Debug: Thor, too low on gain?

[Auke Haarsma]

've build the Thor from ROG. Etched their PCB, used all standard value's. Only replaced 1k r's with 1k2 r's (didn't have 1k's around).  But that shouldn't be a major issue, I think/guess/hope. I also didn't include the switches yet, so they are both 'open' (again, if I understand the schematic correctly that shouldn't be a problem).

Does it work? Yes! But the gain is way too low. It's not at all like the soundsamples posted by MM elsewhere at this board. However, when I crank the input by using a booster in front (in my case a MXR Micro Amp at full gain) and put the Thor on max I get a nice, rich, warm tone. So my guess is the gain at some of the Q's is wrong. I have biased the Q's correctly (I think...) and did swap them around (I have a total of 5 2n5457's around). Also swapped the j201 a couple of times to get the bias right.

Here's my values

PS: 9.5v

Q1
g: 0
s: 0.47
d: 6.41

Q2
g: 0
s: 0.47
d: 4.51

Q3
g: 0
s: 0.32
d: 4.96

Q4
g: 2.37
s: 5.29
d: 9.51

IC1 (tl071)
1: 0. 15
2: 4.74
3: 4.31
4: 0
5: 0.15
6: 4.75
7: 9.51
8: 0

Audioprobe shows signal ok at Q1, Q2 and Q4. But very gated at Q3. Would that indicate Q3 as the faulty one here? (tried several 2n5457's in there).

Thanks for any advice!

[mdh]

Q4
g: 2.37
s: 5.29
d: 9.51

That looks suspicious to me... full supply voltage on the drain of Q4? Maybe look for a solder bridge or something?

[Auke Haarsma]

From MM's reply (#2) in this thread: http://www.diystompboxes.com/smfforum/index.php?topic=53515.0, I learned my value's should be ok - they're about the same as in that thread.

[B Tremblay]

Here are my voltages.  The gate of your Q4 (my Q3, per the schematic designations) is a bit lower.

Q1 2N5457

D: 6.14
S: .656
G: 0

Q2 J201

D: 4.65
S: .341
G: 0

Q3 2N5457 (top of mu-amp)

D: 9.12
S: 5.44
G: 4.17

Q4 2N5457 (bottom of mu-amp)

D: 4.90
S: .539
G: 0

U1 TL071

1: n/a
2: 4.57
3: 4.53
4: 0
5: n/a
6: 4.57
7: 9.12
8: n/a

[Auke Haarsma]

thanks for your values. Do you suggest I up the Q4 gate resistor to lower the gate-voltage? (currently r9 =1M). If so, what value would be suitable? If I double the resistance, would that halve the gate-voltage?

[Auke Haarsma]

Does anyone else have thoughts on this?

[Auke Haarsma]

Bumpertybumpy

I tried swapping around the Q's, placed J201's etc. The voltages Q3 and Q4 vary some, but not much. I do get voltages almost the same as B Tremblay. Stil I get very small gain. But I don't have an understanding of the circuit...I don't know where the gain 'happens'. With an audioprobe I think the gain at Q3 is still very low, and all the volume gain happens at Q4. The 'dirty' sound is mostly generated at Q2.

I'd be really helped out a lot if someone could describe:
- what is supposed to happen to the signal at Q1, Q2, Q3 and Q4. (eg, Q1 volume boost, Q2, dirty added, Q3 etc).
- how to increase the gain (not volume, that's ok) to get those hot, warm tones out of this unit (which I get when I boost, uhm... overdrive it with the MXR MicroAmp).
- if you don't know the Thor, but are familiar with amp-sims and know some useful links for me, please share. I'm willing to learn, but very green at this

thanks!

[Auke Haarsma]

Is this in anyway related to the Mu-Doubler at http://www.geofex.com/Article_Folders/jfetdoub/mudoubler.htm ? I'm trying to read up about this myself, but I'm not sure what I look for, tbh...

[phil]

I'd be really helped out a lot if someone could describe:
- how to increase the gain (not volume, that's ok) to get those hot, warm tones out of this unit (which I get when I boost, uhm... overdrive it with the MXR MicroAmp).
thanks!

