bad oscillation

[Leevi]

This amp starts to oscillate with big volume and when the treble pot is
turned to clock position 3 or more. It can be eliminated when turning the full
CUT on. I have noticed that the oscillation level is changing
when I move the shielded wires from PI to the coupling caps in the power amp.
Since it's quite difficult to change the layout I'm asking if there are some
other ways to get rid of it?

http://www.kolumbus.fi/risto.kivioja/RikstoneAmps/PowerAmpC15aReverb.pdf
http://www.kolumbus.fi/risto.kivioja/RikstoneAmps/PreAmpC15aReverb2.pdf

It might also be related to the problem described in this thread
http://www.el34world.com/Forum/index.php?topic=11207.0

/Leevi

[LooseChange]

You need to kill some of the ultrahigh frequencies. Like fender did on the 70's silverface amps.
I have that problem with one of my builds but you REALY need to crank it up and max out the treble. I left it because I'll never play the amp set like that and didn't want to kill the tone when played within reason.

[Leevi]

I added a 220k resistor to signal way and 100k to ground after the tone stack
and got the oscillation disappeared.
/Leevi

[LooseChange]

How much gain did you lose?

[sluckey]

I added a 220k resistor to signal way and 100k to ground after the tone stack and got the oscillation disappeared.

That's equivalent to turning the master volume down a lot.

[HotBluePlates]

Can you post a photo of the assembled amp?

[Leevi]

How much gain did you lose?

There is some loss of course, but it's now more equivalent with Matchess DC30/Lightning.
They use similar voltage drop in their amp. I would say there was too much gain before.

Can you post a photo of the assembled amp?

You can find the picture in the following location (click Rikstone C15aR)

http://www.rikstone.com/

/Leevi

[phsyconoodler]

Did you use any grid stopper resistors on the control grids of the power tubes?If so,what values? Sometimes oscillations can be from this area.
   Also lead dress is a big factor in a crowded layout amp.
You don't show an interior shot of your amp.That would help.

[Leevi]

Did you use any grid stopper resistors on the control grids of the power tubes?If so,what values? Sometimes oscillations can be from this area.

Yes, I'm using 1.5k and I agree the grid stopper in this type of power amp is a must.
I think the removal of oscillation is very challenging if there is too much amplification.

You don't show an interior shot of your amp.That would help.

Check my previous post, there is a link to the amp (also interior shot)
/Leevi

[Geezer]

guts
http://www.kolumbus.fi/risto.kivioja/RikstoneAmps/rikstoneC15aReverb.htm

preamp schem
http://www.kolumbus.fi/risto.kivioja/RikstoneAmps/PreAmpC15aReverb2.pdf

PA schem
http://www.kolumbus.fi/risto.kivioja/RikstoneAmps/PowerAmpC15aReverb.pdf

Power Supply schem
http://www.kolumbus.fi/risto.kivioja/RikstoneAmps/PowerSupplyC15aReverb.pdf

[Leevi]

That's equivalent to turning the master volume down a lot.

Yes, it is. However it's working when the master is by passed. Maybe in this case a better
place for master is after PI?
/Leevi

[Willabe]

Could the speaker mag. be causing a problem? It looks like it's very close to the chassis.


            Brad       

[Leevi]

Could the speaker mag. be causing a problem?

I think in this case not.
I have built plenty of VOX/Matchless based amps where the preamp has very high gain.
Especially the wiring must be planned carefully around the PI. All the longer wires
must be shielded and there are still very often challenges to get rid of the oscillation.
There are even more challenges if you have foot controls and loop jacks on the other side of the chassis.
Matchless is using in their Lightning similar type of voltage drop probably for the same reason.

http://www.freeinfosociety.com/electronics/schematics/audio/matchlesslightning15.pdf

/Leevi

[phsyconoodler]

Quote:"Yes, I'm using 1.5k and I agree the grid stopper in this type of power amp is a must."

