8
\$\begingroup\$

I'm planning a build for a guitar amplifier with 2 stages, the first being a tube pre-amp and the second stage being a solid state amplifier. Both amplifiers will be powered from the same source, a 31v DC power brick.

The pre-amp will be powered with a DC-DC boost converter from the 31v DC power brick to obtain the high voltage required; the boost converter ground output is tied to the input ground. The audio signal ground will be tied to the same ground.

The solid state amplifier requires a dual supply which I will create with a voltage divider, giving -15.5v, virtual ground and +15.5v.

Voltage divider

As the audio signal ground will be relative to the ground of the 31v power brick (equivalent to the -15.5v of the voltage divider) then this signal would cause an issue connecting from the output of the pre-amp to the input of the solid state amp, effectively shorting -15.5v and the virtual ground.

It would seem that I need to isolate the audio signal so it can be provided as a floating AC voltage.

Can the signal and ground be isolated by adding a capacitor in series with each component like so:

Or would the only solution be a 1:1 audio transformer?

\$\endgroup\$
4
  • 1
    \$\begingroup\$ Normally a transformer, even one rated for audio, will cut off below 40 HZ and above 15 KHZ. For rhythm/lead guitar it will be fine. It would not be good for deep bass or the kick-drum or cymbals. No need for transformer if only one power source. \$\endgroup\$
    – user105652
    Commented Jan 16, 2018 at 0:12
  • \$\begingroup\$ How much current does this power brick supply? \$\endgroup\$
    – user105652
    Commented Jan 16, 2018 at 2:35
  • \$\begingroup\$ @Sparky256 Thank you I've just attempted to listen to a 40Hz signal to see how low it is, apparently MacBooks can't output 40Hz. The power supply is rated at 4A (124w) which should be plenty for both the amps. \$\endgroup\$ Commented Jan 16, 2018 at 5:19
  • \$\begingroup\$ How much current will the solid state amp draw? Do you really think a voltage divider is going to work there? \$\endgroup\$
    – JRE
    Commented Jan 17, 2018 at 13:45

4 Answers 4

2
\$\begingroup\$

I am going to suggest that you want a differential connection from the preamp to the power amp. This can be any of three techniques:

1) audio coupling transformer

2) differential line receiver such as THAT 12xx series THAT differential line receivers

3) if it is available, a differential input on your power amp.

The audio coupling transformer has the advantage that it is easy to implement. Jensen makes nice transformers but they are expensive.

My company makes small PC boards with a semi-custom transformer on it that sells for considerably less than the Jensen transformers. Performance is not as good as the Jensen but is exceedingly good nonetheless. Trinity Electronics transformer card

There are many different differential line receivers available. My current favorites come from THAT Corporation but TI, Analog Devices, many others make pin-compatible parts. It is important to match the input coupling capacitors and use precision resistors in the input stage. Any mismatch causes poor common-mode rejection.

Some power amplifiers have inherent differential inputs, even though the amp may normally be used with only a single-ended input. Again - it is important that the input network (resistors, coupling capacitors, HF rejection capacitors) are all matched so as to achieve the best common-mode rejection.

The reason for using a differential input to the power amplifier is to eliminate any hum or ripple on the negative supply rail. Using a simple singled-ended connection as suggested by others can lead to excessive hum or noise in the amplifier.

\$\endgroup\$
1
\$\begingroup\$

In this case you would want to use a capacitor at the audio input. This stops the input from being ground referenced, you can then reference it to something else by adding DC (through a pull-up/down or other means). Just make sure that the capacitor can handle the voltage between the virtual ground and the actual ground, 15.5 V, with the additional swing of the audio level. You would use another capacitor again at the output when you are switching back to being referenced to the supply ground as opposed to the virtual ground.

You do not need to connect the grounds, the audio can be referenced from either. Capacitors block the DC offset from either supply from transferring between stages.

schematic

simulate this circuit – Schematic created using CircuitLab

The main thing to consider from an audio stand-point would be to choose a large capacitor value. The capacitor will form a high-pass filter with any resistances to ground and so will limit the low-frequency content in your signal. Series capacitance tend to add noise at low frequencies as well so it is good to choose high-quality capacitors.

Also make sure that you are not trying to drive any DC from the circuit using the virtual ground as it cannot handle very much.

\$\endgroup\$
0
\$\begingroup\$

You can use a series coupling cap, but bear in mind that that cap (if the DC on either side is appreciably different) has to charge when the circuit is powered, which might cause an appreciable "thump". You might want to disconnect the output for a while, until the circuit has stabilised.

\$\endgroup\$
2
  • 1
    \$\begingroup\$ Technically it wouldn't be a decoupling cap, it would be a coupling cap. \$\endgroup\$ Commented Sep 23, 2018 at 2:45
  • 1
    \$\begingroup\$ correct. my bad. edited. \$\endgroup\$
    – danmcb
    Commented Dec 18, 2018 at 19:17
-1
\$\begingroup\$

The audio signal ground will be tied to the same ground. The solid state amplifier requires a dual supply which I will create with a voltage divider, giving -15.5v, virtual ground and +15.5v.

If your audio signal ground is tied to the Ground of the Preamplifier and thus the -15.5V Rail Then your clipping all the negative voltage excursions of the AC Signal. You have already Biased the Solid State Amp. Using A Capacitor Connected in series with the pre-amp output and the Power Amp DC Biased Ground you will eliminate prior DC Bias with the AC Coupling Capacitor and the DC Biasing networ you already have will keep your AC Sinusoidal Waveform revolving around your solid state amplifier Virtual Ground

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.