OTL 6c33c Ciufolli v8.2 low voltage rails

[danielrs]

Hi,

I have a pair of toroidal transformers 625VA 65vac-0-65vac. I wish to build low voltage rails (about +/-90vdc B+/B-) on Ciufolli v8.2 design monoblock version but double the output tubes 4x6c33c.

Wondering if anyone could help to estimate the power output at 8 ohm & also 16 ohm load? I've just thinked the possibilities on using fixed bias 90v Vp, 400mA Ip & about -20v grid. What is the best suggested bias for this plan actually?

I 've ever heard that with 400mA the 6c33c could have rp as low as lower than 100ohms. The higher plate current, the lower the rp isn't it? Could someone validate this idea is worth to build?

Thanks and have a great day.

Ciao, danielrs

[tubetvr]

90V is too low to give high output power even when doubling the tubes. The original amp allow the tubes to swing up to 2.5A peak which is necessary to get about 25W in 8 ohm. With 90V a 6C33C can not carry more than about 1.4A and if you allow for some output voltage swing the voltage over the tube goes down to maybe 60V you can not draw more than about 900mA per tube or 1.8A for 2 tubes in parallell. 1.8A peak will give you about 13W in 8 ohm and you will then have much lower efficiency than in the original amplifier. In addition to the lower power it is not that easy to parallell connect output tubes, you need to ensure that current is divided evenly between the tubes, I would not recommend this if you can avoid it.

The best bias scheme is to use about +/-150V - 160V and bias the tubes to around 200mA, this will give at least 25W in 8 ohm and about 20W in 4 ohm.

A 6C33 have a Rp of about 80hm at 500mA cathode current, note that in the Ciufoli amp which is an inverted Futterman the output tubes work as cathode followers in parallell which give an output impedance of about 11ohm before feedback is applied.

You can look here for the schematic of my OTL which is a bit different from the Ciufoli http://www.tubetvr.com/otl.html

Regards Hans

[danielrs]

Hello Hans,

Thanks for your quick respond. I wish to utilise my current pair transformers. As per your suggestion, then if I would avoid that plan, for the power supply solution, what do you think if I can use:

1). Voltage doubler on each channel, so I can get +180/-180vdc. I read on Ciufolli website that the full-wave voltage doubler which using two diodes (actually this run half wave), he have got terrible results. An unknown Japaneese company and Matsuoka using this approach. [http://www.audiodesignguide.com/otl/matsuoka.jpg] Wondering if you have known somebody using it and what the sonic would look alike?

2). Series the pair of the transformers, so I can get 130vac-0-130vac, thus about +180vdc/0/-180vdc. The other thing that I would like to know is the dual power supply for getting the real capacitor less vs. using the floating virtual ground (n.b. there are two output capacitors & 2 series capacitors in the power supply). [http://www.audiodesignguide.com/my/ocl.jpg] Ciufolli also says that the bridge solution is better ! (nevertheless the output current problem ...). What is actually the potential current problem that he says.

Hans... I have read your website... it is interesting! I noted that you are using the floating virtual ground right? In my aformentioned alternative PS solution no#2, with having series the two transformers, I can have the real capacitor less. Please let me know what is the difference between two. If I will use the real capacitor less, do I need the speaker protection? Please advise.

Thanks, danielrs

[tubetvr]

I read on Ciufolli website that the full-wave voltage doubler which using two diodes (actually this run half wave), he have got terrible results.

I don't think he actually built it but he draw conclusions based on simulations. BTW what is shown in http://www.audiodesignguide.com/otl/matsuoka.jpg is a full wave voltage doubler. I don't see that it would be so bad using a voltage doubler IF you design it correctly, i.e use big enough caps 4 times as big as in a normal power supply and as he says some commercial OTL amps have used this method, I would however use 2 transformers to get a higher voltage and use the same power supply for both channels, this is better as you can use smaller caps for the same ripple.

The other thing that I would like to know is the dual power supply for getting the real capacitor less vs. using the floating virtual ground (n.b. there are two output capacitors & 2 series capacitors in the power supply).

What I think Andrea is trying to explain is that a floating power supply is as good as a power supply with "real" ground, this is true as long as the wave forms are symmetrical which is the case for music. If the wave form is symmetric, there will be a virtual stable ground that is as fixed as a "real" ground.

If I will use the real capacitor less, do I need the speaker protection? Please advise.

The virtual ground version is more or less self stabilising for DC but if one tube gets shortcircuited there can still be a lot of DC at the output especially if not each the capacitors can't withstand the full DC voltage, (so the voltage can not "float" up all the way).

So if you believe there is a risk of tube failure I advice to use some kind of protection, in my own amp I use fast fuses in the power supply but in the amps I have built for others I use a comparator that check the DC on the output and switch off the DC from the power supply if it is too much.

Regards Hans