Two things to note: Many claim the 6DJ8 and 6922 are not the same tube. Both of my old school printed manuals say this for the 6922,
"Industrial 6DJ8. See 6DJ8." That's it. Also, if you are willing to use roughly equivalent but not exact tubes... look at the 2C51/396A/5670 tube family as they can be swapped for 6DJ8 with a pin adapter. There are loads of NOS tubes available as they were popular in aviation/military - I have recently picked them up NOS for less than $5 US which is cheap compared to USA/Euro 6DJ8s. I have used both families for audio and found great examples of each pretty easy to come by if you spend $ for NOS tubes from a trusted vendor.
"Industrial 6DJ8. See 6DJ8." That's it. Also, if you are willing to use roughly equivalent but not exact tubes... look at the 2C51/396A/5670 tube family as they can be swapped for 6DJ8 with a pin adapter. There are loads of NOS tubes available as they were popular in aviation/military - I have recently picked them up NOS for less than $5 US which is cheap compared to USA/Euro 6DJ8s. I have used both families for audio and found great examples of each pretty easy to come by if you spend $ for NOS tubes from a trusted vendor.
The 6n2 don't fare so well there do they?
When I put some 6n2p-ev into mine, it was instantly better. I wonder if thats worth a try if you have some?
On my earlier post with the hand drawn schematic - yes I seem to have missed the grid connection from the upper triode to the plate of the bottom triode!!
Well caught!! I traced that out some time back - just to sketch out the voltages, and think about maybe an LED bias on the bottom triode etc. Also to see exactly what the typology was. I've tried other Aliexpress boards that say they are SRPP and in fact are straight triode single stage etc.
When I put some 6n2p-ev into mine, it was instantly better. I wonder if thats worth a try if you have some?
On my earlier post with the hand drawn schematic - yes I seem to have missed the grid connection from the upper triode to the plate of the bottom triode!!
Well caught!! I traced that out some time back - just to sketch out the voltages, and think about maybe an LED bias on the bottom triode etc. Also to see exactly what the typology was. I've tried other Aliexpress boards that say they are SRPP and in fact are straight triode single stage etc.
Every tube should be used/biased properly with different parts values , even supply voltage . There is not such a "universal circuit" where you can just trow them for comparison . Of course will kinda work , but not at the max performance .
If you got that kit you should decide on what tubes you will use and optimize the circuit for them
If you got that kit you should decide on what tubes you will use and optimize the circuit for them
Thats absolutely true. Especially the 6N2P is quite different from the rest as it is biased with only -0.7V due to the 470R cathode resistor. This tube needs to run at much lower anode current than all others mentioned. And with only 150V anode voltage, it is difficult to find a low THD biasing at all (with -1.5V grid bias, only .5 mA anode current possible with 150V anode supply per valve).
The other 6DJ8 like tubes mentioned are all running with 5 to 10 mA in this circuit and the 6N6P is running at 12 mA.
The other 6DJ8 like tubes mentioned are all running with 5 to 10 mA in this circuit and the 6N6P is running at 12 mA.
When it comes to 6N2, i made no adjustments to the circuit. This is how it was supplied as kit. As you can see from the pictures. If the kit maker did not adjust set points properly for this tube, than that is the reason. It it what it is. I measured the kits as i got them. Too bad for 6N2.
On the other hand, 'what the heck is 6N11?' turned out one of the most linear tubes in this circuit.
I may post some spectra later. I will listen to each set of tubes and see if i spot any characteristics in sound.
On the other hand, 'what the heck is 6N11?' turned out one of the most linear tubes in this circuit.
I may post some spectra later. I will listen to each set of tubes and see if i spot any characteristics in sound.
Last edited:
Not a specialist but asked a lot to very experienced folks for an I/V tube DAC project. The Chibese is far to be as good as the russian and european/US parts number above.
The best bank for the buck while sounding very good is the 6NP23-EV but from the plant with the O with two bars - don't remember the name- version in NOS, easily avialabe but not pins compatible with the 6922/E88CC/ECC88. The best of all here being the E188CC Philips made according many enthusiasts : triode are matched within 5% and it's microphonic construction. 6DJ8 is said far better than the chinese's tube of your list and also subjectivly pleasant.
That's because they are not.
While their plate curves are nearly identical, their maximum ratings are very different. Also, because the 6DJ8 was designed for VHF cascode operation, its two triode sections are not identical. The 6922, being a more rugged variant, is more suitable for general purpose applications in my opinion.This was discussed to some degree in another thread.
https://www.diyaudio.com/community/...sistor-on-6922-gain-stage.379575/post-6856475https://www.diyaudio.com/community/...sistor-on-6922-gain-stage.379575/post-6856478
Maybe I should be more explicit. The (quite a few) American 6DJ8/6922 tubes I have and have used over the years of a few manufacturers: test the same in tube testers, different circuits of my own making and Audio Research gear I've used them in. No difference in the sections at all and neither section disliking running with some high heater to cathode V. They look identical physically, down to even the paint on the labeling aside from one says one vs the other. The brands sound different, the same brand 6DJ8 and 6922 sound identical. Based on my real world experience, my 2 different American tube manuals saying they are identical... the only thing I can think that separates them is the 6922 was premium selected, and I can't say I've experienced anything that would indicate that is true. The range they show in tests straight out of the box is the same. Maybe the Europeans ones are different, or very early examples, but I am certain beyond any doubt, all of the ones I have are made to the same specifications.
