• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

Tube preamp for a TDA2040 chipamp

Hi, I'm thinking of building a tube preamplifier for a TDA2040 chipamp that I've built. As audio source I'll probably using an iPod or cumputersoundcard. Therefore I thouhgt that a low gain tube preamplifier should to the job. So then I found this one.

The gain i about 8 db (the page says) and it's supposed to drive a low impedance amplifier. So then I'm wondering, should a build a hihger gain tubepreamp instead, and is TDA2040 considered a low impedance amplifer? Been searching a lot for this, according to the data sheet it could be lowered by decreasing R1, but which value is considered low inpemadance?
 

revintage

Member
2007-05-30 4:12 pm
If the 2040 is built by the book you will not need any gain at all. Just add a pot at the input and that´s it. If you still want a buffer to raise the input impedance a CF is all you need.

Stay away from the example you showed! In general stay away from Diyaudioprojects.
 
Last edited:
With a compressed digital sound source, a tube preamp, and a chip amplifier you could combine into one system all the major types of audio distortion!

The gain of that circuit as claimed by the page is 8, not 8dB. Voltage gain of 8 is 18dB. The RCA book implies a gain of 12 (=21.5dB) (which seems more likely as the valve has a mu of 17 at normal bias). This is probably about twelve times too much. As with most unnecessary circuits, the main effect of this preamp is to add a little noise and distortion to the signal.

If you do decide to build it, ignore the advice about C4 needing to be a good quality capacitor. Probably the most critical capacitor is C3, but even that is not too critical in this application. RCA circuits tend to be a bit "bargain basement". This one makes the common mistake of connecting the grid straight to the pot wiper.
 
If you do decide to build it, ignore the advice about C4 needing to be a good quality capacitor. Probably the most critical capacitor is C3, but even that is not too critical in this application. RCA circuits tend to be a bit "bargain basement". This one makes the common mistake of connecting the grid straight to the pot wiper.

Yeah, I thought so to. But I think it still is the best schematic I've found so far.

I know that a preamp for the TDA2040 really isn't necessary. The reason I want one is just to add a little of tube distortion to the sound. But a low gain tube preamp is what a should go for then?
 

revintage

Member
2007-05-30 4:12 pm
Yeah, I thought so to. But I think it still is the best schematic I've found so far.

I know that a preamp for the TDA2040 really isn't necessary. The reason I want one is just to add a little of tube distortion to the sound. But a low gain tube preamp is what a should go for then?

Don´t go for that one it´s lousy by todays standard. Go for a 6N6P, same mu as the not so very good 12AU7, or something else with acceptable linearity.

If you want something with lower gain than this, build a Anode-follower. Single tube, gain of your choice. So easy and still great.
 

00940

Member
2003-09-03 12:17 pm
Liège
Rather than using a very low gain preamp, I'd knock down the gain of the tda2040 as much as possible (as long as it is stable that is) and use the tube preamp to provide the voltage gain. Chip amps often sound better with lowish gain in my experience.

What's the input impedance of your tda2040 ?
 

revintage

Member
2007-05-30 4:12 pm
The OP wants some distortion!

:D:D, thats an interesting point of view.

Chip amps often sound better with lowish gain in my experience.

What's the input impedance of your tda2040 ?

What would the reason be for a chipamp to sound better with even more GNFB, beats me?

For what I have seen inputZ is high but in the available applications mostly 22k (resistor on the input).
 
Rather than using a very low gain preamp, I'd knock down the gain of the tda2040 as much as possible (as long as it is stable that is) and use the tube preamp to provide the voltage gain. Chip amps often sound better with lowish gain in my experience.

What's the input impedance of your tda2040 ?

Well there's the problem, I don't got clue. Been searching for it a lot. Don't laugh then I say that I think it's about 22k ohm, it might be totally wrong.

You have a good point though, maybe use a higher gain preamp and try to lower the gain of the chipamp instead. I've seem many DIY people who use high gain tube amp in front of there (solid state) poweramps, so maybe thats the way to go I don't know.

Found another tube preamp which makes sense to me, only that the gain is now 51 db, could probably be lowered a little by recalculating some resistor values and so, but maybe high gain preamp is the way to go anyway?
 
the 12AX7 would be one of the last tubes I would use for a linestage... too much gain and it can't drive a low impedance load all that well either. In your case, you don't need gain and trust me - it's no fun to having _too_ much gain.

heck, just put a simple cathode follower in front of the amplifier or even use a FET. If you want to get complex, use a White Cathode Follower which (with the right tube) can drive just about any crazy load you will see at the line level.
 

artosalo

Member
2010-02-16 9:00 am
The reason I want one is just to add a little of tube distortion to the sound. But a low gain tube preamp is what a should go for then?

You can reduce the gain of the TDA2040 by increasing R2. Put a 10 k variable resistor to find what you like. The bigger the resistance, the lower the gain (and bigger NFB).

The input impedance of the TDA2040 is practically the value of R1.
You can put there 100 k if you want.

As a preamp, since you do not need much gain, but want to have distortion, put a 12AX7 stage with unbypassed 10 k cathode -
and 100 k anode resistor. Then the gain will be some 13 dB only (assuming the R1 is 100k). And surely some amount of distortion :D.
 
