DIY Audio Projects Forum

The 6H8C has a maximum anode voltage of 330 volts which should be sufficient for an output of 240 if carefully biased. The 6SN7 has more detailed specs. If the 6H8C can handle the same voltages as the 6SN7 I think it will work as a driver for a GM70.

The GM70 as biased in my amp will need close to 240 volts (peak to peak) of drive for full output power. I'll breadboard a 6H8C driver and see if it fails at a B+ high enough to obtain an output of 240 volts. Not much to lose there. It's certainly worth a try.

Maybe the specs on these tubes are conservative and they can be pushed harder. That's just one more thing I don't know.

_________________
Stephen

Yep, that's the tube I'll probably use if it doesn't blow up when I breadboard a driver circuit.

Thanks.

_________________
Stephen

Thanks to Suncalc for this link:

http://www.tubebooks.org/tubedata/RC30.pdf

It's an awesome book.

Based on the information in the book I was able to determine a couple of important things about the GM70. I'm on fire today!

First, the output power in SE mode will be about 30 watts at roughly 4 or 5 % distortion. Two tubes in paralell would yield at best 50 watts, again with about 5% distortion. This also involves a very large, very expensive transformer.

For class A PP, the output is more than twice that of a SE stage. It's more like 3 times the power. The really nice thing is the transformer is a lot smaller and less expensive.

I also found out that the load resistance for PP can be less than for SE (as seen by each tube) because of lower distortion due to the cancellation of even harmonics in the PP output. The lower load impedance allows more power to be extracted from each tube from higher current. It appears the Edcor transformer I ordered will be ideal for the GM70.

I calculate the output power to be about 85 watts per channel with low distortion for a GM70 PP class A pair. With the output trannys at less than $100 and GM70s going for about $60 a pair shipped to your door this is starting to seem too cheap and easy. The 800VA toroidal power tranny that will give me a B+ of 1130 volts was only $90 on ebay. The total cost of iron for this project not including shipping is less than $300 for an 85 watt/channel amp. Not bad at all. Even better, the total cost of tubes will be about $130. For caps I'm using 3 series 450 volt rated caps on the PS, so there's not a huge expense there. The interstage PIO caps were only a few bucks each.

There's a TON of very practical knowledge in that book. Thanks again to Suncalc.

This amp is going to rock.

_________________
Stephen

I have given my amp a name. It will be called Giganotosaurus. Larger and more powerful than the mighty and better known T-Rex, this 8-ton beast roamed Argentina about 95 million years ago, only to be resurrected in 2010 by a DIY audio enthusiast in Las Vegas, not in the form of an 8 ton beast, but as an 80 watt per channel beast of an amplifier that will both heat and light a normal size room (in addition to making nice music).

_________________
Stephen

Here's the raw schematic of one channel of the Giganatosaurus. The PS is solid state (not shown).



I'll clean it up and add component values as time permits. A resistor on the grid of the top GM70 was inadvertently omitted.

A Darlington pair of high voltage transistors is used to extend the voltage capability of the CCS to 450 volts with a zener regulator providing a fixed voltage reference to the base of the Darlington pair. The cathodes should bias at around 120 volts with no input. The driver is simply two capacitively coupled common cathode triode stages with bypassed cathode resistors. The first stage is only partially bypassed to allow negative feedback to be applied to the lower resistor in the first stage from the output.

The GM70 heaters are directly connected to floating SMPSs, and the 6H8C heaters are biased at about 100 to 150 volts or possibly left floating depending on how the circuit behaves when constructed. If the 6H8C can't handle the voltage a 6SN7 can.

It's a very simple circuit.

_________________
Stephen

Here's the layout on a 17 x 14 x 4" Hammond chassis:



It's going to be a tight squeeze but it can be done.

The output transformers in the picture are identical in size to the ones I ordered. In fact, they would probably work as well as the ones I ordered and deliver slightly higher output power (Edcor CXPP100-MS-6.6K for KT88 vs. CXPP100-MS-10K for GM70). The toroidal power tranny and the SMPS heater supplies will go inside the chassis, leaving the output trannys and tubes on top. If I had the 6H8P tubes I'd have all my major parts.

There is some protective circuitry associated with the LM317HV (some diodes) to prevent failure and a tantalum capacitor to optimize performance that aren't shown in the schematic. There's also another slight mod that I can't talk about.

QUESTION: Notice the orientation of the plates and cathode filament wires. They are in the same plane, "facing front and back" if you will, as opposed to being in parallel planes "pointing at each other". Is this the correct way to orient the tubes to minimize interaction? Or do I have it backwards? I know the KT88 specifies a certain orientation but the description makes no sense to me. It would seem to me the direction of electron flow would be perpendicular to the plane of the cathode, and based on that assumption the orientation should be as shown, but I'm not sure.

As always any help or suggestions would be appreciated. I'm greatly indebted to Ben for saving me a lot of wasted time and blown up driver tubes. He pointed out I was exceeding the heater-cathode voltages in my SRPP driver, resulting in a redesign. It was something that I completely overlooked.

_________________
Stephen

I will also add that I am using both sides of my input and driver.

FTI

Could you sketch a schematic? I'm very interested in your driver. If it can drive an 845 it should be capable of driving a GM70.

Do you bias the heaters on the 6SN7 and 6H8C, and if so, at what voltage? How do you avoid exceeding the 100 volt heater to cathode voltage rating of the 6H8C in a cathode follower?

_________________
Stephen

Not 100% the same as mine but this is what we based ours on

http://www.lampizator.eu/AMPLIFIERS/CHINA/845/new%20circuit%20from%20austria/ma845_v42.pdf

Hope this helps.
My variation should be ready tommorow with more pics

http://entertainment.webshots.com/album/572269202bzvFww

FTI

That looks like it would work. The choice of the 6SN7 is no coincidence I'll bet. There just aren't many small tubes that can handle high voltages needed to drive 845s and GM70s. The 6SN7 has great tolerence for heater - cathode voltage and the plate can run at 450 volts.

I'm going back to a SRPP driver with a 6SN7. I think I've found a way to make it work. If the goal is a low distortion PP amp (and that's my goal) the CC triode gain stage isn't a good choice IMHO. The SRPP is highly linear with very low distortion, so I'm sticking with it.

The schematic for the Giganotosaurus posted above is out of date. The triodes will be configured as a SRPP driver. I think I've licked the heater bias problem.

Thanks.

_________________
Stephen