Building Guyatone PS-24 distortion : Diode question.

[eh la bas ma]

Hello,

The build layout mentions some germanium diodes v.F 0.300.

I wonder what would be the most logical choice between OA1182 v.F 0.255, or D9E v.F 0.345 ?

[antonis]

IMHO, don't bother too much about forward voltage drop..

As far as I can see (no schematic found), those diodes are wired in series, via a 10k resistor,  with another pair of Si diodes to GND so total clipping threshold also depends on Si diodes v.F and current through 10k resistor..
For Ge diodes of v.F lower than 300mV you could raise 10k resistor value where for higher than 300mV v.F you could  lower 10k resistor value..

P.S.
Weird to see a 10nF cap for "smoothing" Si clipping corners and not a respective one across Ge diodes..

[matopotato]

AA112 might be 0.28-0.29.
But I believe (well, we all know) that Antonis is right.
Other things matter more than slight differences in diodes. IMHO.

[antonis]

Antonis follows Shockley diode equation:

I_D = I_S(e^qV_D/NkT − 1)

where,
I_D = Current through the PN junction
I_S = p-n junction saturation current
e = Euler's number ≈ 2.718281828
q = Electron unit charge, 1.6 ×10⁻¹⁹ coulombs
V_D = Voltage across the PN junction
N = Nonideality/Emission coefficient (between 1 & 2)
k = Boltzmann's constant, 1.38 ×10⁻²³
T = Junction temperature (^oK)

which is simplified to:

I_D = I_S(e^V_D/0.026 − 1)

for N=1 and kT/q = 26mV (at room temperature)

and further simplified to:

I_D = I_S(e^V_D/0.026)

for V_D > 115mV, due to (-1) term less than 1% significance

Transforming the above equation for V_D results into: V_D = 0.026 ln(I_D/I_S)

The forward voltage drop difference for various currents can be fould via: 0.026 ln(I_D2/I_D1)
(e.g. for doubling any given current, the voltage drop difference is 18mV - irrespective of both I_S and actual current level)

[idy]

The other way to think about this question: offhanded descriptions of vF are not worth taking as gospel: that number depends in how tested. Unless you have the same machine (or testing set up) as the other guy, you are blind guys describing the elephant in the dark.

And diodes for clippintg are a matter of taste...or faith...

[Rob Strand]

AFIK that unit is similar to a  Boss HM-2.  The germanium diodes are not clippers but "anti clippers" which act like a noise gate.

Just the same, the same things apply, best to try them.  I doubt you will hear much difference in *tone*.

[matopotato]

Are you able to breadboard to try out, incl the R compensation suggested?
Or socket those parts?
Or perhaps some elaborate 4PDT for both configurations as you are now becomen the master of? 

[antonis]

AFIK that unit is similar to a  Boss HM-2.  The germanium diodes are not clippers but "anti clippers" which act like a noise gate.

Thanx Rob..

On a layout second glance, 1k + Ge + 10k & Si diodes form a "voltage divider", the junction of which goes to IC2 pin 5 via a 150nF cap..

So, for sure, clipping thresholds don't add..

[ElectricDruid]

D1 and D2 are definitely the noise gate Rob mentioned, known as "coring diodes". It's unusual to see applications of crossover distortion, but this is one. Until the signal gets over the Vf of those, there's no background noise. Then after that 10K/D3/D4 is a typical hard-clipper as found in a million places from the DOD250 /RAT/onwards.

Choice of the germanium diodes is a trade-off. If you increase the VF, you increase the noise gate threshold, but also you increase the crossover distortion and chop off more of the note's decay. Smaller Vf gives you less crossover distortion and smoother note decay, but lowers the threshold. Somewhere there's the right balance!

For heavy chugging, it's probably not that critical, since you're not looking for a smooth decay and any crossover distortion will get lost amongst everything else.

HTH

[eh la bas ma]

Thanks for your replies ! I am having some troube to finish this build.

Looks like it's a simplified clone of Dead End Fx's Hive :

http://dirtboxlayouts.blogspot.com/2021/10/guyatone-ps-024-distortion-h.html
Schematics : https://drive.google.com/file/d/1Vs2NACcV8-uVlz0vmkIlmWbyJIM-HoQm/view

I only have a signal in bypass, despite all my efforts.

