butwhatif wrote:I took a look at my internal dlator, and was able to fix most of the one sided clipping , here's a scope trace of the before signal out of the CF and by raising the cat bias up to 32 volts I was able to get most of the clipping out. Actually, it had to go to 60-70vdc to pass a very hi clean sine wave, but stopped at 32v as that is where the output section clips.
That's exactly what mine looked like. I put in 1.8k/27k as suggested and my cathode is at 62V. It's completley symetrical now up to the point where the clean stages clip. I hadn't thought about the output clipping but that's a good point.
What values did you end up with to get 32V?
What tube?, with an AT7 64 is probably right.
My amps run a 1K8 and 27K and with 290 volts on the plate I hit about 33 volts.
Gary
Located in the St Croix River Valley- Afton, MN
About 5 miles south of I-94
aka K0GWA, K0 Glas Werks Amplification www.glaswerks.com
FWIW, I used 1.8k/33k. But my dropping string is different...I run both stages off of the same node. The supply is 330v, the CF cathode is at 44v, recovery plate is at 225v. I haven't looked specifically for any clipping in the CF on a scope but I haven't had any problem with the loop clipping.
This is an interesting topic, something I had overlooked with my current mod. There are probably many builds and mods out there that experience this artifact, given the limited schems available for dlators a fair number of projects would have it unless they were investigated with a scope, or had different loop architecture. For the most part, the standard schem works w/o a hitch if drive levels are low, but it's pretty easy to make it clip. I think it sounds better with the cleaner sig path. fwiw
The standard schematic (the one on schematic heaven) is wrong. The cathode bypass cap is a 47uf! I have double checked this with actual Dumbleators. I have only seen 47uf. Just throwing this out there for you guys.
butwhatif wrote:This is an interesting topic, something I had overlooked with my current mod. There are probably many builds and mods out there that experience this artifact, given the limited schems available for dlators a fair number of projects would have it unless they were investigated with a scope, or had different loop architecture. For the most part, the standard schem works w/o a hitch if drive levels are low, but it's pretty easy to make it clip. I think it sounds better with the cleaner sig path. fwiw
butwhatif wrote:I'll investigate this, thanks. Does increasing the 10k tail generally suffice to
rebias the cf stage, or have you found that the cathode resistor also needs to be increased?
Increasing the tail would probably do the trick, but I have not tried it. I would say increasing the cathode is still a good idea as it raises the potential difference between the cathode and grid a little bit.
I knew I would find some more info if I searched enough threads. The collective brain/tone trust of this forum is amazing.
I think I'm going to try 2k/30K and see what happens.
