Finished First DIY amp - ghost note etc.

[fretwizard]

I've scored a Metrix MX55C, and now the voltages are much more accurate than before.

Variac 240vac
All knobs on 0
B+ 365vac
Bias -58v

____pin-1___pin-2___pin-3____pin-4____pin-5____pin-6____pin-7______pin-8____pin-9
V1__175v____0v_____1v________3.0vac___3.0vac___222v____0v________1.7v_____3.1vac
V2__168v____0v_____1.15v_____3.0vac___3.0vac___310v____167v______168v_____3.1vac
V3__228v____26v____42v_______3.0vac___3.0vac___223v____27v_______42v______3.1vac

V4__GND____3.1vac__497v____495v____-58v_____494v_____3.1vac____GND ---------
V5__GND____3.1vac__496v____494v____-58v_____494v_____3.1vac____GND ---------
V6__GND____3.1vac__496v____495v____-58v_____494v_____3.1vac____GND ---------
V7__GND____3.1vac__495v____493v____-58v_____494v_____3.1vac____GND ---------
__________________________________________________________________________________________________________

When biased at 65%
All knobs on 10
Plates: 477v on pin-3
Bias: 34mV

With tubes IN and no pilot light, across the heaters I measure 6.4vac on V7 and 6.1vac on V1.

If I just had the Metrix in the first place

R

[julkke]

Well done man!

[fretwizard]

So it it seems all good in terms of voltages... But...

Now I'm starting to notice ghost notes while I play single notes... especially on the first 3 strings (E-B-G) mostly around 8-13 fret, on both channels.
It is even more obvious when playing with the neck pickup....

I will post some clips when I have time...

Just as a reminder, I'm using this grounding scheme:
http://forum.metroamp.com/viewtopic.php?f=45&t=31328" onclick="window.open(this.href);return false;

The ghost note goes away only when all the tone knobs but the Bass are 0, and I can hear it also when the volume is below 5.
Amp filter caps are all 50+50.
The PI output is twisted very well.

The OT CT is twisted together with the OT primaries to keep tidy, I hope It doesn't cause that...
I've read that the filter-caps are only 1 of the possible causes...

EDIT: I've also read that amps with pentode-triode switch could have more ghosting. I think I tried both...

Screen currents?
Choke?

R

[demonufo]

Ghosting....

Choke is one possibility, but not a very likely one given the high current rating of the ClassicTone that you are using.

Are you 100% certain that it's not wolf-tones from your guitar? Tried another axe?

Amp pics, just in case it's another issue...

[fretwizard]

Very busy at work today. Still haven't tried a different guitar but here's some pics (page 3-4):

http://forum.metroamp.com/viewtopic.php ... 1&start=30" onclick="window.open(this.href);return false;

If you need a closeup of something in particular I can take other pics.
Thanks

R

[fretwizard]

Alright.

I've recorded some clips for the ghost notes.
Strings D-G-B-E. Frets 10 to 13 on each string.

Neck pickup first and then bridge pickup.

1st part of the clip is a 2011 Charvel San Dimas, 2nd part is a 2009 Les Paul Custom.

https://soundcloud.com/overtone85/sets/ ... host-notes" onclick="window.open(this.href);return false;

R

[fretwizard]

I've found this thread about ghosting where Novosibir is suggesting to measure to 2 halves of the PI output.
http://forum.metroamp.com/viewtopic.php?f=3&t=22964" onclick="window.open(this.href);return false;

My Metrix DMM should be TRMS, so it says on the outside
(I don't know if I have to switch anything, I'm just assuming that's the default way it gives AC measurements, tell me if I'm wrong )

So I've generated a 1kHz sinewave with Room Eq Wizard from my Mac and fed it into the amp. I've turned all controls to 10 (bright channel) and presence to 0.
I adjusted the signal until I reached about 35vac RMS on 1 half of the PI, after the coupling cap, on the side with 100k plate resistor.
The side with the 82k resistor measure about 37.5vac RMS. Does it mean the PI is not balance and might contribute to ghosting?

I also tried to short the PI output to see if they were cancelling each other and I could measure about 5vac.

Cheers

R

[fretwizard]

I have measured the 2 halves of the PI again. I've waited for the signal to stabilize and I get 35Vac on the 100k side and 34.5Vac on the 82k side.

If this is close enough to balance the PI, I guess that's no the problem...
I've ordered some F&T caps and I was thinking about forming them according to Larry's Procedure...

R

[fretwizard]

Well...

I've installed the F&T caps yesterday. I've done the 'forming' according to Larry's method... kept them charging for almost 18h. I could measure between 2.5-3.2 Vdc across the 100k resistor after about 6h already but I just kept 'em running overnight.

I tested the amp this morning and the ghosting didn't change . The only cap I haven't changed is the last pre-amp cap on the turret board...

I didn't have resistors with high enough values to balance the PI completely. They should arrive soon and I shall see if balancing the PI closer to 100% is gonna make even a slight difference.

Otherwise I'll just add a spare filter cap on the mains...

Just to make it clear... the ghosting is obvious once the volume is over 4-5, especially with neck pickup...
And my power tubes bias is measuring: 31-25-25-25 mA (I understand within 6mA the tubes are considered matched... )

[demonufo]

Close enough matched on current, but there are other parameters that may have an effect.

Have you actually tried different tubes? (Pre's included)

[fretwizard]

Hi

I tried a different PI, a 12AT7 Genalex Gold Lion, still ghosting. I can try other pre-amp tubes I got but I only have 1 set of power tubes .

By the way... I just realized the pic on the other thread are gone, so, if you still wanna have a look here's the photobucket album link, let me know if it works...:

http://s1302.beta.photobucket.com/user/ ... w%20layout" onclick="window.open(this.href);return false;

I'm pretty sure my ground lugs were pretty snug when I did Larry grounding... plenty of solder there too...

I've found some posts by Toner
here he had to increase mains filtering using the trick Marshallhead suggested, saying the feel didn't change much :
http://forum.metroamp.com/viewtopic.php?f=3&t=9192" onclick="window.open(this.href);return false;

Then 2 years after he says it was bad grounding and he could keep the 50uF:
http://forum.metroamp.com/viewtopic.php?f=3&t=22964" onclick="window.open(this.href);return false;

Then again he says 100uF will decrease ghosting and make the amp stiffer ...
http://forum.metroamp.com/viewtopic.php?f=4&t=36953" onclick="window.open(this.href);return false;

[demonufo]

I don't think ghosting is your issue, from the notes you're describing. Look to the tubes for now. You should have spares for peace of mind anyway.

[fretwizard]

I tried a new set of JJ EL34. This time the difference in the tube bias was only about 3 mV, the hottest being 31mV and the coldest being about 29mV.

Still ghosting.
I guess The quickest test I can do is parallel the mains filtering to get 100uF, and see what happens (hoping B+ doesn't exceed 500v there)...
Otherwise I'm still waiting for some resistors to balance the PI a bit better...

I've noticed the screen grid were all less than 1V lower than the plates, at idle. (478/480V range) I think it was the same with the previous tube set. I've read that at high volume there should be a bigger difference, right?

R

[fretwizard]

It took me a while, but I just installed a Lar/Mar master volume and the ghost definitely stops once the PPIMV is lowered to about 6. It's an audio taper pot too.

Does this narrow the search to anything in particular on the power section? Screen resistors? Screens Filtering?