Minor debugging needed for 1987 Plexi

[Chuck H]

They actually do smooth changes in current flow a tiny bit. More importantly they roll off very high HF. Something about miller capacitance and the resistors inductance, not even going to try to explain it because I'll get it a little wrong. The frequencies affected will be well above the useful range for a guitar amp and the addition of these resistors helps prevent instability.

[mort]

Ok I've installed the 6.8k resistors and replaced the 220k bias resistor with a 150k. After a little adjustment they've settled to pulling 15 and 19 watts. I'll leave them parked right there for now. The guitar sound is still good.

Now I'm only left with the problem of unwanted noise. I've made a recording of it. At about 15 seconds or so I stick my phone right in the back of the cab(so it sounds loud) and turn op the volumes and the presence control all the way up. With the presence control all the way down there isn't nearly as much wash/hiss. And the 'fluttering' isn't loud but it needs to be dealt with.

[big_teee]

I haven't read through all the posts, but have you cycled all your tubes with known good ones?
On my last build I had a couple of New bad preamp tubes that made some strange sounds.
The new bad tubes were all JJs.
T

[mort]

I will try swapping out the 12ax7's. it has new JJ's.

[mort]

swapped in all the 12ax7's from a jtm45 that doesn't make that noise. noise is still present.

[big_teee]

IME, Most noise issues with Marshall type amps, are due to grounding issues.
Are you using the standard marshall multi ground point grounding?
Or are you using the single Preamp ground point grounding?
Most marshalls are noisy with multi grounding.
I have best luck with only one preamp ground near the Preamp B+ Filter Cap ground.
Similar to this.
http://ceriatone.com/images/layoutPi...exi50-lead.jpg
GL,
T

[mort]

mine's kinda close to that but has some separate grounds. I can easily play with it...


what about that cap to ground from the HT fuse on your layout? Do you know what that cap was meant to solve?

[Chuck H]

Thanks T, for getting us back on track. We sort of side stepped into the design flaws of that particular Marshall model.

[big_teee]

A guess here, but maybe on the same line of the dual .22uf Caps on the HT diodes.
More smoothing effect maybe. I also use a 1-2 meg resistor to grd, in that area to drain the B+ when the amp is turned off.
An Observation, on your Preamp B+ readings.
You may want to experiment with the 2 10k Dropping resistors on the input to the preamp B+.
I like the Voltage on the V1 Plate to be around 200v.
Try taking out one of the 10ks and see if that doesn't clean up the tone.
You can run anything from 10k-20k to drop the B+.
I found on my 2204s that depending on the HT voltage that you can vary that resistor value quite a bit.
I would try this after you get your noise issue fixed. A mean-while tweak.
T

[soundguruman]

I've sorted a few preliminary bugs and have only minor issues left. I've posted a photo of my notes.

Items to discuss:

Bias and dissipation check, how to.

Presence control hums considerably as it's turned up.

Little random pops and sizzles in the sound at higher volumes. (edit: actually sounds more like minor fluttering)

The overall sound is good, bold, and plenty of gain. Controls all work normally.

It's pretty much this circuit with 1K 1W grid resistors and I didn't put those 22uF caps across the secondary of the PT(should I add those??). http://www.drtube.com/schematics/marshall/1987u.gif

A. The major issue here is that there needs to be 1K screen grid resistors on the output tubes.
Otherwise, when cranked up to high output (as a Marshall should be), the output tubes will draw too much current. It will blow the fuses, especially the HT fuse.

The design of EL34 was changed AFTER this amp was designed. Previous older EL34 had much lower screen current.
Newer EL34 has much higher efficiency screen grid, hence the need for current limiting resistor, 1K 5Watt at each screen grid.

Install the 2X 1K 5Watt resistors, just as in this schematic:
marshall JCM800_2204.pdf
and I apologize if the schematic does not attach properly...

Failure to install those 1K resistors can damage the tubes, power transformer, rectifier, etc...now you know.

B. The grounding of the power supply:
The power supply filter caps need to be grounded on the power transformer side of the chassis.
Previously, the bias filter capacitor was grounded to the face plate control ground. This causes HUM in the audio path.

So go ahead and move that bias filter cap ground off of the control ground.
Ground the filter caps with a nut, washer, ring terminal- to the power transformer side of the chassis.

The audio ground - should be attached to the chassis at or near the input jack.

Separating these 2 grounds: the audio ground is separated from the power supply grounds.
The power supply is grounded at one end of the chassis, the audio is grounded at the other side of the chassis.
The resistance of the metal chassis is between the two ground points.

This will decrease the power supply hum in the audio path (considerably).

