Reply with quote #1
A short while ago I constructed a DC version of the celebrated 300B SE amp much discussed on this forum. My variation employed the simple DC heater designs used in the Masterpiece preamp. The sonic results were outstanding. Encouraged by this I invested in what Bruce said was even better; the Pinnacle. He was not wrong.
The topology of this amp closely resembles the highly evolved one employed in the Evermore. The mysterious 'regulators' are as in the Masterpiece simply for the DC filament supplies. The build was straightforward as is usual for TS designs, but I must confess here that I am speaking as someone with a long lifetime's experience in electronic design and construction. B± voltages are quite high at >300V and this quickly susses out any dodgy tubes. I have had some rather poor 300B EH samples recently. Anyway, that sorted, how does it perform? I do not have any inclination towards hyperbolic 'audiospiel' so you will have to accept my subjective impressions in relatively plain English. The first thing I noticed was the increased power available which gives a generally more dynamic sound. There is clearly more headroom compared to the 300B SE. I found the sound captivating. You can listen for long periods without fatigue. With a pair of 86dB Wharfedale Diamond 220 speakers the sound is smooth and mellow. I was surprised by the acceptable loudness that was obtainable with these units. 95dB Fostex speakers blow your head off! Speech is presented with absolute clarity and for my musical preferences, classical, vocals and some pop, the performance more than satisfies my requirements. No harshness at all to my ears. I have been using this with the Masterpiece preamp which has enough 'wellie' to drive an elephant! A highly recommended combination. This amp is, for its power class, as well sounding as any I have ever experienced. It represents incredible value for money and, personally, would have paid out a good part of its cost just for the design. These are early days and I will write further on this in the fullness of time. In my opinion this amp represents a bargain not to be missed. Kelvin
Reply with quote #2
Great post. Thank you for the update. After hearing my Beasts for the first time I was surprised how effortlessly Bruce's PP amps handle music. For me it was the first thing I noticed on starting up the amps. Add in the 300B and reasonably efficient speakers then it must be a real killer sounding amp.
Reply with quote #3
Thanks for sharing your first impressions with the Pinnacles.
Is there any audible difference between the DC & PP version driving your Fostex speakers? What speaker design do you use?
Reply with quote #4
Thanks Sailor. Yes, the more I listen the more I am pleased. I am striving to follow your advice and not be too much of an 'Audio Fool'! With my techie background I do realize I go astray all too easily.
I have no wish to fiddle with this amp right now and will get on with some serious listening so I can give any further impressions on the unvarnished, original design.
Reply with quote #5
Hi Wolfgang. The sound quality between my DC 300B SE and the Pinnacle amps is, to my ears, quite close though they are different. Both give outstanding sonics, but the PP's extra power does give it a distinct edge. I am rather limited for space in my 'listening' room: the Fostex FE166En units are mounted in simple 17 litre front ported cabinets.
Reply with quote #6
Thanks Kelvin for your input. The two amps do sound very similar, the big difference being dynamics, authority and punch. The Pinnacle with its higher power allows it to be used with many more speakers too. That's why it sounds better. It also demonstrates that a push-pull amp can be just as smooth as a single ended amp.
Reply with quote #7
Wondering how the bias stability of the push-pull output stage is guaranteed in Pinnacle. What sort of mechanism prevents DC offset drift and uneven bias distribution btw parallel 300B tubes?
Reply with quote #8
I am sure Bruce would be able to give a more authoritative response to these questions than myself, but I will say what I am able to at this time. I have only been running my Pinnacle amp for less than 3 weeks, but I have noticed that there has been been very little, if any, change in the DC offset of a few mV since I initially set up both monoblocks after their first warm-ups. The twin pair, series connected output stage topology is also used in the SOB amp in which I have found the DC offset drift to be negligible even, after many months of daily use, albeit with very different tube types. There is also very little DC offset increase in the presence of substantial signals. Power line voltage variations effects on the DC offset tend to cancel with the use of symmetrical B± supplies. The applied bias voltages for the upper and lower pairs are controlled by Zener diodes and are quite stable. I did not use closely matched tubes. They were however all 300B EH types and the boxes indicated little variation in Ip values when tested at 300V and -60V bias, conditions close to those used in the Pinnacle. My experience has been that new tubes from the same source tend to age at about the same rates; we will have to wait and see with these particular300B EH tubes. I have not measured the individual tube currents at this time, - I am in an anti-fiddle, listening mode at the moment- - but I have no reason to suspect large differences. As far as I am aware, there are no other control features of the type to which you are referring. I hope I have not misunderstood your questions. Kelvin
Reply with quote #9
It doesn't matter. In fact, a slight deviation in push-pull amp is an advantage. There are other mechanisms that compensate for nonlinearities in the output stage such as degenerative and loop feedback. If all the tubes switch from class A to Class AB at precisely the same time, that creates a hard crossover notch at high power. If they all transition at different times, distortion is reduced. The transition is much more gentle.
