Marty Marts Tornado built, but has a small problem

[John Lyons]

I 've cobbled up Marty's Tornado here and it sounds great! There is only one small problem.
There is what sounds like a leaky cap (sounds like a muffled waterfall). Is this a tell tale sign of a leaky cap?

Two people have build this with no problems or noise so something must be going on here.
I've checked the coupling caps between stages and they all seems fine and don't pass DC voltage.



Here are my voltages with a 9.6 volt supply:
Q1
D 4.5v
S 1.2v
G 0v

Q2
D4.5v
S 1.5V
G 0v

Q3
D 4.2v
S .2v
G 0v

Q4
D 9.6v
S 4.7v
G 4.6v

Q5
D 4.5v
S .5v
G 0v




Q6
D 4.5v
S .5v
G 0v

I don't fully understand the Mu amp section (Q3 and Q4)with respect to voltages but other than that the voltages look OK to me.

I've audio probed it and it seems to be after Q2
Any thoughts on what to check?
Thanks

John

[Meanderthal]

 Yep, I've had that happen twice. Bad electro. Someone described it as a 'rumbling grindcore' sound. It comes and goes, can drive ya nuts.

[John Lyons]

There are only two electrolytic caps in this build 2.2uf and the 100uf power supply decoupling cap.
I used poly caps for the 1ufs and .68s. It would have to be a bad decoupling cap wouldn't it?
So the waterfall/rushing water sound is characteistic of a leaky cap correct?

John

[MartyMart]

John, all the voltages line up with mine quite close, though I have a battery
@ 8.6v so I'm proportionaly lower.
I also have a few circuit changes, this does not seem to change the voltage
readings much though.
Tripple check your Mu-amp section wire up and a re-touch of solder joints
won't hurt.
Constant or intermittant "hash" noise ?
Did you try the 1n cap across trim 3's outer lugs ? ( recommended )
Is your C16 a 1uf poly or an electro ?

MM.

[Papa_lazerous]

I recently had a leaky cap on another project and the sound you are describing is exactly what I had.  Start changing caps....

[MartyMart]

John, could be a cap problem but I've spotted a small change from my build
and both my original schem and the clean copy :
From the Mu amp section, I exit with the 47n - C11 from the top end of R13, the 1k
rather than the lower end of it, so from Q4's source !
Not sure how significant this is as these nodes are connected with that 1k anyway .
Could be an impedance change, exiting source rather that the lower fet's drain ?

MM

[puretube]

what a diff`rence a node makes...

[MartyMart]

what a diff`rence a node makes...

Care to expand Ton ?
Or just a friendly "extraction of urine" .... 

I'm just curious that I may be finally learning something ....

[John Lyons]

I used all film caps except C13 at the source bypass cap, and the C20 100uf across the power supply.
Assuming I have a leaky cap I should be able to read DC voltage after the cap right?
None of the coupling caps. .02, .02. , .047 .033 and 1uf .
The Noise is a lowish contant white noise lind of like if you are tuned into a TV channel no reception with the volume turned high with and your hands over your ears. It has a slight fluctuation of crackle, barley though.

I'll check that R13 thing out as well.

My build is stock except I added that .004 cap across the trim pot at Q5. I thought it cut out too much sparkle so I scaled it back to 500pf.
I'm going to figure out this noise issue and then I'll try some of the modifications

Thanks folks.

John

[puretube]

the TCJ can explain it better in depth, (coz the following is not the complete & whole story, but food for consideration):

just imagine Q4 sitting alone there with its R13 between +9V and "ground" (@ the node of C10 & R13);

imagine a signal coming into the grid gate (i.e.: not shunted to "ground" by C10);

You will find no signal at the ground-node, but almost full signal at the node of source Q4 & R13;

compare with a cathode- or emitter- or source-follower, with R13 as the local feedback resistor...

(these nodes are just 1k apart, but they do make a heck of a difference  )

[MartyMart]

John, this could be a particular "fet" causing this, I've had it with a few of
the Jfet circuits, it either requires a bit of "by ear" biasing at the trimpot
or when the Q is identified a swop out for a quiet one.
If you think that it's after Q2, wont be the Mu, so on the 5/6 Q's try adjusting
the trimpot down to "cutoff" and back to where the noise starts/go's
That may well be it .
Could just mean that 1 "Q" has to sit at 5-5.5v rather than 4.5v
It's worked for me a couple of times.
MM.

