HAFLER DH-220  ULTIMATE UPGRADE KIT. THIS UPGRADE ADDRESSES MANY ISSUES PERTAINING TO AUDIO AMPLIFIER DESIGN PRACTICE. SIMPLE UNIQUE JFET CASCODE DESIGN ALSO OFFERS CLEAR AUDIBLE IMPROVEMENTS AND INCREASED POWER OUTPUT. INCLUDES GOLD PLATED BRASS RCA INPUTS AND OUTPUT BINDING POSTS CAPABLE OF ACCEPTING 6 AWG WIRE, 2 RAW PC-1 DRIVER BOARDS WITH ALL COMPONENTS, 17 PAGES OF INSTRUCTIONS AND PICTORIALS, AND ASSOCIATED WIRING.

THIS PACKAGE IS DESIGNED FOR INTERMEDIATE TO EXPERIENCED KITBUILDERS YOU MUST HAVE GOOD SOLDERING PRACTICE SKILLS AND BE ABLE TO FOLLOW INSTRUCTIONS. YOU MUST BE ABLE TO READ COLOR BANDS ON 1% RESISTORS (MATERIAL SUPPLIED) AND INSTALL AND SOLDER SMALL PARTS ON PC BOARDS. YOUR TOOLS SHOULD INCLUDE:

A 40 TO 100 WATT SOLDERING IRON WITH 3/16 OR ¼ TIP WHICH CAN REACH AT LEAST 600 DEGREES F.

ROSIN CORE SOLDER (SUPPLIED)

A DAMP SPONGE OR CLOTH TO WIPE THE TIP OF THE IRON

A WIRE STRIPPING TOOL FOR REMOVING INSULATION FROM 16GA TO 24 GA WIRE

A MEDIUM-BLADE SCREWDRIVER

NEEDLE-NOSE PLIERS

SMALL DIAGONAL OR SIDE-CUTTING PLIERS

A ¼ SPIN-TITE NUT DRIVER

PLIERS, NUT DRIVERS, OR SOCKET SET FOR INSTALLING INPUT AND OUTPUT JACKS

YOU WILL ALSO NEED BOTH AN AMMETER CAPABLE OF RESOLVING 250MA DC AND A DC VOLTMETER CAPABLE OF RESOLVING 10MV DC. A VARIAC WOULD ALSO BE A PLUS BUT NOT REQUIRED.

XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX

Being intrigued with audio since 1973 has given much insight on the topic of amplification and proper design practice. My interest in audio continues to be both my hobby and my passion. Employment at both Dynaco and Hafler Co. was not only enjoyable but allowed me to work and converse with some very accomplished engineers. I continue to work on many audio projects, but will discuss my views on power amplification here.

                So what is it that allows a power amp to excel in its ability to reproduce music?  Why does Brand A power amplifier sound better than Brand B power amplifier? Let us assume both have similar power, distortion, noise, gain, slew rate, etc. specifications. Brand A is proclaimed to be an exceptional amp.  It certainly looks better cosmetically.  Does that mean it sounds better?  It costs $1000.00 more than Brand B.  Does that mean it is better?  We read or hear about terms like triode sound, no coloration, more natural, less fatiguing, cleaner, neutral, transparent, precise imaging, superb stage, etc., but what does it all mean?  We cannot measure these terms quantitatively with standard test equipment.  We can make standard measurements as set forth by The Institute of High Fidelity but that is as far as it goes.  Double blind listening tests are the only way to proceed at this point, revealing the truth about your $500 amp or your $1500 amp.  Of course all testing of this nature is now deemed subjective, but a trained ear will fairly consistently pick the superior amp.

            So what can we surmise at this point?  Why does Brand A amp own the preferred sound?  We can agree that it is surely not because it looks better or costs more. So what is it?  What is the solution? I have concluded after years of listening tests, circuit analysis, and endless bench testing, the answer is not as elusive or fleeting as I once felt but really what one would expect as a now seasoned engineer.

