Park G25R mod time ... please bear with me

[Steben]

So far.....
Settled for NE5532 ... adding an input stage wit fetzer and buffer BJT the output impedance to the IC drops to tens of ohms... Good for such an opamp.
NE5532 should drive the spring tank much better.


The second M5201 stage (before power amp) has 100k resistors (R16-R18-volume pot). Why 100k? Why not smaller?
I sense a kind of urge to use large resistors by most designers. Is this due to habits with tube amps? Large resistors with smaller signal voltages always risks of inducing more noise than necessary.

Yes, Mr noise again. But noise is the most annoying thing in perfectly working amps.

[niektb]

Does a Fetzer not color the sound too much? (especially with high output pickups?)
Assuming that you don't want to alter the sound too much, I'd rather go with a FET emitter follower configuration
I've yet to put it in practice but this buffer I came up with (which is a variation on the Boss Waza buffer) should give you more headroom and higher input and lower output impedance: https://www.diystompboxes.com/smfforum/index.php?topic=125033.msg1188768#msg1188768
note: you might have to tweak the bias resistor values a bit

[Steben]

Does a Fetzer not color the sound too much? (especially with high output pickups?)
Assuming that you don't want to alter the sound too much, I'd rather go with a FET emitter follower configuration
I've yet to put it in practice but this buffer I came up with (which is a variation on the Boss Waza buffer) should give you more headroom and higher input and lower output impedance: https://www.diystompboxes.com/smfforum/index.php?topic=125033.msg1188768#msg1188768
note: you might have to tweak the bias resistor values a bit

Even a BJT buffer can yield high input impedance.
As a matter of fact I want the Fetzer to colour things up. The clean channel is way too sterile.
It is not the only thing I want to add. Planning on swapping R19 (47k) with a 47k master ("power"?) pot just after a power amp saturation sim.

[Steben]

Ok, time for some shopping...

[antonis]

With no offence, I'll suggest you to get a good quality "noiseless" earplugs, Steven..!!
(or get the adress of a noise-pursuer niche therapist..)

P.S.
As in a local island (Creta) people use the verb "hear" for "smell" (e.g. Hey you, hear how this used sock smells..) you "see" noise everywhere..!!

[Steben]

With no offence, I'll suggest you to get a good quality "noiseless" earplugs, Steven..!!
(or get the adress of a noise-pursuer niche therapist..)

P.S.
As in a local island (Creta) people use the verb "hear" for "smell" (e.g. Hey you, hear how this used sock smells..) you "see" noise everywhere..!!

The JFET input is not just noise, it is a triode simulation.
The extra diode is all about drive character.
The power amp sim is al about dynamic crossover.

yeah, I tell a lot of things about noise. I still will do. It is detail. But the essence is here as well.


The diode and some resistors are in place now. Already sounds more dynamic and more useful gain pot range (as always when bringing skyscraper gain down).
(But, uhm, the old amp has some very audible hiss at bedroom level even when no guitar is plugged. )

[Steben]

Video time!
swapped some resistors and added asymmetrical clipping.

I know, some of you guys playing could make this amp sound even better.

[antonis]

The JFET input is not just noise, it is a triode simulation.

So, you're adherent of the trend "Two totally different devices sharing a common characteristic exhibit essentialy same behaviour.."..

[Steben]

The JFET input is not just noise, it is a triode simulation.

So, you're adherent of the trend "Two totally different devices sharing a common characteristic exhibit essentialy same behaviour.."..

Uhm... not really. I just use the Fetzer for the reason it was designed: a solid state input stage with common harmonic content (2/3 law) because of the use of feedback. If a decoupling cap is used at the source for more gain fe, the transfer is square law, more like a pentode.
The clipping of a jFET is not really the same (although in some cases very hard to distinguish), nor is the frequency response.

The power amp sim is somewhat used but slightly differently in Quilter amps. Not a single device, but a circuit.

[Steben]

Start of placing the parts .... never done this via "paint", but it will help

[Steben]

Sag sim part of board v1

[Steben]

Another sim of the "sag control"
This time logarithmic.
Yet I wonder whether linear or log would be better. Log seems a nice sweep here, but the linear control might have more natural feel to it (log would give slight changes in sound impression increment...)



Secondly sag on full and increment in input...

[teemuk]

FYI, term "sag" is used to refer to power supply voltage drop under load. Sag would decrease clipping threshold with sustained distortion with initial attack having more headroom. Dynamic DC bias shifting, and resulting crossover distortion, is an entirely different, and unrelated, phenpmenon.

[Steben]

FYI, term "sag" is used to refer to power supply voltage drop under load. Sag would decrease clipping threshold with sustained distortion with initial attack having more headroom. Dynamic DC bias shifting, and resulting crossover distortion, is an entirely different, and unrelated, phenpmenon.

I know..... "sag" is probably used most for voltage sag, but it does get used for cathode shift as well.... but what happens to an older fixed bias amp with decreased voltage? Response changes.... screen changes....curves change.... bias vs supply changes.... crossover! (even the loss of feedback during clipping will cause suppressed crossover to re appear) Less signal.... result is very comparable to cathode shift .... thats what Quilter realised very early on. Its what peavey does in t dynamics. Check youtube for Quilter explanation. You see clearly the signal on a fixed bias marshall going to crossover cut off. Anyway both the supply drop and the shift cause less effective power.

[anotherjim]

Going back to the input -  does the discrete circuit have an RC power filter? I know the supply has filtering but opamps have good supply noise rejection while single-ended discrete usually does not.
Do the NE5532's have ceramic bypass capacitors on their power pins? Many low-cost mass-market designs don't fit any local opamp bypass.
The DC grounding scheme isn't clear on the schematic. The IEC inlet safety ground is correctly shown directly connected to the chassis - but what about the PCB? I've seen problems where although the pcb layout star distributes the ground, it connected to the chassis at every metal support standoff ( a Park might have those plastic standoff things).

[Steben]

Going back to the input -  does the discrete circuit have an RC power filter? I know the supply has filtering but opamps have good supply noise rejection while single-ended discrete usually does not.
Do the NE5532's have ceramic bypass capacitors on their power pins? Many low-cost mass-market designs don't fit any local opamp bypass.
The DC grounding scheme isn't clear on the schematic. The IEC inlet safety ground is correctly shown directly connected to the chassis - but what about the PCB? I've seen problems where although the pcb layout star distributes the ground, it connected to the chassis at every metal support standoff ( a Park might have those plastic standoff things).

- At the moment input stage design gets power from original 15V supply (with 150R / 100µ filter). You suggest adding an extra RC?
- IC do not have bypass caps. (note: original opamps are 4558's)
- Yes, pcb is put on plastic standoffs...

[Babajoux]

Hey, do you have the original schematic?

[Steben]

[anotherjim]

We lost the thread of this one, didn't we?
How did it turn out?

[Babajoux]

Thank you a lot Steben! I just got this Park and wondered what's inside. Did you do the Sag mod? And what is a fetzer? A jfet booster?