6C33C SE-OTL Project

[Wolfgang]

Maybe a test with the IXCY10M45 as we have already the Maida regulator which improves the SQ so much that I wouldn't want to gamble and lose anything.
Compared to the RCR which I used originally it sounds like a 100% upgrade. Choke loading the input stage I consider rather as difficult mostly from my own practical experience with this. I did tests with buffers and different chokes and it was quite revealing how it can affect the sound and frequency response despite good looking calculations. For the #26 buffer I used a tube anode choke from K&K which worked actually pretty satisfying.
P. Millet made an interesting comparison.Maybe someone has built a circuit like this already and could chime in?
http://www.pmillett.com/current_source.htm
Lower distortion and higher gain is what could be achieved. The lower distortion part would of course be interesting although the 6C33 with the maida reg. sounds already incredibly clean.

[brommermartin]

The test with the SSHV2 regulator faild due to problems with current setpoint
Next plan was to try CCS units as load for the 6sn7 driver tubes ( I have very good experience with the CCS work of Ale Moglia )
from Bartola valves
Today I have tested one channel and ik looks great hopefully it will sound great also .
This is what I have build ,design from Ale Moglia !

[brommermartin]

[Ray P]

Maybe a test with the IXCY10M45

As these CCS modules just slot in place of the load resistors it should be easy enough to experiment. I think I'll order a pair of the K&K kits so I can (eventually) try them on one channel vs a resistor loaded channel.

[Ray P]

...hopefully it will sound great also .

Watching with interest.

[Ray P]

I've done a little reading on using current sources and it seems the Maida reg will deliver many of the benefits of using current sources but they still offer advantages with respect to increasing the gain and linearity of the triode sections of the 6SN7s. WIll increased gain and linearity result in improved SQ or will some sort of interaction between the reg and current sources compromise things? I guess there's only one way to find out.

[Wolfgang]

WIll increased gain and linearity result in improved SQ or will some sort of interaction between the reg and current sources compromise things?

This depends also on the rest of the circuit. Increasing linearity is the most important benefit we are looking for when using CCS in this context. But how much better it will be also depends on our original operating point and B+. If we decide to adjust parameters regarding the load line we can get pretty good results already with typical resistor anode loads especially if we consider that the ideal looking "flat" load line which we get with CCS will be watered down again when connected to the next stage (300k/50k grid leak resistors).

The most important part to consider is the "sound" we are aiming for. Sometimes it can get clinical and emotionally uninvolving if we do too much. Many high end solid state amps (and some tube amps)sound like this but have incredible low distortion and flat frequency response almost into the MHz range. But we listen to music and not to low distortion and flat frequency curves. So all this needs to be considered and practically tested.
My feeling is that - for the input/ driver stage - either the Maida regulator with a very good operating point of the tube (enough B+, symmetric swing) or CCS would be the right thing to do. Choke loading is somehow in-between but needs very high quality anode inductors or it can backfire especially with big amounts of NFB. I cannot see any benefit of using a combination of both here but would like of course to listen how it sounds if we combine them.

For the output stage it looks a little different again. That's why I want to try a combo of both although I was skeptical from the beginning. CCS is necessary but the psu should be cleaned up and less capacitance could sound better. The SLB psu with the InvOTL was a real eye opener for me and is the main reason why I want to try this. Will be interesting how we can make things work (or not) choosing the right settings for the Maida regulator. I have some ideas that need to be tested practically.

[Ray P]

Talking with Kevin at K&K he tells me that small CCS boards are designed to be mounted onto a typical through hole resistor footprint using some solid copper wire - no other fixing required. It should be very simple to arrange some header pins to be able to switch between CCS and resistors. They're not expensive so I'll get some and experiment.

[Wolfgang]

Talking with Kevin at K&K he tells me that small CCS boards are designed to be mounted onto a typical through hole resistor footprint using some solid copper wire - no other fixing required. It should be very simple to arrange some header pins to be able to switch between CCS and resistors. They're not expensive so I'll get some and experiment.

So you will get this one:

CCS.jpg (16.62 KiB) Viewed 8117 times

The 2W max is very borderline for the first triode if you don't plan to reduce B+. Did you discuss this with Kevin and he said it's ok?

[Ray P]

The 2W max is very borderline for the first triode if you don't plan to reduce B+. Did you discuss this with Kevin and he said it's ok?

Yes, that's the one I've discussed with Kevin; just needs a slightly larger heatsink.

[Ray P]

I'm shutting this project down and putting my eggs into the 13E1 SE-OTL basket.

So what to do with my 6C33C valves, expensive ceramic sockets and various other parts? I thought I might put them aside and start collecting the other parts for a Tim Mellow OTL build;

http://www.vk6fh.com/vk6fh/triodeamp2013.htm

It'll be nice to compare a cap-coupled SE-OTL to a direct-coupled Push-Pull OTL.

[Ray P]

I ordered a set of PCBs fro the Tim Mellow 6C33C OTL amp this morning. I'll start a new thread when they arrive.

[pre65]

I ordered a set of PCBs fro the Tim Mellow 6C33C OTL amp this morning. I'll start a new thread when they arrive.

Ray, have you a link to the supplier please ?

[little eddy]

Another benefit of the CCS is that voltdrop is simply the difference between the supply and valve Va at the chosen current.

As valves age, or you try another, so long as you have enough voltage headroom, just drop in and test/listen at the same Ia.

[Ray P]

Ray, have you a link to the supplier please ?

Hi Phil, is this your next project?

The webpage for the amplifier project is here;

http://www.aresaudio.com/otl-power-amplifier.html

and the PCBs are available here;

http://www.aresaudio.com/kit-e-pcb.html

The site is only in Italian but Google Translate will take care of it and when I mailed them the response was in good English.

The PCBs they sell are those with the black solder mask. The boards featured are for a stereo build consisting of left and right driver boards, one power supply board and one start-up/protection board (microprocessor controlled to ensure everything is done in the right order after pressing a momentary contact switch - the programmed microprocessor is supplied with the start-up boards. ). I ordered an extra power supply and start-up board to give the option of two mono builds.

You shold be aware that this amplifier is hot - literally! The 6C33C envelope will give a severe instant burn and there are issues with pin temperature with cheap septal sockets so you need to spend a little on decent ceramic ones - I can point you at a source if you need some. I recommend you read the links listed near the bottom of the website I posted a link to previously;

http://www.vk6fh.com/vk6fh/triodeamp2013.htm