6C33C OTL Amp Comes to Life
Posted 8th January 2013 at 01:45 AM by wlowes
This is the continuation of my personal journey to build my first tube amp.
It was a very happy day yesterday when the first mono bloc powered up and produced music. I was pretty comfortable that I had not made mistakes that would result in smoke, but really did expect some issues with hum, grounding or tube substitution. But nothing! Hum was barely audible with my ear on the speaker or from the chassis. And music played!
My first test flight was one channel only driven by a Sony Walkman on a small Totem speaker. Even under these conditions the sound was pretty good, and really started to open up after the first 60 minutes. I was like a proud father 
The first picture shows the pair with a Rotel CDP in the middle for scale.
I will not be able to fully know the sound until both are wired, broken in and hooked to the good speakers and source. I think they will be amazing. Already better than the very respectable chip amp they replace. There is a switch that toggles between light feedback and even less feedback which will be interesting.
What follows are some underside pics and the build process.
I read everything I could on this site with respect to hum and wire dressing. Taking a cue from postings by Sy, I deviated from the design spec which called for DC filament supplies and used AC on driver and power tubes. Input tube is DC. I was very particular about wire dressing. I have separate filament transformers for drivers vs input and 2 more torroids for power filaments. I used 22 guage teflon coated wire for the 6np6 drivers to get very tight twist. Even though they are wired in series at 12v AC, I ran separate lengths of twisted pair to each tube so they could easily be rewired to DC or to use 12AU7A tubes specified. On the power tubes there is one 12v torroid for each pair of tubes. I put them as close to the tubes as possible to keep the high current 12v runs short and fed 115v AC to them with as much separation from other wiring as possible. To get enough voltage from these transformers I added about 6 feet of wire to each secondary. This bumped the filament voltage from 11 volts to around 13. A bit high, but within spec.
I took a fair bit of time to wire all connections on a fairly geometric plane with good separation everywhere. In many cases, I removed insulation and just use air. Signal wires are combination of small gauge silver on input, 22 silver plated copper with Teflon on drivers to the power tubes, and 16 gauge copper on output.
There are 2 main ground buses to a star ground. Each power supply collects its ground and then separate ground back to the start ground.
I think the wiring looks good, and it is really easy to follow to double check the circuit before firing it up.
The modular power supply made it very easy to check out. Each section of the amp was individually verified prior to final assembly and final stage was really just an integration test.
I built a slow start for the B+ using a 12R 25W resistor in series. Basically it means I power on the main amp which fires up all the filaments, the small tube B+ and the bias voltage. After the tubes heat for 15 mins, you flip the power tubes on. After a few seconds a second switch shorts out the slow start applying full power to the tubes. The other reliability precaution was to heat the 6C33C tubes for 3 hours before applying B+. Then I kept volume down to minimum for 30 minutes. This burn in process is reported to result in 6C33C tubes that will be reliable. So far first 4 tubes are stable.
Next step is to clone the wiring into the right channel and hear the finished product. This should go faster than the first one as the design is done.
Will report the final sound in about a month.
It was a very happy day yesterday when the first mono bloc powered up and produced music. I was pretty comfortable that I had not made mistakes that would result in smoke, but really did expect some issues with hum, grounding or tube substitution. But nothing! Hum was barely audible with my ear on the speaker or from the chassis. And music played!
My first test flight was one channel only driven by a Sony Walkman on a small Totem speaker. Even under these conditions the sound was pretty good, and really started to open up after the first 60 minutes. I was like a proud father The first picture shows the pair with a Rotel CDP in the middle for scale.
I will not be able to fully know the sound until both are wired, broken in and hooked to the good speakers and source. I think they will be amazing. Already better than the very respectable chip amp they replace. There is a switch that toggles between light feedback and even less feedback which will be interesting.
What follows are some underside pics and the build process.
I read everything I could on this site with respect to hum and wire dressing. Taking a cue from postings by Sy, I deviated from the design spec which called for DC filament supplies and used AC on driver and power tubes. Input tube is DC. I was very particular about wire dressing. I have separate filament transformers for drivers vs input and 2 more torroids for power filaments. I used 22 guage teflon coated wire for the 6np6 drivers to get very tight twist. Even though they are wired in series at 12v AC, I ran separate lengths of twisted pair to each tube so they could easily be rewired to DC or to use 12AU7A tubes specified. On the power tubes there is one 12v torroid for each pair of tubes. I put them as close to the tubes as possible to keep the high current 12v runs short and fed 115v AC to them with as much separation from other wiring as possible. To get enough voltage from these transformers I added about 6 feet of wire to each secondary. This bumped the filament voltage from 11 volts to around 13. A bit high, but within spec.
I took a fair bit of time to wire all connections on a fairly geometric plane with good separation everywhere. In many cases, I removed insulation and just use air. Signal wires are combination of small gauge silver on input, 22 silver plated copper with Teflon on drivers to the power tubes, and 16 gauge copper on output.
There are 2 main ground buses to a star ground. Each power supply collects its ground and then separate ground back to the start ground.
I think the wiring looks good, and it is really easy to follow to double check the circuit before firing it up.
The modular power supply made it very easy to check out. Each section of the amp was individually verified prior to final assembly and final stage was really just an integration test.
I built a slow start for the B+ using a 12R 25W resistor in series. Basically it means I power on the main amp which fires up all the filaments, the small tube B+ and the bias voltage. After the tubes heat for 15 mins, you flip the power tubes on. After a few seconds a second switch shorts out the slow start applying full power to the tubes. The other reliability precaution was to heat the 6C33C tubes for 3 hours before applying B+. Then I kept volume down to minimum for 30 minutes. This burn in process is reported to result in 6C33C tubes that will be reliable. So far first 4 tubes are stable.
Next step is to clone the wiring into the right channel and hear the finished product. This should go faster than the first one as the design is done.
Will report the final sound in about a month.
Total Comments 1
Comments
-
Thanks Ryanj
Now paired with your Distinction TDA1541a board there is some amazing sound coming from these beasts!
WalterPosted 30th March 2015 at 12:13 AM by wlowes
