So I’ve been reviewing EL34/6CA7 data sheets preparing for an EL34 SE-UL optimization study like the two I’ve done before for the 6V6 and 6L6. However, I’ve run into a little snag.
Choosing what I think will be the best SE-UL load impedance is a combination of calculation, estimation, and intuitive feel. The DC bias points for UL optimization come off the triode strapped characteristics for a given tube. This is because at DC, UL topology is really just triode topology with a small resistance between the plate and suppressor. But the AC operation is a compromise between the triode and pentode operating characteristics. One way to estimate this is to plot a DC load line on the triode characteristics and then estimate the how much the plate characteristics will straighten and shift to the left under UL operating conditions.
The long and short of all this is that I have determined that the load impedance I want to use for the EL34 SE-UL optimization study is 2.5kΩ. And this is where the problems arose. When I went digging through my parts stash I discovered that I currently have only four output transformers available; all 5kΩ primaries. These are all Edcor GXSE10-5k 8Ω transformers.
The astute reader may point out that I could just terminate these transformers at 4Ω and then the effective primary impedance would be 2.5kΩ. And I did consider that option. However, whereas the impedance transformation would be correct in this case, the core magnetization characteristics and parasitics would not. Since the study itself requires a significant amount of testing, measurement, recording, and analysis I decided that I really didn’t want the added questions hanging over my head due to the suboptimal output transformer load. So until I can get Edcor to send me a pair of GXSE10-2.5k transformers, the EL34 SE-UL study will have to wait.
But wait… there may be another option in the short term.
The thought occurred to me that I’ve been doing these studies on big octal based tubes and it might be nice to perform an optimization study on a smaller 7 or 9 pin “miniature” tube. There are lots of candidates in this class. I can take the 6AQ5 off the table because it’s just a lower powered 6V6 and the data for that study can be used for the small tube (pursuant to the latter’s power dissipation limitations). But that leaves some other tubes I could use with a 5kΩ load.
One candidate is the 6AS5 beam power tube. This is a small power tube originally designed as the audio output amplifier for automotive radios. It’s a low power tube with a maximum plate dissipation of only 5.5 watts. Another candidate is the always ubiquitous 6BQ5/EL84 power pentode. This is a higher power tube more akin to the 6AQ5 with a maximum plate dissipation of 12 watts. And a third candidate that I found while search through my stash is the 6CL6 power pentode. This is a pentode designed primarily as a video output amplifier in televisions. It is a low distortion wide band amplifier tube. The 6CL6 has a maximum plate dissipation of 7.5 watts, between the two other candidates.
The question obviously is which one do I choose. All will optimize well with the 5kΩ primary output transformers I have on hand. The 6AS5 and 6CL6 are both readily available on resale sites and eBay for very little money. The EL84 is widely available but more expensive than the other tubes.
My first inclination was to choose the EL84. But two things gave me pause. First was that EL84 designs are numerous, widely available, and widely varied. In short, the EL84 has been “done to death” on the internet. Both as a basic audio amplifier and in guitar amps. The second was that the 6AQ5 and 6BQ5/EL84 are so close in specifications and performance that I am not sure an 6BQ5/EL84 optimization study will provide any great additional knowledge or insight verses the 6V6 family optimization data set.
The 6AS5 is admittedly a lower power tube, but I’ve always wanted to see what I could do with it. In the 6AS5 µ-Power UL Amp I built over ten years ago, I don’t think it ever produced more than a watt per channel and I know for a fact that the operation was not optimized in any way. It would be nice to revisit this tube and see if I could do it justice.
The 6CL6 is a different kind of beast. It is a true pentode but it was specially designed to be low distortion and to have very low parasitic capacitances so it could be wide band. With it’s higher plate dissipation limit I could probably get maybe up to 2 watts out of it. It might make a really nice example of a tonally neutral lower power amplifier for a small room or office.
The other benefit that both the 6AS5 and the 6CL6 have over the EL84 is the lower power supply current required. This limits the size and the expense of the power supply and case.
