Triode connection for Pentodes (1/3): in-depth look plus alternative triode connection!

Since 4 days I've used a different connection which I think is more 'triode-like' than the usual triode connection for tetrodes and pentodes. In this video I explain the limits of triode connection, where non-linearity comes in and what changes I made to my amp.
See also 2 music clips which are posted separately so you can hear the difference of the new setup compared to earlier posted clips:
• Soundclip after improv...
• Soundclip after improv...

[@RWBHere]

You could also use a beam tetrode (kinkless tetrode) for significantlylower distortion and better power efficiency than a pentode. At around 10:00 to 11:00 I think you are talking about the infamous tetrode kink, which introduces the distortion which you are trying to limit. It's a kind of negative resistance effect, which varies with the valve type.

[@leandrocarg]

Once the diode is forward biased, ac only will see its dynamic impedance, remember your ac rides on top of the DC bias so the diode is always forward biased unless you reach cutoff but at that stage the whole tube will not be conducting.

[@EduardBroekman]

You may well be right there. I dropped this setup just because it didn't improve the audio anyway.

[@loureda5443]

Hi Eduard remember starved DHTs should only be used with thoriated heaters it may damage oxide types so be careful.I like the 01 , 26, 12a and I am not sure about the 3a5. While I have the highest respect for Sakuma san I do not adhere to all his policies in his transformer coupling concepts. Firstly his choice of Tamura iron with frequency responses of 30 to 15 khz is underwhelming and his choice of high step up ratios transformers tend to roll things off further. Metal oxide resistor use and electrolytic caps are also a no no in my experience. Do not get me wrong he is still in the top designers of his time and is to be revered as a genius.I like the 1 : 1 interstage approach with split secondaries as it enables me to drive my output stage in parallel single ended mode or push pull mode with the addition of a push pull out put transformer if I choose.It also makes the best phase splitter given it offers the same impedance and frequency response simultaneously with no addition tubes and sockets and resistors and capacitors. Cheers Lou

[@EduardBroekman]

Hi Lou. Good point although I still run these DHTs within specs... hope that works out! Thank you for the tip on 01 and 12a. Not having heard Sakuma's work in person, I can only speculate but with him 'hifi' was just not goal when he talks about bringing a highlight to music which he compared to Rembrand's plays with light and dark.. and so that inspiration and intent I can imagine leads to interesting choices other than what anyone would do. Otherwise, I totally get your points on his choices, especially the bypass caps which influence the overall sound quite substantially. I tried some Japanese mox for a crossover and there they seemed fine but I've never used them otherwise.. in the budget price class, I tend to prefer high wattage wirewound resistors.. I have no problem using 6W or 15W ones where others will use 1 watt ones.. low current seems to work well in preventing colouration. To jump back...with a 'poor spec' Tamura.. with him using very low Rp tubes as drivers (before ITs).. he'd get a shitload of bandwidth out of them even if the 'standard' specs look poor. A 7900 Ohm Rp for a type 26 has a much higher -3dB than a 4P1L with an Rp of 1200... on the same IT. Or was that spec within in circuit? And then I also remember he was taming Lowthers, Denon 103 cartridge and only used amorphous core.. all known for brightness...anyway all to say is that without considering the whole setup, I hesitate to cast 'out of hand' judgements as on the whole they may just work (aka piss-poor for a component by itself, but super well judged in a whole chain?) Since I've been using Lundahls.. I'm sort of used to having split primaries/secondaries :) but never switch between SE and PP that way (and technically my setup is differential, not P/P.. the return loop is different from PP). And I do love having perfect symmetrical split using ITs + very easy to implement some fixed bias. Do you actually end up switching a lot between SE and PP?

[@loureda5443]

Hi Eduard, I use both pentodes wired in triode and DHTs in my system. I use each for the desired effect.There are pros and cons to each, obviously the triode wired pentodes I use for high gain in my phono and line stage. They give me a level of quite over DHTs and the use of fewer stages over the DHTs as well.What are the cons well a triode wired pentode will never be as linear as a DHT. Advantages of those old DHTs are only 3 elements spaced widely apart, as well as a widely spaced grid element so the electron stream is less altered and more like a diode in its linearity.Just look at a 10, 211, 845 ect, they sound amazing and are some of the most linear devices made by man. On inspection of these tubes it becomes apparent why they sound so good. Unfortuneately they have low gain compared to the triode wired pentode so there is no free lunch. As far as the larger screen resistor I found in the past I prefer just the plain 100 ohm value. I do not know if there is a correlation between the sonics of a larger grid resistor and the use of a larger screen resistor. I find the larger grid resistor 1k and above to sound less extended and transparent as compared to 200 ohms or so. I will have to experiment with the screen resistor value to see what happens. By the way you can lower the distortion of a thoriated heated DHT by lowering the heater voltage slightly in case you were wondering..I have to go. Happy holidays Cheers Lou

