In this second installment of the tube of the month series I'd like to present the tube which I chose as driver in the single ended 6CB5 amplifier, the 6N7:
Another tube which is mostly ignored by the audio crowd, although it is actually quite well suited for audio purposes. It has a medium amplification factor of 35, a plate resistance which is still in a range which allows to use it in transformer coupled circuits or choke loaded. And most important: it is very linear. A detailed datasheet can be found here.
The tube has been developed for a Class B output stage. This is probably the reason why it is not so popular. It is a double triode. Both halves share a common cathode. Both triode systems need to be wired in parallel if the tube shall be used as a Class A driver. Again a practice which is not very popular among audiophiles. Unrightfully though. With both halves in parallel it can be used with interstage transformers like Lundahl LL1660. It has enough drive power to be used with smaller triodes like 45, 2A3, 801A and of course the 6CB5A triode connected. With it's reasonable high amplification factor it is perfect for a two stage amplifier concept. The 6N7 is available in various different forms. The basic type with metal envelope as pictured above. The 6N7GT has a straight sided glass bulb and the 6N7G has the beautiful coke bottle shape. This tube was manufactured by all the big tube companies and is still available as new old stock for very reasonable prices. Here is a selection of 6N7s from various brands:
The three variants with the different shapes are electrically interchangable. All of them are very nice sounding tubes. Don't shy away from the metal versions. The fact that it is quite cheaply available (about $5 for NOS tubes) doesn't mean it is bad sounding. Most people avoid it because the glow is not visible. The 6A6 is the direct predecessor of the 6N7, it is electrically identical but uses the UX7 base instead of Octal. The 53 is yet an older version which shares the UX7 base with the 6A6 but has a 2.5V heater instead of 6.3V.
The next picture shows some of the different shapes:
From left to right: 6A6 (this one was only made in the coke bottle shape), 6N7G, two 6N7GTs, one with graphite coated glass the other with clear glass and a RCA metal 6N7.
Another nice aspect of the 6N7 is the similarity of it's pinout to that of the 6J5, which is pictured on the right. Heater and cathode connections are at the same pins. The pins nr 4 and 6 which are used in the 6N7 for grid and plate connections of the second triode system are unused in the 6J5. A 6J5, which is basically one triode system of the popular 6SN7, will drop right into a socket which is wired for a paralleled 6N7 and will bias up correctly in a circuit which is sized for the 6N7. It will just provide less voltage amplification. The 6J5 has an amplification factor of 20. This will result in 3dB less gain. This might be just the right reduction if a system has too much overall gain.
So, if you are planning a project, check out the 6N7, it might be just the right tube.
Best regards
Thomas
Hi Thomas,
ReplyDeletecongrats for the new camera. Good choice ;-).
Cheers: Holger
Hi Holger,
ReplyDeleteyeah, I thought my blog deserves some better pictures than those from a crappy cheap camera ;-)
Thomas
Nice tube... I have seen it servicing as a microphone input stage, phase splitter and output tube. Very versatile!
ReplyDeleteCan anyone tell me why triodes can not be biased into a class ab2 push pull arrangement.
ReplyDeleteThey always are! Class B strictly means zero quiescent current and that is never how push pull audio tubes should be biassed as it causes too much crossover distortion. There is an optimum quiescent current that minimizes open loop crossover distortion. You have to do a bit of computer work to find it, as determining the characteristic curve with both grid and anode voltage varying is complicated by hand
DeleteHi! Who said they can't???
ReplyDeleteThe 6N7 was designed for PP Class P service. I have no experience how far it can go into AB2. Try it out.
Best regards
Thomas
Given 6N7 is your favourite indirectly heated driver for 300b, would you please comment why do you prefer it to something like D3a for example? Do you thin 6N7 sounds "better" or is it your designer statement not to use some obvious choices. Thank you!
ReplyDeleteHi! I do not recommend the 6N7 as driver for the 300B. I use it with smaller triodes like 45, 801A. In my experience ether 300B needs a driver with lower plate resistance. Although the D3a seems popular as driver tube, I don't like it as diver. I think it has too little headroom. See my post gain, headroom and power for my thoughts about this
DeleteThomas
Ah, sorry. I've thought that you had a 6N7-300B amp. Maybe I mixed it with someone else.
ReplyDeleteHi!
DeleteI did build a 300B amp which could use the 6N7 or 6J5 as driver. My preference comes from that amp. I preferred the 6J5 there
Bets regards
Thomas
Many thanks!
ReplyDeleteHello Thomas I pushed preview and here I am.
ReplyDeleteAnyway I would be interested in a schematic of 6n7/45 se
Thanks, Bill Larson
Hello Thomas
ReplyDeleteDo you think this tube will work well as a concertina phase inverter? The gm looks pretty good.
Hi!
DeleteI never built a concertina phase inverter, so have no practical experience with it. But I see no reason why the 6N7 would not be usable as such
Thomas
Hi Thomas,
ReplyDeletemay I ask what you think if there are some reasonable contraindications for using 6N7 in the INPUT stage of phono preamps, except for microphonics and common cathode? I asked numerous people, but yet nobogy gave me any intelligible answer.
I would very, really apprecciate your opinion.
thank you, Alexander.
Hi!
DeleteWhy don't you simply try it. Probably nobody has used it as input stage in a phono and that's why you don't get any sensible feedback on your question. I have used it as second stage in a phono with success
BR
Thomas
yes, I saw that you did, that was the main reason I asked, I`ve heard pretty many industrial examples where it is used in a second stage, all sound wonderful, so I thought maybe it`d be as good in a 1st stage? all whom I asked are murmuring something indistinct about microphonics and other, but nobody has ever given ANY reasonable, "scientific" answer, why one shouldnt` do it- like resistance, amplification (too much/too low) etc.
ReplyDeleteWell it depends on the gain you need and overall circuit if it is a good match. If you find it tempting simply give it a try, a bread board circuit is quickly done. This will give you actual experience.
Deletesure I will try, and report later.
ReplyDeleteHallo Thomas! Thank you for all the good stuff you share on this blog, lots of good info and tasty pictures :) I ordered a few 6N7 to try my hand on making an amp using them as outputs and am waiting for them to arrive. Any ideea on how much power is to be expected in SE? 3-4watts for one tube maybe? Freundliche gusse aus Rumanien!
ReplyDeleteSorry but these are not useable as outputs in SE. Plate impedance is too high and even if you could find a suitable OPT (which you won't find) you would probably get 100mW. As outputs they are only useable in Class B push pull
DeleteOups.. Based my asumption on plate dissipation figure and purchased away hehe. I will experiment with them as soon as they arrive, I am sure I will learn and have fun with them :) thank you for taking time to reply!
DeleteThis comment has been removed by the author.
DeleteMax plated issipation is only 1W per unit
DeleteI see it in the datasheet you kindly provided in the article. A quick google search tho yields a Tungsol datasheet which only mentions the 5.5W figure, I assumed it would be the same no mater the class.
ReplyDeleteHi Thomas, could you please comment on following topic - with all known differences on electrical parameters- if you would rate linearty of PARARELL connected 6N7 vs single (half) of 6SN7 triode? You say in your post (great text by a way, thanks for that) that 6N7 is "very linear". I am trying to understand how "very" by comparing to 6SN7:) I am ignoring working point, currents impedance so far deliberately as it's topic for later, careful design consideration
ReplyDeleteAs much as I understand that an easy answer would be desirable, you cannot rate this without the operating conditions. And especially the voltage swing you need to generate
Delete