Home Tube DIY Asylum

Do It Yourself (DIY) paradise for tube and SET project builders.

  Register / Login

An interesting excerpt from Sinclair Audio and Hi-Fi Handbook 3rd edition




The reasoning of a designer is always inspiring for a DIY design.

>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>
The Earmax headphone amplifier

In 1994, a delectable little German designed headphone amplifier called Earmax was released to the world - anyone who saw it fell in love with its diminutive dimensions, and it redefined hi-fi jewellery.
Unfortunately, at £325, it also redefined bank balances, so when the Earmax was later reviewed, the review information was used to
reverse engineer the circuit.

The photograph showed three valves, each of which
contains two triodes in one envelope.
The input valve was stated to be an ECC81, and the ECC88 was said to be
acceptable for the output valves.
The amplifier is stereo, so each channel can only use three triodes.
The amplifier was stated to be able to drive a low impedance load
(200 Ω- 2 kΩ), so the output stage must be a cathode follower.
Low output impedance from a cathode follower implies high gm, which confirms the suitability of the ECC88 (≈ 10 mA/V at high(ish) anode currents).
But even a gm of 10 mA/V still gives an output impedance
of 100 Ω in normal cathode follower mode, and we need
much lower than that to be able to drive 200 Ω; a
reasonable design target would be 10 Ω, or less.

There are two ways of achieving a 10 Ω output impedance. We could either use a conventional cathode follower, and apply 20 dB of overall negative feedback, or we could use a White cathode follower.
Maintaining stability with 20 dB of negative feedback might be a problem, so it is more likely that a White cathode follower was used.

There are two versions of the White cathode follower, one of which has a very low output impedance and can only be operated in Class A, whilst the other can be operated in Class AB, but requires a phase splitter.

Both forms use two triodes, leaving only one valve for the input stage.

It is possible to make a phase splitter with only one valve, but this would give the entire amplifier a voltage gain of ≈1, which would be insufficient to drive 2 kΩ headphones from the typical line level available from a Walkman Pro.

From these arguments, it is possible to say with a fair degree of certainty that the circuit topology is as shown in Fig. 10.11.

Earmax is alleged to be able to deliver 100 mW into an unspecified load, but loads of 200 Ω- 2 kΩare deemed to be acceptable.

100 mW into 200 Ω would require a far higher current than the ECC88 can deliver, so the assumed target was 100 mW into 2kΩ.

Using P = I2R, 100 mW into 2 kΩ requires 7 mARMS, and implies a minimum of 10 mA quiescent current in the output stage (Class A).
Perusal of the anode characteristics of the ECC88 reveals that at this operating current, gm=10 mA/V, and ra=3.3 kΩ.

The equation for the output impedance of the version of the White cathode
follower used is:

rout = 1/gm2 R'

Where R1 is the parallel combination of ra(anode resistance), and RL
(anode load).

Although a high value of RL reduces rout, it wastes power and increases heat dissipation.

A good compromise is
RL=ra = 3.3 kΩ, giving a theoretical output resistance of 6 Ω, so the choice seems justified.

In order to drive 100 mW into 2 kΩ, we must swing quite a large voltage. Using V2 = PR, we need 14 VRMS , or 20 Vpk.

Now, this might seem to be a trivial voltage swing for a valve, but in combination with the 10mA standing current, it is easy to choose an operating point that exceeds maximum allowable anode dissipation.

Although
Pa(max.)= 1.5W for an individual triode within the ECC88 envelope, total anode dissipation is only 2W.

Va = 91V,
Ia = 10 mA gives an anode dissipation of 0.9 W, and would just allow a 20 Vpk swing before running into grid current.

Once the operating point has been deduced, the required HT voltage and biasing is easy to calculate, and an input stage is needed.

Although an ECC81 was used for the original circuit, it forces a rather low anode current (≈ 1 mA), and rather high gain, which must have resulted in increased noise and sensitivity to valve variation.

As an alternative, the input stage could use another ECC88.

The advantage is that the ECC88 can be operated very linearly at the anode voltage required to directly bias the output stage, whilst drawing a reasonable anode current and hence minimising noise.

Because the gain of the ECC88 is only half that of the ECC81, input sensitivity is reduced, and noise becomes still less of a problem.

The amplifier now has a total voltage gain of ≈28, and needs ≈540 mVRMS for its full output of 14 VRMS into a 2 k Ω load.

Turning to power supply requirements, an HT of 217 V @ 26 mA is required, whilst the heaters could be wired in series, requiring 18.9 V at 0.3 A, which could then easily be supplied by a regulated DC supply to ensure an absence of heater induced hum.

With the calculated sensitivity, there should be no problem with hum, even with AC on the heaters, and simple two stage RC smoothing of the HT supply will be quite adequate.
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>


Omnes feriunt, ultima necat.


This post is made possible by the generous support of people like you and our sponsors:
  VH Audio  


Topic - An interesting excerpt from Sinclair Audio and Hi-Fi Handbook 3rd edition - cellailuca 03:49:54 06/12/25 (2)


FAQ

Post a Message!

Forgot Password?
Moniker (Username):
Password (Optional):
  Remember my Moniker & Password  (What's this?)    Eat Me
E-Mail (Optional):
Subject:
Message:   (Posts are subject to Content Rules)
Optional Link URL:
Optional Link Title:
Optional Image URL:
Upload Image:
E-mail Replies:  Automagically notify you when someone responds.