1. Simple circuitry - This means not pushing the audio signal through more semiconductors and capacitors than needed. We are dealing with non-linear components and need not compromise the audio essence with more ingredients than needed. This will only lead to unnecessary phase shift, distortion, and loss of clarity.

2. Good design practice - This is where many amplifier designs fall short. The first problem area overlooked is PC board layout.  It is crucial that 2 layers or multi layer PC boards have minimal or no track crossover from layer to layer. This will tend to inject capacitance into unwanted areas and cause instability.  In turn, this leads to the addition of more capacitors to stabilize circuitry, effectively robbing bandwidth and introducing additional phase shift. The amp design is now flawed by poor PC board layout. Amps that have multiple small value pF caps or caps tacked on to the foil side would be suspect of this flaw.  The second problem area and one that will corrupt the sound of even the best amp designs is proper grounding technique.  Input, output, and power supply grounds must follow the star ground scenario. For some reason, many manufacturers still to this day do not seem to get the importance or take grounding serious enough. Ground is ground. Period! It drives me crazy when I see resistors in the ground path to balance out ground loop situations or daisy-chaining unrelated grounds to save 12 inches of wire. These acts will muddle the audio content and ruin what we are trying to accomplish. If you have any of these situations, correct them or save your money and go buy a $159 receiver. When evaluating a design, look at the B+ and B- rails in relation to power supply and signal grounds. Then, try to visualize the circuit backwards where the grounds now become the B+ or B- and the B+ and B- become the grounds. This look can be very revealing. Would you insert current limiting on power supply rails and expect music to sound dynamic and accurate. I already have a problem with fuse resistance in the rails skewing dynamic capacity and incorporate bypassing measures when required but that is a topic for a different discussion.

3. Quality parts - Audio engineers are often limited to the final costs requirements of a new design. Unsurprisingly, the first parts to be downgraded to save on cost are capacitors. (Except for their own unit of course) This is a poor choice. Even under the double blind listening test, most audiophiles can consistently pick out the difference between quality polypropylenes and inexpensive metal films. Semiconductors should also be of good quality with low noise and distortion characteristics. Now all that is needed is a properly sized power transformer and main power supply filtering capacitors along with some desirable bypassing.

            Pure amplification does not have to be elusive. Listen to simplicity and enjoy.

                PC-1 DESIGN GOAL REALIZATION

A. Precise PC board layout with no track crossover.

B. First stage single ended Jfet/bipolar cascode differential amplifier with CCS regulation. Jfets incorporated are genuine Toshiba 2SK170BL devices with IDSS closely matched. If you know anything about this Jfet, I need not say anything else. High input impedance, high gain, tube like presence, and eerie low noise in listening tests makes it a no brainer. Cascode configuration reduces Miller effect, drastically increasing bandwidth and lowering distortion. This all important first stage configuration eliminates the need for massive gain and feedback used in many amps to cover up distortions and design flaws.

C. Second stage also uses SE bipolar cascode amplifier with CCS regulation. Again, Miller effect is reduced and bandwidth is not compromised.

D. Simple bias control and emitter follower circuit to drive output stage is added and all that is needed to complete the design goal. Clipping now occurs at 150-160 watts into 8 ohms instead of the 110-115 watts from original PC-6 or PC-19.

E. Polypropylene capacitors throughout with all HE electrolytic caps bypassed with polypropylenes

F. Vital star grounding realized with pictorial showing proper wiring arrangement

SPECIFICATIONS

Signal to Noise Ratio, unweighted:

Exceeds 100 db referred to 100 watts into 8 ohms                       106 dB Typical

Typical THD at 100 watts into 8 ohms:

20 Hz: 0.002%

1 kz: 0.002%

20 kHz: 0.006%

Input Impedance:

33,000 ohms.

Input Sensitivity:

1.285 volts rms for 100watts into 8 ohms. 

Frequency Response into 8 ohms:

- 3 db, 4.8 Hz to 225 kHz at 1 watt. 

Semiconductor Complement per board:

9 transistors, 2 Jfets, 5 diodes, 4 zener diodes