Hi Pong - Just to let you know, I've very new at building effects so take my help with a grain of salt ... but since no one else has answered yet I'll jump in and give you my 2 cents and let someone more experienced correct me if I'm leading you astray (I have done 2 successful builds of the Thor though). From the Thor schematic it looks like the 1m pot is what controls the gain so from this I would suspect that Q2 is where the gain happens. First I would check that the 1m pot is wired correctly and with a dmm check the pot itself, or replace it with a known good pot. Then I would put the audio probe on the drain of q2 with a signal into the box and turn the 1m pot. If the gain does not increase then I'd know to focus on this area - specifically the 470k resistor and the 470pf capacitor into the gate of q2. Give that a try and post your results then I'll see if I have any other ideas for you to try. Since it sounds like you only have a gain issue, I think you are really close to getting it completely working ....

[MartyMart]

Ponq - few things to check out :
First, check that your resistor values are correct, 390 Ohm's are not 3k9 or 39Ohm's for instance
or your 220k being perhaps a 22k ??
It's easily done, so there's no shame if this has happened and that kind of little mistake can make
a circuit perform very badly, or even not at all.
So get your glasses on and read those bands  !

Second, Q1/Q2 have outputs from the "Drain" and Q3's output is from the "source" pin , are you
sure you have this correctly wired ?
Also check the 120k/39k/56k/1n  network at the + pin of the opamp is wired correctly.

Third , as stated above check also that your gain pot wiring is correct into Q2, your 1k2 sub wont make
much difference ti the overall gain.
I know that I may be "stating the bleedin obvious" BUT the circuit works from the schematic, so
you MUST have a error somewhere .... "mother nature" is telling you - as RG would say 

Keep looking and you will see the fault in your build, dont do this when you're "angry" or "fed up" with
it - this is the time when you just WON"T find it 

Marty.

[stm]

Ponq, please notice the following:

1) Typical gains for each stage are roughly 2x, 5x and 70x, respectively, assuming you use the indicated JFETs with "average" parameters.  Said values are for 2N5457's with Vp=-1.6V and J201 with Vp=-0.8V.

2) Based on the above, if you max out the Gain pot and use an audio probe to test the output of each stage you should notice increasing amounts of gain, particularly after the 2nd and 3rd stages.

3) The JFET voltages you posted have three problems:

-First, readings for Q3 and Q4 are interchanged as Q3 is supposed to be the upper unit and Q4 the lower one. Perhaps you just interchanged the readings when posting.

-Second, the difference between your Q4-Source and Q4-Gate is way too large.  You shouldn't have more than 1 to 1.5 volts of difference, however your readings indicate over 2.9V.  This voltage must always be less than the FET's Vp parameter.

-Third, and key to find the problem, your Q4-Gate reading (Q3-Gate reading according to the original schematic) is only at 2.37V, instead of Vcc/2 or roughly 4.5V by virtue of the voltage divider formed by the two 10k resistors.  Here you have to concentrate your efforts first.

Good luck!

[Auke Haarsma]

Thanks all for the replies! I'm determined to make this beast roar and this forum is a great hulp doing so.

To clear up the Q3 Q4 misunderstanding. I see that the layout in the pdf from ROG differs from the schematic at their site. The names of Q3 and Q4 are swapped. For clarity: in my first post I did refer to the pdf file, but from now on will follow the schematic at the ROG site.

(maybe usefull for ROG to correct this difference?)

[B Tremblay]

Thanks all for the replies! I'm determined to make this beast roar and this forum is a great hulp doing so.

To clear up the Q3 Q4 misunderstanding. I see that the layout in the pdf from ROG differs from the schematic at their site. The names of Q3 and Q4 are swapped. For clarity: in my first post I did refer to the pdf file, but from now on will follow the schematic at the ROG site.

(maybe usefull for ROG to correct this difference?)

It was corrected on January 28!

[Auke Haarsma]

I stand corrected! I used the pdf I downloaded (version 1.1).

[Auke Haarsma]

update (references to version 1.3 of pdf from ROG):

I went trough the board, checking the parts. I found R6 (Q4 gate resistor) to be of an odd value...either the resistor was broken or way too big (DMM measures up to 2M... I'm never sure if I read the colourbars correctly...blue colour of resistor seems to change the colourbars). I changed it to the listed 220k and voila: Out comes the gain! Gain pot does work now.

