Debugging a new amp!

[mth5044]

Hi everyone! First post on this fine forum as well as my first problem with a solid state amp build 

I've put together an amp from pieces around the web. The schematics are below. The first is the preamp and the second the power amp and power supply. The power amp is based on a TDA2005R chip. The FC+ and FC- are for an offboard cap, two 1000uF caps in parallel for 2000u for filtering. Left them offboard for space. Diodes are 1N4001 rather than 7's. Speaker is a 25W 8ohm Jensen.




TDA2005 datasheet: http://www.paia.com/KRUKits/K143/tda2005_datasheet.pdf

The LM317 was initially running at 17V, but I've turned to it down to 12V to see if it would help - no dice. Both 2005 and 317 are heat sinked and isolated.

The problem:

The presence control causes loud crackling when turned and the signal is always strangely distorted (as you can hear in the sound clip below) even when preamp volume is low.

Sound clip: http://snd.sc/13Ui8O8   Ephiphone Casino right into the amp.

0:00 - 0:19 presence all the way down
0:20 - 0:35 presence up a bit
0:35 - 0:49 you can really hear the crackle when the presence knob is turned up
0:50 presence knob full up and stays that way until the end of the clip.

Voltages:

Q1
D - 11.06V
S - 0.24V
G - 0.00V

Q2
D - 12.41V
S - 2.18V
G - 1.76B

Q3 - presence full down
D - 0
S - 0
G - 0

:loco

Q3 - presence full up
D - 6.93V and decreases as probe is held on
S - 1.51V
G - 0.47V also decreases as probe is applied.

TDA2005
1 - 1.16V
2 - 0.76V
3 - 7.68V
4 - 0.72V
5 - 1.14V
6 - 0V
7 - 12.18V
8 - 6.62V
9 - 12.62V
10 - 6.27V
11 -12.17V

LM317
1 - 19.62
2 - 12.62
3 - 11.33V

Something is clearly going on with Q3 but I'm not sure how to go about debugging! Thanks for any help, I look forward to exploring the forum.

[phatt]

In my limited opinion,, You have NO Ground reference to ground on gate of Q1. A real Worry!!!

And I'd guess,,neither does Q2. The 2m2 R needs to go to ground not some 1/2 bias setup.
Q3 same problem,, no ground reference.
Start there  and see if it improves.
Other minds here know more.
Phil

[Roly]

Welcome @mth5044!

The presence control causes loud crackling when turned and the signal is always strangely distorted (as you can hear in the sound clip below) even when preamp volume is low.

As shown the Presence control is actually acting as the Source bias resistor for Q3 and so it isn't at all surprising that you will get crackling noises when you move it.  It looks to me like there should be a cap in series with the Presence pot so it a) doesn't change the bias of Q3, and b) acts as a tone control.

Q1
D - 11.06V
S - 0.24V
G - 0.00V

Assuming a supply of 12 volts this FET is almost cut off, too much bias, and as @phatt has pointed out this is almost certainly because the Gate isn't ground referenced.  So stick a resistor between Gate and ground, at least 1 Meg, and if you are using a passive guitar you might care to try higher values, anything up to 10Meg, which will allow the pickups to resonate with more chime.  You should also have a similar resistor from the input to ground to avoid popping due to charging C4.

Your Volume control at 100k is loading the tonestack and entirely defeats the advantage of following it with a source follower - in fact you could omit Q2 and all its associated components because the input impedance at the Gate of Q3 is similar.  Taking C20 directly to the Volume control wiper, and given the high impedances C20 doesn't need to be an electro or have a high value, a 0.1uF poly would do.

Again as @phatt points out, Q3 also doesn't have a ground reference on its Gate, and a 1 Meg to ground would do here.  It seems that you are trying to apply negative feedback (NFB) to Q3 - but for the time being, at least until you get Q3 biased correctly, pull out R23 100k.

The regulator is marked LM350 on the circuit, but you discuss using a LM315.

There is no point in trying to measure the Gate voltages unless you have an ultra-high (~100 Meg) input resistance voltmeter.  In these common Source amplifier stages the important voltage is the Drain, and this should be around half the supply voltage.

You should try feeding in a signal from an MP3 player or similar to the TDA2005 at point "A" (ground referenced, via C23, and disconnected from Q3) to verify that the chip amp is working correctly.

You should be able to set the supply voltage back to ~17V provided that the voltage source is at least about 3 volts higher (so that the regulator has some headroom to regulate).

HTH

[Kaz Kylheku]

I recommend looking at the schematics for a vintage Roland Cube 60.  It has JFET circuits like this in the input stages, including a similar tone stack.

Speaking of which, notice how it uses a 1M pot after the tone stack which doesn't load it down as much.

And check out the proper references of the JFET gates to a voltage. For instance near Q1, there is a voltage divider R2/R3 which cuts the 38V to about 3.5V. This is conveyed to the JFET gate through R4.  This 3.5V voltage is written down next to the gate.

You need to similarly a voltage at each gate, and then superimpose your signal via AC coupling onto that voltage. The JFET is not properly biased if you leave the gate floating (every path out of it leads to a capacitor).

[Roly]

While you certainly can bias the FET Gate a bit positive, as you can with a triode, unless you are expecting quite high signal voltages or some other specific need it generally isn't necessary, as here;