if you're looking in the newer build doc), which does not measure 0V, when probed in the circuit.  I'm getting 4.71V at VA., or specifically at resistor "27", the pad right next to pin 1 of IC2.

DC voltage reading at VA (or VB) doesn't mind 'cause this poin is (or should be) considered AC ground..
(in the mean of signal sees it as 0V point, due to either C22 or C23 cap..)


edit: I was refering to schematic posted by Paul..
As I saw your posted schematic, VA isn't AC ground anymore due to C24 omission..
There is a permanent 7k5 resistance (15k//15k) is series with diodes clipping pair(s)..!!

[FlyingWild]

I think C24 is another capacitor tagged onto after the voltage divider for the Spectrum transistor, it's the same value as C23, I've not seen it in any nobles ODR 1 schematics except for Aion FX version, it's clearer in the latest version of his build docs.

https://aionfx.com/app/files/docs/andromeda_documentation.pdf

[idy]

Ryan's trouble is with the clipping diodes, everything else seems to work. Does an Opamp with feedback-loop diodes output a voltage more that 2 x vf?

If not there is a problem with the design... But this thing is known working.... I like the idea of LEDs in the loop and Si after, or Si and the Ge, but the Aion suggestion is the opposite, doubling the hard clipping diodes. Mysterious.

If you short that node (top of the hard clippers) to VB does your sound cut out? I.e. is that really AC ground? The last response points out that the C left out was redundant, there should still be 47uf to ground on VB. Is there?

[antonis]

If you short that node (top of the hard clippers) to VB does your sound cut out? I.e. is that really AC ground? The last response points out that the C left out was redundant, there should still be 47uf to ground on VB. Is there?

According to OP statement:

• I omitted the 2nd 47uF capacitor in the power supply circuit "C24"

it isn't there..!!

[idy]

[ryanjdenaro]

Ryan's trouble is with the clipping diodes, everything else seems to work. Does an Opamp with feedback-loop diodes output a voltage more that 2 x vf?

If not there is a problem with the design... But this thing is known working.... I like the idea of LEDs in the loop and Si after, or Si and the Ge, but the Aion suggestion is the opposite, doubling the hard clipping diodes. Mysterious.

If you short that node (top of the hard clippers) to VB does your sound cut out? I.e. is that really AC ground? The last response points out that the C left out was redundant, there should still be 47uf to ground on VB. Is there?

All good responses so far, thank you!

idy, yes, I have the 47 UF capacitor to ground (C23) installed.  I also initially figured the second 47 uf capacitor was not necessary and wanted to try the circuit without it first. the build document mentioned something about an attempt to filter low frequency oscillations at higher gain, but also mentioned that it was not in the original design. I kind of wanted to approximate the original design more closely to start.

There are a few things I want to try after reading all of the responses up to this point, but it will have to wait until I have time later tonight. I have a feeling the solution might have something to do with the balance of the forward voltages in the diodes, as mentioned.  But, I may also try installing the 47 uf capacitor on c24 just to see if that changes anything, although I kind of doubt that it will make a difference.  I forgot to specifically mention it but I did in fact use 1N4148 diodes for both the soft and hard clipping.  The build doc mentioned 1N914, but I believe those are supposed to be very close to 1N4148.  I have a selection of diodes some with very low forward voltages (maybe a schottky somewhere in the .2xx range?)  So, I may try that and see if I'm getting any difference in sound.

One other thing I was curious about is the capacitor in the feedback loop. The build document calls for a 120pf, which is what I ordered, but the polystyrene cap that I got measured about 160pf, a little off from the nominal.  I do have a ceramic cap that I measured to be exactly 120pf.  Would that be significantly better?

[ryanjdenaro]

I don't recall if I tested shorting the hard clipping diodes to VA. That's something else I will try.

[antonis]

Short them to any convenient ground point..

[r080]

One other thing I was curious about is the capacitor in the feedback loop. The build document calls for a 120pf, which is what I ordered, but the polystyrene cap that I got measured about 160pf, a little off from the nominal.  I do have a ceramic cap that I measured to be exactly 120pf.  Would that be significantly better?

I wouldn't bother. A lot of caps are 20% tolerance. That is just slightly out of tolerance, and should only matter at high gain settings (slightly less treble).

[ryanjdenaro]

OK tried the following:

Test   Description                                               Result
1   Add 100uF Capacitor per official schematic      No audible difference
2   Try shorting diode connections to VA              Lose all sound
3   Try shorting diode connections to ground        Lose all sound
4   Try 120pf ceramic capacitor, as measured       (didn't try this yet, but may still at some point)
5   Tried 1N5819                                                Large volume loss, lots more more compression
6   Tried 2X 1N5819 in series                              Some volume loss, a bit more compression

So, as FlyingWild had posted the link to the official Nobels ODR-1 schematic, I took a look at that one and compared it with the Aion Electronics schematic.  I saw that there are two capacitors from the VA node to ground: one 47uf and one 100uf; whereas Aion Electronics has 2X 47uf capacitors there.  I decided to give the 100uf a try essentially where C24 would be since that is what was on the original schematic.  The Aion build document states that C24 (47uf) was "added" (page 2, build note 2), which sounded to me like it wasn't part of the original circuit, and thus why I decided to omit it at first.  I did this on a breadboard with clip leads going back to short leads that I had temporarily soldered on the PCB.  There was no audible difference in adding it, which is kind of what I expected.

I tried shorting the hard clipping diode pads together, this resulted in absolute loss of all sound.  Same thing with shorting the diode pads on the signal path to ground, which would be obvious.

I still want to try swapping the polystyrene capacitor that I measured at 160pf with the ceramic one that I measured at 120pf.  If it's true that the 160pf cap may just make it a bit darker, (or have less treble) then perhaps I might like it better, but I figured that's 40pf over or ~33% off from the nominal schematic value.  In my opinion this sounds like a significant amount, which makes me want to test it.  Of course I can't make any final judgments before I've done that.  I'll repost once I get the chance to do that.

The change that actually made an audible difference (as I suspected after reading through some of the first post) was changing the hard clipping diodes to something with a lower forward voltage.  I measured some 1N5819 Schottky diodes with my multimeter on the diode mode at about .18-.19 each, and running two in series yielded measurements of ~.37 to .38 for each set.  This seemed to sound the best, not too much volume loss, and seemed to compress more.  I didn't have many other types that were on the lower end and anything that measured ~.5 and above seemed to not really clip at all (the 1N4148's were like just under .6).  I'm guessing the threshold for this circuit (at least the one I have made up here) is maybe a bit higher than .4, where anything higher than that won't hard clip at all.

Aside from trying the cap that I measured at 120pf, I may just swap the 1N4148 diodes in the hard clipping pads to the double-series 1N5819 diodes, and call it a day.  I considered messing with the soft clipping diodes as well, but I'm not sure I want to really bother since those do seem to work and sound ok.  When the hard clipping diodes are disconnected it does sound fairly open, which I kind of prefer, and I'm not sure I'd want to go any more open sounding to, say, LEDs or anything.

[duck_arse]

I might cop a hiding at the mere suggestion, but, ryan, can you please post some photos of your build? it's just possible we might see something.

[ryanjdenaro]

I don't mind at all.  I usually take a lot of photos of all my builds.

[idy]

You left the 'spectrum' pot off?

[r080]

You left the 'spectrum' pot off?

Documentation shows 2 options for orientation of the spectrum pot depending on whether you are using the bass pot.