Ok, let's see if I've got this right. What artosalo suggests is that I increase the input resistance of the TDA2040 amp to 100k (R1). And then use a single 12AX7 stage like the one a linked earlier only that a changes the values of R10 (anode resistor) and R9 (cathode resistor) to 100k and 10k.

I now calculated the gain of the preamp circuit to be 8,5 db. artsalo wtote that it would be about 13 db, so what have I missed? The input impedance of the TDA2040 does not affect the gain, or does it?

Or maybe he didn't refer to the circuit mentioned earlier at all, just a single 12AX7 stage. I'm really confused, please help.
 

artosalo

Member
2010-02-16 9:00 am
I now calculated the gain of the preamp circuit to be 8,5 db. artsalo wtote that it would be about 13 db, so what have I missed? The input impedance of the TDA2040 does not affect the gain, or does it?

Below is what I calculated. The gain is 13 dB. There are also the equations so that you can check again.

But what is your problem ?


[IMGDEAD]http://kuva.termiitti.com/image/15661.gif[/IMGDEAD]
 
My problem is that I'm stupid. You're calculations seems right, don't know why I didn't got it right.

However like your idea to use a low gain 12AX7 stage to get som distortion, so I quickly throw this schematic together. Haven't calculated the value of the capacitors yet, but I guess it should be something like 0,1uF for C1 and C2 and 3-5uF for C3. All 400V. Donät sure if this could work, feels to simple some how. What do you think?
 

Attachments

  • preamp1.jpg
    preamp1.jpg
    15.8 KB · Views: 119

artosalo

Member
2010-02-16 9:00 am
Leave the last 12AX7 half (cathode follower) away and connect the signal directly from C2 to TDA2040. Then the gain will be 13 dB, assuming R1 is changed to 100 kohms. Then C2 should be min. 220nF.

By the way, the schematic you made has such an error that there is no bias arrangement for the cathode follower.
 
Leave the last 12AX7 half (cathode follower) away and connect the signal directly from C2 to TDA2040. Then the gain will be 13 dB, assuming R1 is changed to 100 kohms. Then C2 should be min. 220nF.

By the way, the schematic you made has such an error that there is no bias arrangement for the cathode follower.

R1 should be 100k, I know. And yeah, I missed a resistor at the schematic. But maybe one stage is enough however. Think I'll build it only with one stage, connecting the TDA2040 to C2 and then I'll see how it sounds.
 

johnr66

Member
2009-03-05 1:55 pm
The chip amp will become a nice RF oscillator if you provide too much negative feedback (reduce gain). The datasheet recommends gain be 24db minimum.

The 22K resistor on the input pin to ground permits bias current to flow for the input stage. Without the bias resistor the amp falls silent or spits and crackles if the input signal is strong. If you change this value much, I'd be doing some testing.

My recommendation is to leave the TDA2040 components close to the datasheet values and add a divider network or pot and attenuate any gain from the tube stage.

I'd use a "poor" tube or really louse up the circuit to get some distortion or your just going to have a pointless stage. A well designed tube stage is going to be low distortion anyway.
 

artosalo

Member
2010-02-16 9:00 am
The chip amp will become a nice RF oscillator if you provide too much negative feedback (reduce gain). The datasheet recommends gain be 24db minimum.

Correct. Then R3 should be 12 k, but not less. Then the gain of TDA2040 is 25 dB.


Without the bias resistor the amp falls silent or spits and crackles if the input signal is strong.

Absolutely so, but 100 kohms should be fine.

When R3 is 12 kohms, the sensitivity (to full output) of TDA2040 is some 550 mV.
And after adding the 12AX7 (with suggested circuit) the overall sensitivity is some 120 mV. Therefore you should place the volume pot just at the input of 12AX7.
I quess you typically use line level source. There would be also a good place for fixed attenuator.
 
I'm not sure if this is what you want, if you're looking to add some distortion that sounds like a poor tube circuit. But a 12B4 is a very linear low gain tube with a low Rp. This is a very good sounding circuit. Gain of about 6-6.5.

Use a 10M45S as a plate load; 9-11 1.7V red LEDs in the cathode circuit and adjust the 10M45S for about 130V on the plate. Should be around 15-20mA.

For an output cap, use the minimum size that gets you the LF response you need. PPFX Multicaps sound very nice here. Of course V-Caps are much better and 10x the price.

Stuart
 
I was actually thinking of putting an passive attenuator or at least a volume control in front of the 12AX7. Lowering the gain of the chipamp to 24 db sounds like an good idea.

For tube I was thinking of using one half of an 12AX7 for each ch, so I only need to use on tube then. Can't think of why it wouldn't be a good idea. I also got some really old 12AX7 tubes, maybe the sounds worse (but in a good way) than newer ones.

I'm not sure if this is what you want, if you're looking to add some distortion that sounds like a poor tube circuit. But a 12B4 is a very linear low gain tube with a low Rp. This is a very good sounding circuit. Gain of about 6-6.5.

Use a 10M45S as a plate load; 9-11 1.7V red LEDs in the cathode circuit and adjust the 10M45S for about 130V on the plate. Should be around 15-20mA.

For an output cap, use the minimum size that gets you the LF response you need. PPFX Multicaps sound very nice here. Of course V-Caps are much better and 10x the price.

Stuart

Sounds interesting. Doesn't really get it though, got any schematic, exemples or so?