• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

OTL designed by Tim Mellow with 4 6C33C?

dwhitf

Member
2008-05-20 12:11 am
I have decided to build the OTL designed by Tim Mellow, which appeared in the February issue of Audio Express. I am considering using 4 6C33C's instead of 2 outputs for more power.
Do you think the driver section will be able to push 4 output tubes? I Would like to hear opinions.
I am planning 2 mono blocks.
Thanks in advance!
dwhitf
 
I'm considering the same, using all Russian tubes (6N2P-EV in place of the ECC83; 6J32P for the EF86), and am also wondering about adding a second pair of 6C33C-B's. I was thinking about designing my own, but I don't have the EE chops - and Mellow's schematics look fairly straightforward.
I don't know that I'd use that small a chassis, though - those 6C33s are REALLY close to each other... I'll probably go with monoblocks, also.

PDF of article from audioXpress:
http://diyaudioprojects.com/Forum/download/file.php?id=2101
 
I am collecting parts for that build as well. I am very happy to see this thread start. Maybe we can be support to eachother.

Are you saying you want to use 4 tubes per side?

Yes, the chassis Tim used is way too small. I have an artsy looking chassis that I am using. It is two levels and I will put the tubes on top and the power on the bottom. Those tubes are known for getting pretty hot.

I was going to follow Tim's scheem to the T, but I would like to do tube rectification. Has anyone thought about that or worked it out?
 
I was thinking of a hybrid bridge, using a pair of 6D22S - I like the delay and power ramp-up they provide. I was trying to design a PS using PSUII, but that doesn't allow hybrid bridges (or, more than likely, I don't know how to set the program up for them...). I've got the tubes, sockets, and caps, so I'd like to use them.

djn - yes, 4 6c33s per side. I'm hoping for a little more power, and less stress on the tubes. 6C33s may be fairly cheap, but I don't have a lot of cash to spend on this - I would like the tubes to last a while...
 
Last edited:
I've got an extra pair of the 6d22s as well. I am using that tube (2 of them) in my GM70 amp and yes, the warm up time is great for protecting the tubes. I get into the PSU thing tomorrow to see if the hybrid can be done, but I've never seen it.

My speakers are high eff so I don't need the extra power, but the longer life is something to think about assuming they are not run as hard as just two of them.

I have not really looked at the PS in Tim's article that hard but enough to know it is not that high powered so there should be room to play. My GM70 is running 850v B+ and that is just a big pain in the *** to ensure there is addiquite room to avoid shorting and corrona effect.
 
...My speakers are high eff so I don't need the extra power, but the longer life is something to think about assuming they are not run as hard as just two of them.

Mine aren't. 90dB AR M5s. Can't afford to replace them, and I really like the way they work with the room's acoustics. They can be bi-wired/bi-amped, though - I could just build 2 full amps and use 2 channels per speaker.
Hmmm...

I have not really looked at the PS in Tim's article that hard but enough to know it is not that high powered so there should be room to play. My GM70 is running 850v B+ and that is just a big pain in the *** to ensure there is addiquite room to avoid shorting and corrona effect.

I have. HT3 is at 430v - not huge. Schematic may need proofing, though - Mellow states in the text that R29 limits the current through the output stage and speakers in the event of a fault that forces the output stage to latch either up or down; looking at the schematic, I don't see how R29 could do that. Also, the schem shows -50v to the grid of one 6C33 (V4), and -200v to the grid of the other 6C33 (V5) - is this correct?

What am I not seeing or understanding here?
 
Yup. All my DIY exp is in SS; this will be my first glass project. From what I've read here and elsewhere, I get the impression that OTL is generally not something for a noobe to tackle - but this amp circuit schem is less complex than the last ss headphone amp I built. I've dealt with HT in the past, and respect it.

Hopefully some of the more knowledgeable members will chime in...

Re bi-amping: Yeah, the M5 have seperate connectors for the woofer and mid/tweeter sections - easy.
 