The minimum grid stopper on most all EL84 amps is 8.2k. 1.5k grid stoppers are for Fender amps with 6V6's and 6L6's.They are totally different tubes.It may be part of your issue.The rest might be lead dress and use of shielded cables in the right places.
  Too hot of bias will do nasty things too.I wouldn't run those tubes past 13 watts cathode bias and never past 65% fixed bias.
The gain from an EL84 can be too much and backing off on PI values and attenuating preamp circuits is necessary too.
  If you look at a typical EL84 Pi you will notice how it differs from a 6L6/6V6 design.

[Leevi]

phsyconoodler,

The minimum grid stopper on most all EL84 amps is 8.2k.

Traditionally almost all EL84-VOXs (AC30, AC15, AC10) and the copies of VOXs like Matchless Lightning/DC30
are using 1.5k grid stoppers. I would say 1.5k is more popular than any other grid stopper what
comes to EL84 amps.

/Leevi

[phsyconoodler]

That's on original designs yes.But when you alter lead dress and use higher voltages all bets are off.It couldn't hurt to try a larger grid stopper and see what it does.
  Case in point: Fender 5F6A amps have no grid stoppers at all on the original amp schematics but nearly all new builds need them.They oscillate otherwise.I have had to go as high as 50k on some 5F6A builds with no change in layout from one to the next.
  The oscillation came in the form of a distortion through the speaker.
But in my experience most unstable amps are lead-dress related.Your amps look neat,but some wires may cross in odd ways that can set up an oscillation.
  PI wiring is the most prone to this in my opinion.i had a princeton reverb once that howled at certain volumes levels and a move of a couple of PI wires completly eliminated it with no other changes.
  I find i have the most problems when I deviate from known working layouts to fit the amp into odd chassis.Grid wire length is a reall issue.Sometimes a grid wire that is too long will set the amp up for an oscillation.Simply shortening all the grid wires can make a huge difference.I mean ALL the tube grids,not just the power tubes.
  Contrary to popular belief,I have had less issues with amps crammed into small chassis vs amps put into oversize chassis.
 Long wire runs seem to cause me grief.

[jojokeo]

You're reminding me of advice from Weber's books where he says, plate wires can be as long as you need but grid wires must be kept as short as possible due to the sensitive nature of tube grids' sensitivity and how the wire acts like an antenna. Also, as soon as the plate output wire hits a coupling cap, it then becomes part of the next stage's grid wiring. This is why putting all of the components on a tag or eyelet board (ala Fender et al) looks so pretty but electronically and functionality-wise - this isn't wise to do for these exact reasons especially in a high gain amp. You may get away w/ things in normal gain amps but the higher the gain w/ more stages the more important this subject (& technique) becomes.

[Leevi]

I don't use bigger than 1.5k grid stoppers in the VOX based EL84 amps.
Furthermore the power tubes normally operate quite hot. I agree with you that there is a risk for
oscillation which now and then requires extra effort. Small chassis works better but is
not so nice to wire.
/Leevi

[Leevi]

A real point-to-point using terminal strips
/Leevi

This is the guts of a Matchless DC30 custom made by Phil Jamison for Simple Minds guitarist Charlie Burchill. The Matchless point to point system looks wild but it makes for one of the best sounding amps on the market. This is one of three that Charlie uses in his live setup and was just needing a few minor tweaks to get it ready for the Simple Minds tour with the Blue Nile.

[jojokeo]

That looks pretty but also very busy and a bit out of control err lack of orgainization according to layout principles. But, it does illustrate many coupling and tone stack caps soldered directly to their respective pots which likely makes for minimizing grid wire lengths - btwn that and all of the shielded cable runs. He isn't using the philosophy of keeping the filter caps at their respective stages' grounding points though.

That chassis looks just like the demo amp chassis' of what I've seen at the NAMM show in their booth - not a real used amp to be installed inside a "normal" head/cab. Why? Look at all of the 6v bulbs lit up and installed. Why would anyone include all of those for a "normal" giggin' amp unless there was a special plexi front for all of them to light up? But then why all of that for a "giggin amp"? 

[phsyconoodler]

I have seen a few Matchless amps that have those bulbs inside and they were indeed for lighting up a transparent front faceplate.
  That one looks like it a a couple of modded items inside but it's largely Matchless all the way.

You can get away with lots of stuff with point to point amps,but as the previous poster suggested,the grid wires are all really short.