Clementine ULtra

[Shannon Parks]

Here are my basic Clementine ULtra notes. I can add more commentary later. In a nutshell, I discounted an ultra-linear Clementine due to the high plate impedance thus extremely poor damping and high ampifier Z. But with 6dB of negative feedback, we gain parity with the triode configuration plus we get more output power as the mu is higher in an ultra-linear configuration. Damping factor was measured at 5.2 (UL without feedback is less than 1 and selective distortion wreaks havoc) and the distortion was mostly 2nd harmonic even at max power - key triode 'signatures'. Read more about this in RDH3 pages 16 and following. Note how beam tetrodes are superior to pentodes for this application. The only negative in the ULtra is the high input voltage needed - 1.6V for 7W - so a preamp is needed and an iPod will not drive it.

Mods from the Clementine 6L6 Build in the main thread:
- Remove R15 and R16 (no triode strap resistors)
- Connect ultra-linear taps to J3-2 and J4-2
- Use connectors at J10 and J11 or solder directly to the pads. I used CDROM audio cables since they are cheap shielded cables. The yellow output secondary (+) goes to J10-1 and J11-1 respectively. The white output secondary (-) goes to J10-2 and J11-2 respectively.
- R27 and R28 are 680 ohm resistors with 6800pF (i.e. 6.8nF or .0068uF) strapped across them. I am using cheap ceramic discs. A good temp rating or voltage rating is not important.

You are then done, and have about twice the power. It sounds pretty awesome, with the 3D 2nd harmonic is 100% there and no irritating dominant 3rd harmonic.

7W and 4W output power and distortion:
http://www.diytube.com/clementine/7wultra1640mV.pdf
http://www.diytube.com/clementine/4wultra1208mV.pdf

Shannon

[miker]

Very nice! I may follow this path with my second Clem board. What operating parameters did you settle on for the 6L6 and are you still using 6L6GA's?

[Shannon Parks]

These are biased up at 50mA and dissipating around 18W. I'm technically at 706 ohms due to what I had on hand, but suggest the parts list 720 ohms cathode resistor network. I'm still using the cheapo Russian 6L6GAs. Probably can squeak out more power with EL34s with no circuit changes, and I intend to test them for harmonic content.

Shannon

[carpenter]

Hi, I'm a bit confused:

Are you saying that J10 and J11 are being jumpered to one another? The specific wire color is throwing me, makes me think there's another component involved.

To me the mod reads: use two wires that are shielded to couple J10 to J11 (noting number 1 and 2 connection points).

Am I right?

I don't know where J10 and J11 belong on the schematics. I've been reading both 1626/6L6 versions. Is there another schematic variation on this site?

[Shannon Parks]

Indeed, it is a bit confusing. I'm working on a drawing and will upload it it the morning.

Shannon

[carpenter]

Thanks Shannon. I'm seriously considering this series for my 100hz mid bass horns. They may not be as efficient as my paper horns, but they might fair quite well with seven watts.

The 1626 Clementine drives the paper horns with ease, creating breathtaking music in the process. I'm so pleased I discovered this forum!

[carpenter]

These are biased up at 50mA and dissipating around 18W. I'm technically at 706 ohms due to what I had on hand, but suggest the parts list 720 ohms cathode resistor network. I'm still using the cheapo Russian 6L6GAs. Probably can squeak out more power with EL34s with no circuit changes, and I intend to test them for harmonic content.

Shannon

What became of the EL34 concept? Seems like a rather interesting approach.

[Shannon Parks]

I've uploaded a new Clementine manual with the last three pages being ULtra notes: http://www.diytube.com/clementine/clementinemanual.pdf
Let me know if this clears up the confusion.

The Clementine 6L6 supports all 6L6-types, so 6L6, KT66, KT88, 6550, KT90, KT120, etc. EL34s require a jumper at V2-1 to V2-8 as well as a jumper at V3-1 to V3-8 (very easy - you can use clipped resistor leads). I think it would work very well for your application.

Shannon

[carpenter]

Shannon, you're time is valuable, and it's such a pleasure when you share you insights. The added information wiped away all the confusion. I can't thank you enough! I'll be ordering another clementine PCB :)

[coolhandjjl]

If I understand correctly, if wired in Ultra, negative feedback is added, but it is okay as the 6L6 is a beam tetrode and produces just 2nd order harmonic distortion and no 3rd order distortion. Benefits, higher 7wpc output.

What the is output wired normally? Is it a strapped triode mode?

And if built Darling style, it is very low output, 3/4 watt.

[Shannon Parks]

If I understand correctly, if wired in Ultra, negative feedback is added, but it is okay as the 6L6 is a beam tetrode and produces just 2nd order harmonic distortion and no 3rd order distortion. Benefits, higher 7wpc output.

The negative feedback is needed due to the higher plate impedance in a ultralinear connection. It is just enough to raise the damping factor to around 9, or basically a little higher than a good triode (5 or 6). The higher output is a result of the power tube having higher gain in UL vs triode. I briefly FFT'd the amount of 3rd order harmonics, and it is still far from dominant. Note that the extra gain power is related to the limiting grid bias voltage, and that the ULtra total gain will be less than a standard Clementine.

What the is output wired normally? Is it a strapped triode mode?

Correct.

And if built Darling style, it is very low output, 3/4 watt.

Correct. It requires high sensitivity speakers. But I can rock a huge room with my Klipsch Chorus IIs and Darling - it's pretty cool.

Shannon

[carpenter]

Shannon, this is a crazy question, but I'm new to tubes, so it's worth looking foolish...

Would the 813 tube work with the Clementine board when set up as a triode, or perhaps ultra mode? I've noticed amps on the net using this tube and the power excites me.

[carpenter]

I read a bunch of websites. Because separate power voltages are involved, it appears the the clementine would make a great driver board and the 813 could be wired point to point with ceramic sockets and its 1000 volt supply. Any suggestions, anyone?

[Shannon Parks]

Some of these transmitting tubes like the 813 can be used in SE amp projects, but due to the much higher voltages involved I've never messed with them. I certainly wouldn't recommend going that direction just yet.

Shannon

[carpenter]

I can understand your concerns and agree.

I noticed an interesting power valve, the 6C33C, which requires far less B+ voltage. It's driven by tubes similar to those on the Clementine. That seems like a safer concept.