DC 9V+ : 9.86V

Q1 BC550C, oriented according to layout :

E= 3.713 V
B= 3.954 V
C= 9.59 V

Q2 BC550C

E= 4.47 V
B= 5.02 V
C= 9.59 V

Q3 BC550C

E= 4.49 V
B= 5.06 V
C= 9.59 V

IC1 RC4558P

1= 5.22 V    8= 9.59 V
2= 5.18 V    7= 5.17 V
3= 5.14 V    6= 5.17 V
4= 0 V         5= 5.17 V

IC2 RC4558P

1= 5.16 V     8= 9.59 V
2= 5.18 V     7= 5.19 V
3= 5.15 V     6= 5.19 V
4= 0 V          5= 5.3 V

IC3 M5226P

1= 5.05 V      16= 0 V
2= 4.34 V      15= 5.17 V
3= 5.05 V      14= 9.59 V
4= 4.35 V      13= 5.15 V
5= 5.05 V      12= 5.18 V
6= 4.35 V      11= 5.15 V
7= 5.05 V      10= 4.35 V
8= 4.36 V       9=  5.05 V


All EQ potentiometers's lug 1 are connected together and to "All band 1" pad. Every lug 3 are connected together and to "All band 3" pad.

The Distortion pot's lug 3 is connected to lug 2 with a jumper, lug 1 and 2 are wired to the board as instructed.










I wonder if you can spot something suspicious, on the voltage readings or on the pictures ?

[antonis]

I don't like Q1 V_BE..
(actually, I don't like V_B)

[matopotato]

You are quite serious with the cuts...
Probably nothing but in pic 3, NW corner there are a few cases of frayed wire not cut off.
Sooner or later they can cause issues IMHO.
One of the cuts middleish and south seems to ha e residue copper from the trace going off, but could be fooled by the see-through cuts.
/my five centimes worth...

[eh la bas ma]

I don't like Q1 V_BE..
(actually, I don't like V_B)

I'm not sure what I can do about Q1 voltages, but at least it passes signal on each legs.

I audio probed the circuit, following the layout and DEFX schematics.

It stops after IC1. Near IC1 pad 1, I have a loud signal at the 1uF positive leg, but not on the negative side, nor on the 10k following it. According to Hive schematics, I should have a signal here, right ?






Edit: This cap negative side wasn't on the right row. I have signal at IC2 now. Not at the output jack, unfortunately.

I'll keep probing, and find the next mistake....

[ElectricDruid]

[It stops after IC1.

It's IC3 that I see on the schematic you'Ve posted, but no matter.

Near IC1 pad 1, I have a loud signal at the 1uF positive leg, but not on the negative side, nor on the 10k following it. According to Hive schematics, I should have a signal here, right ?

You're talking about that C9/1uF, right? Yes, you should have signal on both sides of that.

[antonis]

I don't like Q1 V_BE..
(actually, I don't like V_B)

I'm not sure what I can do about Q1 voltages,

Let it be..
Low Base voltage is probably due to DMM loading..

Although it's a terrible schematic, let's have a reference:

[eh la bas ma]

You are quite serious with the cuts...
Probably nothing but in pic 3, NW corner there are a few cases of frayed wire not cut off.
Sooner or later they can cause issues IMHO.
One of the cuts middleish and south seems to ha e residue copper from the trace going off, but could be fooled by the see-through cuts.
/my five centimes worth...

Thanks, these two suspicous aera have been secured.

I always like a good clean cut, to avoid shorts. Shorts are evil. Philonide, greek poet born in -600 said :

"We must cut the evil at its root". I might humbly add : "let's use a small drilling machine to do it, and it will be even more fun"

[antonis]

Philonide, greek poet born in -600

>spam on<
If you're talking about Philonides of Laodicea, he was an Epicurean philosopher and mathematician, born in 200 BCE..
>spam off<

[antonis]

Now, according to above posted schematic, till where you have signal..??

[eh la bas ma]

Philonide, greek poet born in -600

>spam on<
If you're talking about Philonides of Laodicea, he was an Epicurean philosopher and mathematician, born in 200 BCE..
>spam off<

Thank you for this correction.


It's working....

Wouhouuuu ! Ayyaaaaah !


There's a 82n cap at the end of the signal path, it's a bit dodgy on my build, signal comes and goes when i touch it. I'll have to reflow the aera, and see if everything is fine inside the box. Every controls are working... I'll make a build report once i've played a bit with the circuit.

Thank you for your help !

[antonis]

Those MKT (82nF and 5nF) caps aren't much convenient, are they..??
(I bet their voltage rating is higher than 63V)

P.S.
Out of curiosity, I can't get the reason for binding 2 gyrators (L1 & L2) for LOWs, set via a single pot, and use an extra Emitter follower (Q3) for HIGHs ..
Meaning, simultaneously varying 2 bands for LOWs (81Hz & 173Hz) seems to me like an overkill..
(especially when there isn't any single gyrator left availiable..)