Self induced oscillation

[John Caswell]

Hi all,
Quite often, when having built something it appears not to perform as the designer specified, with either the voltages not being in the ball park area or the valve getting red even with no signal.
This is often caused by oscillation, usually supersonic and normally only viewable on a 'scope, and is generally triggered off by poor layout/construction techniques and the like.
It can be the very devil to cure, but simple things like keeping control grid wires away from anode wires will effect a cure. In output valves for instance when using them in UL mode a small (100R) resistor wired right on the valve base of the screen grid pin will often stop oscillation dead as will the same remedy on the control grid albeit a slightly larger value. Similarly if strapping a pentode/tetrode to a triode a small 100R resistor between screen and anode will help alleviate any possibility of oscillation
It is also worth noting here that the resistor on the screen grid does not have to be a high wattage rating. Indeed a small wattage rating is to the good, as if the valve goes s/c screen to grid then the resistor will fry protecting the transformer and cathode bias/fixed bias components.
With small signal valves oscillation can often be stopped by fitting a small metal screen across the valve base physically between the grid and anode circuits either earthed or connected to the cathode.

John aka Dr John

[john & Jake the dog]

hi John,
is this just a general observation that you are sharing with us or does it relate to a specific problem?

[bikerhifinut]

Originally Posted by john & Jake the dog

hi John,
is this just a general observation that you are sharing with us or does it relate to a specific problem?

John is trying to give us some useful information.
What you do with it is up to you.
Some no doubt is relevant to your problems.
The advice on grid stoppers is very pertinent.
ok time for bed.

[Richard]

Hi J&J, Yes, it's general info rather than aimed at any particular amp. Very useful too, so we'll copy it to the FAQ as it will eventually get submerged in this section. Rich.

[Mark]

John,

The bit about keeping anode wires away from grid wires caught my attention - I will check the KEL84 layout tonight.

You mention that oscillation may need an oscilloscope to detect. I don't have one, but someone near John might, but they may lack the knowledge of what to look for. How would you go about looking for oscillation using a scope ?

Mark

[john & Jake the dog]

Originally Posted by John Caswell

Hi all,

It can be the very devil to cure, but simple things like keeping control grid wires away from anode wires will effect a cure. In output valves for instance when using them in UL mode a small (100R) resistor wired right on the valve base of the screen grid pin will often stop oscillation dead as will the same remedy on the control grid albeit a slightly larger value. Similarly if strapping a pentode/tetrode to a triode a small 100R resistor between screen and anode will help alleviate any possibility of oscillation
It is also worth noting here that the resistor on the screen grid does not have to be a high wattage rating. Indeed a small wattage rating is to the good, as if the valve goes s/c screen to grid then the resistor will fry protecting the transformer and cathode bias/fixed bias components.
With small signal valves oscillation can often be stopped by fitting a small metal screen across the valve base physically between the grid and anode circuits either earthed or connected to the cathode.

John aka Dr John

hi John and Mark,
amp%20009.jpg
I built my 84 as a newbie and followed the instructions.
I put the pink wire where it should go and I put the purple wire where it goes etc.
Grid wire and circuits and anode wires and circuits do not feature in the instructions.
How is one, as a Newbie supposed to find a fault with a badly laid out anode wire when one hadn't been told and doesn't know what one is?
Techno speak is great for those in the know but it just isolates us lesser mortals.

[david counter]

Originally Posted by john & Jake the dog

I built my 84 as a newbie and followed the instructions.
I put the pink wire where it should go and I put the purple wire where it goes etc.
Grid wire and circuits and anode wires and circuits do not feature in the instructions.
How is one, as a Newbie supposed to find a fault with a badly laid out anode wire when one hadn't been told and doesn't know what one is?
Techno speak is great for those in the know but it just isolates us lesser mortals.

Hi John
as Richard has already said this isn't about the Kel84 which has a pcb,it's about amps in general,

may I suggest that if you want to understand more, get yourself a book, perhaps the very good Morgan Jones one and read up about valve amps

[john & Jake the dog]

Originally Posted by Mark

John,

The bit about keeping anode wires away from grid wires caught my attention - I will check the KEL84 layout tonight.

You mention that oscillation may need an oscilloscope to detect. I don't have one, but someone near John might, but they may lack the knowledge of what to look for. How would you go about looking for oscillation using a scope ?

Mark

Hi David,
Put yourself in my place and tell me what you understand from Marks post.
To me it implies that the kel84 has grid wires and anode wires.
Perhaps I should go and read a book then I wouldn't be a Newbie.

[david counter]

yeah, I can see where you're coming from,

honestly though there's nothing you need to do to the Kel84, there's nothing really wrong with the pcb layout,

[Ianm2]

well I have seen a lot worse builds, and considering that's the first, its pretty good indeed.

There's a 'trick' you can do with the wiring, desolder all the bits running together on the same line, hold it up and pull it tought, together, and put a cable tie every inch, so it forms a big mass, manipulating the bits to still keep it tought, and then route it to all its terminations, that way it will remain slightly pliable, and it will look fully professional. that's how to loom it, takes a bit of practise, but its quite easy really

if you've cut them to short already, it may not work

its 50/50 with the output transformer orientation, but i would be inclined to rotate thro' 180 degrees and have the secondary pointing to the output terminals for neatness, that's nitpicking, but I think important, I am a bit of a perfectionist with wiring ( it may cause possible interaction with the inputs tho', so perhaps that's why it wasn't done on the build)

keep them all at perfect 90 degree angles, left to right, front to back, that's how to route for looks

with the output tx wires, its probably best to keep them initially in the same 'plane' as they exit the transformer, looks like vertical, and easier if someone does the slight tension, whilst someone else ties the ties, then route it to its destination.

looks very nice tho', well done