So does anyone have any opinions about the path I should take? Should I just wait for my EL-34 transformers from Edcor (probably March or April)? Or should I choose one of the small tubes? And if so, which one? I’d like to know what people think.
As always, questions and comments are welcome.
Pingback: Minor Work Stoppage | Cascade Tubes
Let me go look through my tube stash…..that will make my choice 🙂.
In all seriousness, I think I understand the arguments for EL84 (which I have in plenty) but my inner nerd personality likes “the odd”..
A quick look at my tube inventory says I have no 6CL6 or 6AS5 (which surprises me). So I have to choose 6CL6 as a second choice because I know my friend has some. Very scientific process here…..
happy EL 34 is still a go… i vote for the EL 84.
A little “tongue in cheek” here:
Both the EL34 and EL84 were weapons of choice in the “British Invasion” of the early 60’s.
As a guy who many years ago didn’t quite know what he was doing yet still managed to cobble together an EL84 PP UL amp that (to my amazement) easily went head-to-head with commercial gear costing 10x as much, I’d be very interested in your rigorous scientific take on all things EL84.
Any of those options would be interesting to me but those tubes that haven’t been explored much or if they had, their operation could be improved, might be preferable. Some time ago while designing the 6AS7, you mentioned also the 6DQ5 which also looked like an attractive tube–https://www.cascadetubes.com/2021/07/24/another-option/
Casting a vote for the “clickbait” crowd, I say KT88. It is a tube people choose for power, and I have never seen a scientific study that shows how to get the most clean power out of the tube. People not familiar with this site might visit to see your results.
I started to do a KT-88 SE-UL optimization about a year ago. But I rapidly figured out that I was power supply limited. My bench power supply only goes to 400V and has a 100mA current limit. I might be able to perform the UL study within the voltage limit, but to really get the most out of the KT-88 in UL mode, the total cathode current needs to be between 100mA and 190mA.
My current target for the KT-88 in SE-UL is a 1.7kΩ primary output transformer, Vp≈280v, Ik≈150mA, Vg≈-30v. I need a higher current bench supply before I can test around this operating point. Power out in this configuration is around 14W.
I didn’t realize that 1.7kΩ was the target for the OT. The Mikael Abdellah & Allex Gendrano designs on the DIYAudioProjects page used a 5K OT.
14W is an impressive goal. That result would be around double your 6L6 Spalted Alder amp. You could certainly use “Optimized” in the blog title.
hello Matt:
I think you brought up a very good point about the internet over promoting the same old tubes. So for sure I would move away from the 6bq5 it has been done before too many times.
One of the things I really like about your site is that you focus on interesting tubes that are a little less well known, yet still are available and perform very well. You have already explored low power output tubes very well. I was hoping for something in the 3-5 W output range.
I mentioned to you once in the past that I am currently exploring compactron tubes. A somewhat popular one is the 6lu8. A medium mu triode and a pentode in the same bottle. The pentode section has a dissipation of 14W. This may be a candidate worth exploring, not sure if it optimizes to a 5k output transformer though. There are many others as well some up in the 30w dissipation range.
Otherwise my vote is for the 6cl6. Something uncommon.
Thanks
I built the 6AS5 Micro-Power UL amplifier some time ago, I have always liked how it sounds. Even now 6AS5 tubes are still a bargain- I checked eBay and they are available in the $2-$5 range. Certainly, I’d like to see your optimization study for the 6CA7 / EL34 ; but in the meantime I’d say let’s give the 6AS5 tube a serious optimization study- I’d be very interested to see what could be done with it.
The 6BQ5/EL84 is a current production tube – a very big plus! Also who says that the 6BQ5 amps that flood the internet are truly optimized? There are a lot of “tinkerers” out there that slap something together that sort of, kind of works. You, on the other hand would use a methodical scientific approach to arrive at the optimal operating conditions.
As far as the 6CL6 goes, there is a kit out there called “The Spud Kit”, which uses one tube per channel to get a claimed 2+ WPC out of it operating as a single ended pentode. But, I am not sure that the 6CL6 is currently in production.
My vote would be for the 6BQ5/EL84 study.
Just my $0.02 worth.
My thoughts exactly!