Star grounding- connecting audio and power supply ground to a single point...
will cause a lot of hum in the audio.
Star grounding is a very bad choice for a tube amp. Lots of extra noise.
Separate these 2 grounds, and you will have much less noise in the audio.

C. Oscillations:
High frequency oscillation will cause a loud 60 cycle hum. You won't hear the high frequency, but you will see it on an O scope.
Depending on the severity, it will also tend to overheat the output tubes, or turn the output tubes RED.

This type of oscillation will typically originate at V1A input stage. Then it will be amplified...
You may also notice a loud HISS in the preamp, or the first preamp tube will have microphonic ringing.

Factory fix: Connect a 6-7 pF 1000V capacitor between grid and plate of V1A. Do this right at the preamp tube socket.
This stops the high frequency oscillation, and will kill the HISS and microphonic ringing of the first preamp tube.

You can use 2X 12 pF 500V in series. Instead of a 1000V single cap.
The factory used ceramic disk capacitor for this.
I use Silver Mica, I just like it better than ceramic.

YES you will find this cap on several versions of Marshall heads, installed by the factory.

[big_teee]

mine's kinda close to that but has some separate grounds. I can easily play with it...


what about that cap to ground from the HT fuse on your layout? Do you know what that cap was meant to solve?

Here is what the Valve Wizard says about the small cap and also adding a small resistor in series.
It however is more of a Bridge Rectifier thing.

"Whenever the diodes switch on and off, as they deliver large pulses of current to the reservoir capacitor, they tend to produce a voltage spike known as a 'switching transient'. This can introduce high-frequency 'hash' noise into the power supply. This can be alleviated to some extent by placing a small resistor in series with the rectifier before the reservoir capacitor. This effectively 'slows down' the rate at whch the diodes switch on and off, smearing the transistion and reducing the switching transient. If too much voltage drop or power supply sag is to be avoided then a small value of say, 10 or 22 ohms, 3W should be sufficient. A small ceramic or poly' shunt capacitor of around 10nF to 1uF may also be added (shown dashed) to reduce the power dissipation in the resistor without spoiling the noise reduction effect too much. This effect is automatically produced by valve rectifiers since they have their own internal anode resistance, which is why they are often decribed as quieter than silicon rectifiers. In most cases though, switching transients are not a problem as they will be swamped by other noises within the amp".

This exerpt is from here, the 3rd figure down the page.
The Valve Wizard

[Chuck H]

A. The major issue here is that there needs to be 1K screen grid resistors on the output tubes.
Otherwise, when cranked up to high output (as a Marshall should be), the output tubes will draw too much current. It will blow the fuses, especially the HT fuse.

The design of EL34 was changed AFTER this amp was designed. Previous older EL34 had much lower screen current.
Newer EL34 has much higher efficiency screen grid, hence the need for current limiting resistor, 1K 5Watt at each screen grid.

Install the 2X 1K 5Watt resistors, just as in this schematic:
[ATTACH]28033[/ATTACH]
and I apologize if the schematic does not attach properly...

Failure to install those 1K resistors can damage the tubes, power transformer, rectifier, etc...now you know.

Hey SGM. You have a habit of reading the first post and answering (often to something other than the question) without reading the other replies. Mort said in post #13 that he did use 1k screen grid resistors.

[mort]

Thanks for all the replies.

I'll talk for a second about how my ground scheme is set up and maybe you can comment on likely corrections...

The two grounding busses are grounded through the end that has the screw, the other end is soldered to an eyelet that's only there for mounting purposes. I made two of them to separate the signal grounds and power grounds. The filter cans are grounded with the PT CT's with the mains ground located right next to it. PT cathodes are grounded right at the sockets, and finally the pots are grounded to themselves.

I was considering grounding the pots to the signal buss. And possibly moving the other buss' grounding point to be closer to the PT end. Anyone see any other possible spots?

[big_teee]

mine's kinda close to that but has some separate grounds. I can easily play with it...


what about that cap to ground from the HT fuse on your layout? Do you know what that cap was meant to solve?

It looks like from your photo that your pots are grounded to the Pot backs.
I would redo that and run a lead to your ground bus on the board.
Tie all preamp grounds together and ground to a chassis screw near the Preamp Caps in the center of the board.
The Power tube grounds are fine.
No pot back grounds. is my advice.
The first 2204 I built, I wore it out trying to get rid of noise, until I did the single point preamp ground.
Now I just build them that way from the start.

[mort]

Ok so I disconnected the pot grounds and connected them to the signal buss. That may have cleared up a little bit of wash but the main noise is still there. I did discover that it intensifies or fades wit the bright channel volume pot. Also, shouldn't the plates of V1 be getting the same amount of voltage? I'm getting 197v to one & 151v to the other.