I won't try to comment on how sellers acquire their tubes and brand them. I don't doubt that some vendors will take the same tube and mark some as 6DJ8 and others as 6922. But true 6922s and 6DJ8s are different. The discussion in the links above point out the significant differences.
Even if their plate curves are virtually the same, their maximum ratings are not and that has to be considered as part of the design and when selecting one or the other for a given application. But everyone here is free to make their own choices.
Even if their plate curves are virtually the same, their maximum ratings are not and that has to be considered as part of the design and when selecting one or the other for a given application. But everyone here is free to make their own choices.
I don't want to argue but differencies are minor ... if the curves are the same . Some max ratings are different , after all E88CC is a special quality tube . This happens all the time in soviet tubes , standard and -EV sufix .
The sections are identical for audio purposes , for RF cascode you have to use them as required .
The sections are identical for audio purposes , for RF cascode you have to use them as required .
I really don't disagree with anything you've said. At audio frequencies the two tubes behave pretty much the same. But I do think that the maximum ratings differences can be important and that they are worth being aware of. For example, I wouldn't want to plug an NOS Sylvania 6DJ8 with its 130 volt Vp-k rating (datasheet attached) into a circuit that puts 200 volts on the plate, which would be within the limits of a current production JJ E88CC (6922) (datasheet also attached). [As an aside, I realize that the 130 volt "design center" rating for the 6DJ8 might not actually be its "maximum" figure, but I tend to interpret ratings conservatively.]
When I design a circuit using the 6922 which might be built or used by someone else, I intentionally stay within the limits of the NOS Sylvania 6DJ8 (using a Vp-k of 90 to 100 volts, for example) in case that user decides to "tube roll." As I said, I admit to being extremely conservative in my designs, so these differences mean more to me than they do to others. But I'll probably keep doing what I'm doing, and so will everyone else.
Cheers!
When I design a circuit using the 6922 which might be built or used by someone else, I intentionally stay within the limits of the NOS Sylvania 6DJ8 (using a Vp-k of 90 to 100 volts, for example) in case that user decides to "tube roll." As I said, I admit to being extremely conservative in my designs, so these differences mean more to me than they do to others. But I'll probably keep doing what I'm doing, and so will everyone else.
Cheers!
The characteristics that vary between 6DJ8 and 6922 are:
- 6DJ8 section with pins 1,2,3 has max cathode-heater voltage of 130V
- 6DJ8 section with pins 6,7,9 has max cathode-heater voltage of 50V
- 6922 has max heater-cathode voltage voltage of 100V for both sections
- 6DJ8 has max plate voltage of 130V (although that might be a conservative spec)
- 6922 has max plate voltage of 220V
- Max plate dissipation of 6DJ8 is shown as 1.8W.
- Max plate dissipation of 6922 is shown as 1.5W (but Philips data sheet says, "Max anode dissipation of section 1 plus section 2 = 2W").
Some 6DJ8 have heater current of 365mA as compared to 300mA for 6922, but when measured some 6DJ8 heaters measure 300mA. They're probably not different in this regard.
6DJ8 and 6922 are essentially the same, with some differences which are minor in most applications, but could be important in certain circuits (especially the max Vp spec, and some cascode or other totem-pole configurations where the top tube's cathode voltage is far above ground).
(I used Sylvania data sheet for 6DJ8 and Philips data sheet for 6922.)
Yes, you did. Because the Sylvania manual says they are the same tube. And I can assure you all my Sylvania ones are indeed, the same.
I have thousands of dollars worth of these tubes new in the box on shelves - and I've tested most of them. A small slice is commercial boxed but most are JAN tubes. If they all test and perform the same and pass the testing, QA and such to be boxed JAN in the first place, I think we can accept they are.
Not to mention there are are discrepancies and mistakes in tube data sheets sometimes, they aren't always right. They also were never meant for the engineers not to do their due diligence and try and test their circuits. There are also plenty of examples right here on this forum, where people have noted improvements in tubes parameters and/or construction that were never reflected in the data sheet and a new name or suffix was never designated. There sometimes simply happened as matters of economy; creating common manufacturing techniques across multiple tubes or create a new version of 2 similar tubes that can be manufactured and sold as both. This is not unique to the USA manufacturers either.
I have read the above discrepancies many times. But I don't see them in real life use, and I don't see others having those issues either. It's debated endlessly, but DIYers and manufactures have been swapping both (and the listeners!) for decades and without issues.
As I said before, it is possible there are examples out there that are indeed slightly different. But if there are, it is surely in a small minority or it would be known well in the community as fact - instead of this continued debate. I say knowing everything in this thread, test your own tubes and make up your own mind as I have. And have fun building and listening.
Since I felt bad for poor 6N2, I did test the distortion from 100 volts supply to 200 volts, in 20 volt increments. There is smiley effect, distortion going down, then up. I will post results when I have a chance. I made no mods into the circuit, except used variac for HV part. Heater part was separate and kept as before.