I also sockteted R9 (gate resistor of Q3) to find a resistor to change the voltage of Q3. Normally a 1M resistor should be at R9, I did get the right voltages with 160k... but this didn't help the sound. While change resistors I kept biasing Q1 and Q2, so that shouldn't be the problem. Debugging takes me a while, since I often look at the pcb and wonder what I'm doing, however, I feel I'm getting used to it Of course, tips and suggestions are always welcome.

Idea/question: I have not placed the bright and bottomswitch yet. But if I change R15 (100k resistor at "add bottom switch" section) to a 100k pot, would that allow me to add some sort of bass control (switch open/removed ofcourse)?

[MartyMart]

Ponq - few things to check out :
First, check that your resistor values are correct, 390 Ohm's are not 3k9 or 39Ohm's for instance
or your 220k being perhaps a 22k ??
It's easily done, so there's no shame if this has happened and that kind of little mistake can make
a circuit perform very badly, or even not at all.
So get your glasses on and read those bands  !

Marty.

I thought this may have happened   
220k 4 band colours : Red Red Yellow Gold/Brown
220k 5 band colours : Red Red Black Orange Gold/Brown      / = option !

Seems that you had a 2M2 there, which of course would reduce gain quite seriously.
I dont know why you're having trouble at Q3 ( top of Muamp ) that 1M taps off the 2x10k junction
so has 4.5v there.
I'd check some more resistor values, 1M is Brown Black Green Gold/Brown

Almost there

Marty.

[Auke Haarsma]

Thanks Marty! Your reply (and the suggestion in the replies by others) got me on the right track, that's why I looked at the 220k resistor

Any thoughts on my "pot instead of switch" suggestion? My logic is that with a closed swithc the 22n-33k-22n section will be bypassed. If I change the switch with a pot I can gradually get some current to go over the 22n-33k-22n setction thus adding (or removing?) some bass from the signal.

[MartyMart]

That may work, you have to be mindful that it's a "bridge T" network so odd things may
happen to your tone if you reduce that 100k down near "0"
The ROG chaps may come in with a better solution ....
I did this which worked quite well :
Changed the 33k to ground at BridgeT for a 10k & 22k log pot as variable resistor to ground.
This works quite well as a tone control, seems to reduce high/mid as rolled back .
Just an idea !

MM.

[Auke Haarsma]

Ej guys,

I checked all resistostor value's and they are all ok. I checked all capacitor values: ok.

The trimpots are adjusted correctly so I have the right voltages (still the same as me first post). I've swapped around 2n5457's and j201's to no avail.

I have two questions:

-Gate of Q3 stays at 2.37V, no matter if a swap around other Q's. This should be at around 4V. What part could affect this? All value's are as per the schematic. If I change R9 I can change the gate voltage,but the sound gets killed.

-with the current circuit I get a octave-sound. I think this is caused by the misbiasing of Q3. Right?

-C12 and C11 are those red Wima capacitors. When I approach these parts with my fingers, I get a strong hum. The other parts don't respond like this. If I approach it with a tool, I don't get this hum. Is this the magnetic/energy (?) field of my body interfering with the capacitor?

[stm]

Ponq, the hum induced into capacitors when touching C11 and C12 is sometimes a "normal" thing I've seen in other circuits. In particular, when I worked for an AM/MW transmitter manufacturing company, once we replaced our typical boxed-style caps for some poliester "greenies", and that introduced the problem you mention.  When we finally got our usual caps the problem was gone.  Nevertheless your circuit will be later put into a metal enclosure, so that shouldn't be a problem.

Regarding the odd voltage reading, it is definitely something wrong in there. I'd suggest the following:

1) Measure Vref as indicated on Thor's schem v1.0.  If it is not Vcc/2, then see what's loading the 10k+10k voltage divider, and check both resistors are actually 10k (sometimes a resistor may have its color bands printed wrong!).  If voltage is still low at Vref, start isolating everything that's connected to it... hey, wait a minute!  CHECK THE POLARITY OF THE 100uF capacitor: if it is installed backwards it'll drop the voltage at Vref due to its leakage.

I await your feedback.