MUSIC REFERENCE OTL-1 White Paper
A good white paper should start with a story, If you don't care about the story skip down to “How it works”
Forty years ago I got my hands on a pair of Harvard Electronics OTL amplifier. They were not identical, which I found strange. However the difference was only in the output tubes employed. I had plenty of experience with tube amplifiers at that time but I had never seen an OTL. The owner had a single page of typewritten instruction on how to adjust the pots on the amplifier. It also went into great detail of what to do if adjusting the pots did not produce the specified results. One of the instructions I found a bit curious was about swapping tubes from one side to the other if things would not bias up properly. This takes a lot of time because you have to wait for the tubes to re-stabilize every time and one has to be careful not to burn oneself on the very hot tubes. Adjusting this amplifier was not a simple matter. Without considerable practice I doubt many owners could adjust them, hence they came to me. There was nothing wrong with the amplifiers and nothing had to be replaced. I observed they were well built. Upon their return the customer was very pleased that he found someone who could sort out the rather odd instructions.
Thirty years ago I designed the RM-6 amplifier which I licensed to Counterpoint Electronics. I often get compliments that it is the best amp this person has ever heard. It's sound seems to linger in listeners ears long after hearing it. To my knowledge it was the first OTL to be direct coupled and employ a servo to automatically balance the DC output to zero. Other direct coupled OTLs require manual adjustment periodically.
Two years ago I started thinking about designing another OTL. Actually I never stop thinking about OTLs and I have made several breadboards in the interim, but none of them satisfied me. I did experiment with 6AS7 and 6C33 triodes in both totem-pole and bridge configurations. I experimented with pentodes and triode connected pentodes in SEPP. I made some 5,000 Volt SEPP OTL amps for Acoustat electrostats (those are very successful and we are currently making them). However SEPP circuits are not useful at speaker levels and impedance.
HOW IT WORKS:The Music Reference OTL-1 is a low powered, ultra short path amplifier with minimal parts, no output cap, DC coupled internally from input to output, self balancing via a Servo, single bias adjustment, single output pair amplifier. Via its direct output it can source 1 amp of current. It can also supply 100 volts into a high impedance load. Via its Impedance Converter tapped inductor it can deliver up to 4 amps from 20 Hz to 200 Khz. If one slows the servo, the amplifier is capable of DC response.
The amplifier uses NOS tubes designed and made at the height of the tube era by the best factories of the 1970's when tube production was at its highest level due to the popularity of Color TV. It is the Japanese invention of the transistorized TV that closed tube factories all over the world. The factories that stayed open the longest were in the countries where tube TVs were kept in service the longest (China, Russia and in Eastern Europe). I tell you this little known bit of tube history so you can appreciate that we are using the best NOS tubes made that are far superior to the current sweep tubes as the EL509 and PL509. Those tubes were great while needed for Color TV construction and replacement as they had to be rugged and at Sylvania the goal was to have them last 10 years.
To obtain low distortion, low noise and good damping a circuit was developed by Roger A. Modjeski that would have the shortest and fastest path possible. It is desirable in an amplifier to have DC coupling to produce the best bass and wide open-loop bandwidth so that feedback can be applied and stability into a variety of difficult loads is guaranteed. The resulting amplifier has a bandwidth from below one Hz to above half a megahertz.
This amplifier is unique in the field of past and present OTL amplifiers. Distortion is lower than most conventional amplifiers at 0.5% or less. The damping is higher at 20. The frequency and power response are far wider as stated above. The tubes deliver their power directly to the speaker with no intervening coupling cap or relay.
Unique to this amplifier is the inclusion of an Impedance Converter (IC) that allows the amplifier to output 10 watts into a variety of speakers from 1 ohm to 32 ohms. The impedance converter can be bypassed, however the amplifier power is then reduced to 3.5 watts for 8 ohm loads and doubles as the impedance doubles up to about 32 ohms. The IC along with a fuse for each output tube provides excellent protection from harmful DC present on the speaker. The amplifier will not oscillate into any load. The circuit description has more details of how this was accomplished.
Thanks to the advent of more sensitive speakers it is possible to achieve enough power with only one pair of output tubes. This reduces the power consumption of the amplifier to 140 watts, a fraction of what most OTL amps draw. Tube replacement is easy and matched tubes are not required as the servo adjusts for unmatched grid bias. However for best performance matched tubes are recommended and are available from us at a fraction of the cost of modern output tubes such as the KT-88, KT-120, EL-34 etc.
CIRCUIT DESCRIPTIONThe driver consists of a single tube for each channel. The pentode section of this tube provides the entire voltage gain of the amplifier which is about 500x or 56 dB. The triode section is connected as a bootstrapped split load phase inverter which (contrary to some misinformed individuals) provides equal drive to the triode connected output tubes. These tubes provide the current gain and can source one ampere from the positive or negative supply.
Feedback is a controversial subject among audio enthusiasts. It is considered more bad than good and many amplifiers tout the fact that they have no feedback. In recent issues of Stereophile Magazine there have been several reviews of amplifier which have feedback switches so that the listener can determine for himself the effects of feedback. In the examples I have read the feedback is 2 dB which is virtually nothing. In each case the reviewer reported that this small about of feedback did not please them and they preferred the sound without feedback. I cannot imagine properly applied feedback of 2 dB would worsen the sound of the amplifier as reported unless it was very poorly applied. It would take careful listening to hear it at all being vanishingly small.
The application of feedback is a tricky thing. Amplifier circuits do not have a place marked “put the feedback here”. One has to create a place to inject the feedback. I imagine the amplifiers previously mentioned might have created this injection point with the switch thus modifying the circuit as a whole when the feedback is switched in. The designer of the $75,000 amplifier admits putting the feedback switch on the amp to show people how bad feedback is. Well, one can certainly install a switch to make an amplifier sound bad.
In the OTL-1 the input tube literally rides on a the output terminal. So as the input rises the output follows it perfectly in phase. Rather than being an injection point for feedback this point exists in the fundamental circuit. This allows the amplifier to react immediately. In conventional amplifiers the feedback comes through the output transformer (with considerable phase shift that is load dependent). The feedback then needs an injection point which is often the cathode of the input tube or grid of a differential amplifier. Internal delays in the loop (phase shifts at low or high frequencies) can cause the negative feedback to become positive and make the amp oscillate at low or high frequencies or both. Many amps are unstable without a load and many have low frequency instability that causes the woofer cone to wander about its rest position.
The power supply consists of a 300 VA toroid bridge rectified and connected to two 2700 uF filter capacitors for exceptional energy storage. There are separate supplies for the driver, bias and servo.
The servo is a simple integrator that keeps the DC at the output below 20 mV at all times It does this very slowly and much slower than the lowest bass note from any instrument. Bias is set on the negative going tube and, due to DC coupling the positive tube adjusts itself to equal idle current.
A high inductance, low loss tapped choke provides protection from DC should a tube fail. Fuses prevent damage of other components if that should occur. Music Reference amplifiers are known for their excellent application of fuses. With only a pair of output tubes per channel tube failures are reduced and locating a bad tube is simplified. The choke is in parallel with the output of the amplifier to ground. It has three taps which allow for more power and lower distortion that would otherwise be available without it. It does not harm the sound of the amplifier in any way. For those who want to confirm that claim there is a direct output also.
A great deal of consideration to the user has been employed in this design. We hope that this amplifier will begin a new chapter in the history of OTL amplifiers.
