MT Audio Design 'Quadimodo'



Some Comments on the Design
 
My first design is based on the famous Quad II amplifier. I bought a couple of those a few months ago for a reasonable price, well below what the transformers would cost today. The amplifiers did not work, but that did not matter as I from the beginning was thinking about a few changes in the design. The reason for buying the Quad II amplifiers was mainly an interest in the output transformers with their partial cathode load.

The MT Audio Design 'Quadimodo' will be designed as an integrated amplifier with line level inputs, a volume control and the possibility to run the pre-amplifier and power amplifier separated. In the first stage I do not intend to include a balance control, but there will be one fitted on the chassis if I feel the need for it.

The KT66 certainly is a good tube, but the tetrode connection used in the Quad II design has it drawbacks. I wanted to try a triode coupled output stage with split load transformers, and with KT66 in triode connection it is possible to get around 10W class A in push-pull, this is a little low, so I will try using KT88 instead. This will increase the power in triode mode to around 15W class A, and as I have a couple of KT88 lying around this is a perfect choice.
 
The output stage uses triode connected KT88 in push-pull with 10% of the load at the cathode. The load is approximately 7.8k anode to anode (I measured the ratio to approximately 28.2+3+3+28.2:1+1+2) which means about 15W class A1 with low distortion.
 
The driver stage is a mu stage design similar in the Alan Kimmel fashion. The tubes used are ECC83 as the voltage amplification tube and ECC88 in cascode as the upper cathode follower tube. This stage will easily swing the necessary +/- 100 Volts needed with very low distortion, and the amplification factor is high enough to make this a two stage power amplifier. The phase splitter is a 2:2.25+2.25 interstage transformer with 'para feed'.
 
The used fixed bias unfortunately means that an additional transformer must be used, but this is not too expensive.
 
The pre-amplifier stage uses a 6BL7 in SRPP. Simple but effective I hope. 

This design is only in the building stage, and I can not guarantee that it actually works. Feel free to try it, and please inform me about the result.

Power Amplifier Stage

'Quadimodo' power amplifier schematics

Pre Amplifier Stage

'Quadimodo' pre amplifier schematics

A very simple but good pre-amplifier stage. The 6BL7 is very linear and the output impedance is also low enough.

Power Supply

'Quadimodo' power supply schematics

The 6.3V winding is designed for 2xKT66 + 3xEF86 + 1xECC83 = 2x1.3A + 3x0.2A + 1x0.3A = 3.5A. I use 2xKT88 + 1xECC83 = 2x1.6A + 1x0.3A = 3.5A. I bought an extra filament transformer with two 6V windings (@1A) with primary voltage 220V, 240V. In Sweden we have 230V and with this connected to the 220V primary the secondary windings are 6.3V. These transformers were not expensive at all and gives me the opprtunity to use more powerful tubes. At the same time I bought a little 48V (6VA) transformer for the bias supply. The cost for four transformers (two per channel) is around 30$. I found the transformers at Farnell Electronic Components, but I don't think it is hard finding 6V transformers with selectable primary windings from other dealers.

Load-line ECC83 (mu stage)

ECC83 mu-stage loadline

The driver stage is a mu stage with ECC83 as lower tube and a cascode coupled ECC88 on top, this is coupled to a "para feed" phase split transformer with the ratio 2:2.25+2.25. The amplification factor is around 70, and the input sensitivity is 0.7V RMS.

Load line and analysis KT-88

KT-88 SE Amp CAD results

KT-88 SE Amp CAD loadline

KT-88 SE Amp CAD transfer curve
KT-88 SE Amp CAD mirror transfer curve

The cathode load is 1/10 of the total load, which will cause a change of the amplification factor as follows:
A0=5.6, beta=1/10 -> A=5.6/(1+(0.1x5.6))=3.59 (3.9dB feedback)
This means that the driver stage must be able to swing around +/- 64 Volts for full Class A1 output power (12.5W). The feedback will increase both the output power and the needed swing (because the feedback is higher at low grid voltage and lower at high grid voltage). The negative swing for 41V is 260V (gain 6.35) and the positive is 202V (gain 4.93). The driver stage swing for full negative swing is 67V and this will give 45V at the grid for positive swing. The output power at 43V swing is 7.04W. The push-pull configuration will also add 1W or something like that. The estimated Class A1 power is 16W (with around +/- 70V swing).
 
The transformer load is 3.9k (at 8 Ohm). The calculated distortion figures (390V/70mA) are good for a triode connected beam tetrode without global negative feedback. The 2nd harmonic is 6.2% and with the cathode feedback the  2nd harmonic distortion is reduced to 4%. Note the even transfer function (before cathode feedback), which will transform into an almost straight line with push-pull action. The push-pull configuration decreases the THD at least 2-3 times and this leaves us with a THD well below 2% at full power (16W) without global negative feedback. Push-pull is supposed to cancel out the 2nd harmonic, but the driver stage still have some. The 3rd harmonic in the output stage is pretty low, and I hope the 2nd harmonic will be dominant at all power levels. No global feedback is necessary.
 
A very interesting aspect of the cathode feedback in push-pull stages is that it acts separately on each half and makes them more equal, this will of course mean better cancellation performance (less even harmonics and less 'cancellation distortion').
 
The output impedance should be below 1 Ohm (2.41/2/(5.6/3.6)).

Tube data used here can be found at The Tube Directory

MT Audio Design Logo


 

This site uses frames, but if you have found this page via a search engine and do not see any frames, you can use the button below to get to the main frame of the MT Audio Design site.