No???
That makes no sense to me.
A magazine is text - like reading the screen you're on right now.
A magazine's authors/contributors are people who you don't know, or know of their qualifications - just like on the internet.
I've been educated professionally, gotten my degree, certified, experienced from 4+ decades of working in shops.
And I've seen enough flaws and blabbering in both platforms, yet you try to tell me they're different?
Hell no! - I'm not one to succumb to and believe everything that I read.
That depends entirely on the circuit, and is often difficult or impossible to calculate. For example, in an op-amp circuit, it has to be low enough not to cause oscillations, but there is usually not enough information available to calculate how many ohms are allowable at what frequency in order to keep everything stable.
In Yoshimitsu's case, low enough not to get noticable parasitic feedback across the output valves. I think that means << 1/(2 gm) of the EL84's.
Edit: wrong, I forgot about the extra RC filters between the supply of the EL84 and the anode resistor of the first stage. My criterion is much stricter than needed.
Last edited:
@wiseoldtech , I do not think you get what I want to say, but that is not important.
@MarcelvdG , Most of the enthusiasts works are not sota...multiple board with wires in between, poor pcb with no power layer plane, mixed ground with signal that cross loops, islands everywhere... no saying most'of us not knopwing really how to use a scope efficiently... At analog level it is certainly less a problem than for digital though...
When I look at my old Philips players with often the ground on the top, mkt radial everywhere, I become nostalgic
@MarcelvdG , Most of the enthusiasts works are not sota...multiple board with wires in between, poor pcb with no power layer plane, mixed ground with signal that cross loops, islands everywhere... no saying most'of us not knopwing really how to use a scope efficiently... At analog level it is certainly less a problem than for digital though...
When I look at my old Philips players with often the ground on the top, mkt radial everywhere, I become nostalgic
4 layers have advantages with modern digital...you at least needs 4 layers todays, I do not talk tube amps.
NONONONONO..... you only need as many layers as you actually need. 4 layers are only there to make a device more compact. A tube amp only needs 1 layer, a DAC typically has only 2 layers.... it depends there is no general rule. The possibility of multiple layers is just something the designer can work with to make a more efficient device - a means to an end.
And idk why we are talking about PCBs now if the original thread was about bypass capacitors in the power supply.
Why ? Because of the inductance and compacity. And also because you can not cross some traces on a 2 layers so you need vias that go to a sub layer...not only compact goal. Also plans help certainly to go near the text book scenario with ground and power on two layer, signal on another plus its ground on another and so on. I know the pcb designers have today expensive tools to probe all what they want but this is not the topic. But a tube amp can be be air wirered as well..no pcb. But we are talking decoupling that means a power source and a load...this is the topic I beleive.
Last edited:
The KISS principle is unbeatable, and it is also the cheapest, thus NO additional bypass, as they have the potential to degrade matters, and practically no potential to improve them.
If the physical distance between the main caps and the point of load is large, you can consider a local bypass, but it has to be done properly and you'll need to give details about your physical layout before any sensible advice can be given
About what (decoupled) supply impedance is allowable. I tried to estimate that for this circuit https://www.diyaudio.com/community/threads/can-i-measure-resistors-in-circuit.385293/post-6998582 but calculated it incorrectly.
It's off topic, but the main advantages of four- or more-layer PCBs are that you have more options to shield things from each other and the fact that for a given characteristic impedance, microstriplines can be much narrower than on a double-layer PCB.
Last edited:
Just try to make a shielded crossing between two wires that aren't allowed to couple to each other on a two-layer board. Which layer are you going to use as shield?
FWIW (I make boards for tube amps, not digital, UHF or otherwise) my boards are all 2 layer - no ground plane. Even then I need some vias here and there. I once made a pair of 4 layer boards, but after to work I was able to make it into 2 layers and save money and time.
Tbh, as I said the only reason someone does multiple layers is because the circuit requires it, to save space for instance. Nobody will do a 4 layer pcb if 1 or 2 layers are enough.
Of course, but you wrote in an earlier post that four layers are only used to make things more compact. That's simply not true.