[John Lyons]

Well... I replaced all the caps except the pf caps to ground at the gates and the source bypass caps. I wouldn't think these would contribute to the "noise"
I tried biasing the fets with more voltage at the drain but until I got up to supply voltage the rushing noise was still there.
I replaced all the fets to no avail.

Here's a clip with clean, some chords and then the noise at regular volume and then brought up with the recording software to 4 times the level...then some single note stuff with reverb added.

www.mrdwab.com/john/Tornado.mp3

The gain was maxed for the clip and the tone was a about half up.

Unless someone thinks the PF caps and source bypass caps would contribute to the noise I guess this is where it stays.

Marty. When you replaced the cap and resistor with .1 and 82K in the tone control...what would you say is the difference this makes?
Another thing. Since you have modified it with a Second gain control after Q4, you would be attenuating the back ground noise unless you were full up. Pulling anthing after Q with mine and the signal and noise is cut.

John

[MartyMart]

John, I'm puzzled and sorry to hear that you have this problem 
Before my little "tweaks" I didn't have a noise issue and I dont think that I've
reduced the overall gain, just added a pot to vary the gain after the Muamp.
It's a 2m2 pot wired in the "normal" way.
It's still quite "full on" when dimed and is totally noise free !
Can you identify exactly where it appears in the circuit ?

MM.

[John Lyons]

Marty

Mydementia built the same layout as I did and his was apparently fine so the layout is probably fine.
Hopefully I can figure it out as there are still a few components left unturned.

Any thoughts on the tone control changes and the difference it made?
Also why the 2m2 pot instead of a 1M and then 470K/470K divider? already have my enclosure finished so no room for the second gain pot. Although I may put in a trimmer and see how that goes...
Seems that the 2M+ resistance to ground is cancelled by the 470K to ground right after, at half it would be 1M in series and 1 me to ground then 470K in series right after. So 1million 470K in series before Q5!!!!!!  Given that large resistances are a big noise issue it's amazing the circuit isn't really noisy... I used a 820K carbon film resistor but that shouldn't cause that much noise. All the other resistors are metal film...


John

[TELEFUNKON]

the TCJ can explain it better in depth, (coz the following is not the complete & whole story, but food for consideration):

here you go:
http://www.tubecad.com/articles_2002/SRPP_Deconstructed/index.html

[MartyMart]

Any thoughts on the tone control changes and the difference it made?
Also why the 2m2 pot instead of a 1M and then 470K/470K divider? already have my enclosure finished so no room for the second gain pot. Although I may put in a trimmer and see how that goes...
John

The slight TC tweak is subtle and IMO the original works fine.
The 2m2 was a pot that I had around and I wanted to be able to drop the gain a lot
after the Muamp, for some more clean tones, this works quite well and doesn't induce
any noise at all.
Interaction of both gain pots is quite cool.
Keep lookin John, the answer is in that build somewhere !!

MM.

[runmikeyrun]

yep- most likely a leaky cap.  I saved my studio engineer the price of a new Sansamp.  Something was sounding strange and every time you plugged it in you heard whoooooooooshh...

[John Lyons]

I got fed up and etched a new board this afternoon, spent a few hours polulating the baord and wiring it up,  flipped on the amp and guess what.... Same thing!!!!!!!!!!!!!!

I think there must be an error in the schematic, layout or ? I spent a lot of time checking the correct values with a meterwhen I put the parts in this time.
Film caps except for the pfs to ground (which are ceramic) shielded wiring to the switch and in/out. Metal film resistors...

It does sound like a leaky cap but after making two units with new caps each time it's very unlikely to be a leaky cap.
The noise isn't very loud but it's surely there and I trust Marty would know if his was the same and also noisey.

It's after midnight here so I'm not going to look over the layout against the schematic now... Argh!

John

[MartyMart]

OK, so it must be a layout issue, schem's a worker and my vero build is a worker also.
There's a "LOT" more room on my build and power/ground/gain/tone leads are about as far
away from each other as possible !!
Seem like I've created a bit of a gain monster which is sensitive to layout.


MM.

[John Lyons]

 Marty, I checked your hand drawn schematic against Gaussmarkov's redraw and layout.
I only found on thing different. The 120pf across gate and source was drawn gate to ground on Gauss's shem and layout (C2).
That will not make much difference, maybe a little more treble roll off with the 120p gate to ground.

I'll have to take a look at the layout and see about any sketchy areas where critical components might be crossed. Then make a new layout...
FUN FUN FUN!