[@EduardBroekman]

Hi Lou!! So you used the classic setup a triode connect pentode as receiving tube then? I was also experimenting with a fixed voltage on the screen, one advantage is that the gain can go up a lot.. but while it improves some sonics, I found it musically overall less appealing.. so makes me think SRPP is also a no go :) However I do like this diode/high resistor setup. And yes, I run all of them slightly 'starved' on the filament (-5%). The next thing I will try tube-wise is the go with a type 26 and replace the 2P29L (which is very, very nice sounding pentode and extremely affordable and easy to power with

[@loureda5443]

Hi Eduard, I need to order a push pull output trans to switch between the two output configurations and I do not have it yet.On cathode bypass I ended up with either Black gate non polar or audio note Kassei non polar caps in a super e configuration with a film bypass over those for more resolution.I am sure you realize that using Transformer coupling as I do I have to use a bypass if I want to keep the plate resistance low towards the manufacturer specs. On the plate resistor front I prefer a transformer or choke first then non inductive wire wound of decent wattage or Tantalum because of its temperature stability. For grid resistors I use tantalum Shinko or Audio note or carbon compositions or films , same with cathode positions for low power tubes. There is no resistor in my experience that works everywhere. Horse for the course so to speak. With regards to Sakumas work I have issues with his high step up ratios in his amplifiers which will severely limit bandwidth and life.I love Transformer coupling for its texture and dynamics but if I use high step up ratios it limits extension and transparency so I avoid high ratios . This applies to everyone using step ups from cartridges to amplifiers.Having said this I have heard a couple of you tube videos that Sakuma did many years ago of his system playing some Doris Day and it sounded very realistic. What do I mean? it sounded like Doris Day was singing in his restaurant where his system resided it was way beyond anything on you tube. But I could hear some extension limitations in the higher frequencies As far as the difference bandwidth goes my money is on the 4P1L over the 26 as you only need a 5000 ohm primary with the 4P1L as it is only 1200 ohms. The 26 would need at least a 20000 ohm primary which means greater inductance which in turn means more high freq roll off. The 26 on the other hand will probably have more tone and ease to its presentation by comparison. Cheers Lou .

[@EduardBroekman]

Hi Lou. Great to hear from you again. With regard to the bypass, I've circumvented that by using filament bias on the 2P29L so the bias resistor is about 40 Ohms there, it's plate is about 2900 Ohm and the interstage 43H although I could upgrade it to using two interstage transformers, non-splitting with 130H.. so then there's really no roll off. At the differential output stage the 4P1Ls (chokefed) share a common cathode in the form of a diode tube, which doesn't need a bypass either and again is pretty low value impedance anyway.. and the 1200Ohm plateresistance is met by a 125H transformer ... so I've gone all out on avoiding caps and keeping resistors to a minimum. At the some point I do want to take that approach more all-out with real triodes, better chokes, maybe even parallel feed.. but keep this course. The only issue by going low-mu, low Rp in the output stage is the higher bias requirements and I may have to resort to tube-diode and fixed bias. Not sure just yet. The high ratios, yes every step up quadruples the input capacitance, so -depending on what's driving it- I'm sure that roll-off would come into it. I'm working on a headphone tube amp and I am experimenting with using a 1:4 step up driving the 4P1L just to see if I can get away with a 1-stage DHT. I had previously done a 1-stage 4P1L differential amp driven directly by 4V RMS XLR connection and that amp had 90% of the imaging and clarity of a single ended amp.. so it seem worthwhile to try it and Jacmusic has a similar design with Lundahl stepup and 20B true DHT (20-22mu I think) so this has my interest. I'll have to find that Doris Day clip :)

[@hpfctif7tx7t]

Can you do a part 2 where you explain on how to calculate the resistor of the screen grid?

[@EduardBroekman]

No I won't but you can simply applies ohm's law to limit the current to the datasheet's specs.

[@tmomedrano]

just tried this.. seems like the bass is a bit bloomier. did it just convert a triode strap to a pentode through decouling with a diode instead of a cap

[@onemintjulep]

Very interesting. Pentode question : Does the suppresor grid share the negative ground with the cathode or is a resistor in between the two to offset deflecton ?

[@EduardBroekman]

It's connected to the ground. Have you seen designs putting a resistor in?

[@onemintjulep]

@@EduardBroekman no i haven't but i don't understand how electrons are relased from a screen of zero volts if the cathode initially releases them at zero volts. wouldn't the elctrons get clogged at screen with zero volts ? I don't expect you to explain particle physics in YT comments . ty

[@EduardBroekman]

@@onemintjulep I'd say the electrons being 'boiled off' at the cathode and pulled by the electric field set by grid2, pulls electrons in from both ground/cathode resistor and the screen. Hence, I don't see how they get clogged at the screen having a direct connection to the cathode which pulls them in.

[@user-pq3bz2jh8l]

Parasitic oscillations