As far as DC instability is concerned, that is also a nonevent. The 300B is extremely stable. Once set, drift is only a few millivolts, which is less than negligible. It is essentially nonexistent. There is no need to compensate for something that doesn't exist.
Reply with quote #10
When people read stuff on the internet that sounds authoritative, they tend to believe it. If they start believing things that affect how my products are perceived, then I have to speak out.
You are very smart there is no doubt. But I know more about OTL's then anyone else on the planet. I care absolutely nothing what others publish on the internet about audio. I look at none of it. I'm at the top of the pile and I got there by being the best. My products speak for themselves. There is no authoritative audio press to compare or analyze my work. Such a body does not exist. In the world of high end audio, it's 99% crap. I'm a proud member of the 1% who isn't.
Reply with quote #11
I personally feel it is very important to differentiate between transformer based pp output stages and OTL pp ones. Most of these autobias tricks are probably more relevant to the situations involving magnetics. SE output stages with parallel tubes are quite a different kettle of fish and not the intended primary focus of this thread. I only wish my aging ears were up to detecting some of these sonic nuances. So for me some of these 'events' will have to remain 'non-events'. My build of the Pinnacle still sounds very sweet to the said ears!
Reply with quote #12
If it was all about the measurements we would all be using opamps and transistors.
Bruce, thanks for designing quality products at fair prices. Your Beasts are the best amp I have ever heard and I have heard thousands. Some day I hope to hear your new 300B line. I'm sure it sounds great.
Reply with quote #13
Re: Chasing ghosts and non-events
Hi Bruce, I would like to add that as an owner of both your 300B 25Hz heated SE and your Masterpiece (DC heated SE) I have definite experience with ghosts in the machines that are really happening. You have designed small variations in the circuits, optimizing the listening results within the frame of your listening environment (your speakers, your amps, your RF surroundings, your ears). In the T&C book you mention for example that you add a bypass cap in the NFB connection of the push-pull OTL not by textbook considerations, but by listening and also judging from the oscillator image - because you know what makes an PP OTL sound good. Now this sounding good is really the ghosts having their party - and attendance at that party is totally dependent on who was invited by the framework of the system (speakers, amps etc.). So letting bias drift, and thereby fattening the output may be just the right thing in your environment, but in other environments, a finer output trace may be the optimum. Then again, in the PP fattening and thereby shrouding the crossover distortion may be the thing to do, I don't know. I would also like to thank your for your work and your input and or being the host of such a civilized forum where great contributors meet. Cheers, Achim
Reply with quote #14
I think we should wait (or try by themselves) until more field experience with PP 300B is published. The working points in PP vs. SE are very different and also in PP the tubes operate under way more dynamic conditions, electrically, and thermally as a consequence This may average those tiny differences if they existed here.
Reply with quote #15
Putting a bypass cap on the feedback resistor is common place in all tube amps. Only the cheapest amps leave that cap. out or possibly bad designers. Bruce used an Oscilloscope to determine what size to use. In the old days it was done by trial and error. So what is the problem. All amp. designers [voice] there amps. The cap. is a small tweak. Pull up an old Dynaco schematic and you will see the cap. on every amp except on the SCA35 where cost was a factor. Other than the fact Bruce builds the best OTL's, I think Bruce's amps are a step above all others in the driver stage of push pull amps where he controls the voltage to the plates. Few of the commercial amps out there will put that many parts into the driver stage. But it makes a big difference in detail.