I have. HT3 is at 430v - not huge. Schematic may need proofing, though - Mellow states in the text that R29 limits the current through the output stage and speakers in the event of a fault that forces the output stage to latch either up or down; looking at the schematic, I don't see how R29 could do that. Also, the schem shows -50v to the grid of one 6C33 (V4), and -200v to the grid of the other 6C33 (V5) - is this correct?

What am I not seeing or understanding here?

R29 is a typo for R33. The -50V and -200V on the grids of V4 and V5 are perfectly correct. The cathode of V4 is at 0V, whilst the cathode of V5 is at -150V, so each of them therefore has a grid-cathode voltage of -50V.

I've built this amp, and I'm very pleased with its performance. I used a somewhat bigger chassis than Tim Mellow did, and I included a quiet cooling fan and made a few holes in the top of the chassis. It runs a bit warm, but nothing too excessive I think.

Chris
 
I more or less just followed Tim Mellow's design. I didn't bother with the ammeters; I just put 0.1Ohm resistors instead, with tip jacks so I could check the current flow with an external DVM occasionally. I'm just using 200V electrolytics in the HT2 and HT4 output-tube supplies (two 2,200mfd
in parallel for each supply), rather then the string of 4 lower-voltage capacitors in Tim Mellow's PS.

I'm not using any start-up HT delay, and I checked before first running it that the EF86 pentodes V2 and V3 heat up and start functioning long before the 6C33C output tubes begin to conduct. This is important because with those 100K resistors from grid to anode on the 6C33C's, if the EF86's aren't running properly first, something quite unpleasant could happen! But anyway, everything works just fine. When first testing, I recommend using 8Ohms power resistors in place of speakers, for a first check that voltages are OK.

My speakers are Lowther DX3 with about 100dB sensitivity, and the amp can drive them much louder than I'd ever want to play them. Probably 90bB speakers will be fine too, I would guess. By the way, the hum level is so low its inaudible, even with 100dB speakers.

djn: I would have thought tube rectification would be quite a challenge. You really want the output tubes' HT supplies to be able to give two or three amps without significant drop in the voltage. Also, with HT3 needing to be negative (-430V), it all sounds a bit of a bother. Rectification is one place where I prefer to make life simple!

Chris
 
R29 is a typo for R33. The -50V and -200V on the grids of V4 and V5 are perfectly correct. The cathode of V4 is at 0V, whilst the cathode of V5 is at -150V, so each of them therefore has a grid-cathode voltage of -50V.

I've built this amp, and I'm very pleased with its performance. I used a somewhat bigger chassis than Tim Mellow did, and I included a quiet cooling fan and made a few holes in the top of the chassis. It runs a bit warm, but nothing too excessive I think.

Chris

Chris is correct on the neg voltages. One issue I have is that the 6c33 data sheets i have all show different socket pinout numbers then the ones in the article. Can someone confirm i have attached a pdf datasheet for the tube.
 

Attachments

  • 6S33S.pdf
    104.2 KB · Views: 805
For N1, I used a 65V 0.7mA neon lamp, type A9A by Chicago Miniature. Available from Mouser for 32 cents.
Any small neon lamp is probably fine.

For N2, by the way, I'm using an A81-C90X gas discharge tube made by EPCOS. Again, available from Mouser, for about $2.00.

It looks like Tim Mellow's pin numbering scheme for the 6C33C assumes you take the anode to be pin 1, and then count round from there. The seven pins on the 6C33C are at equal angular spacing, so there is no obvious choice for which should be called pin 1. The only distinguishing feature is that the anode pin is thicker than all the others, so calling it pin 1 could be considered natural, I suppose. But anyway, since his schematic doesn't indicate the pin connections for the two heaters, you need the full data sheet to get that information, so its probably best to stick with their numbering convention.

Chris
 
By the way, I'm pretty sure that mislabelled R29 resistor is the only typo in the article (if you don't count the valid, but non-conventional, numbering of the 6C33C pins).

One other remark; I recommend using an NTC thermistor in series on the mains AC supply side of the transformers, to reduce the start-up surge. I am using a couple of hefty toroidal transformers for the output-tube HT supplies and the 6C33C heaters, and I was blowing pretty hefty fuses until I added the thermistor. I'm using a CL-40, again available from Mouser.

Chris
 

dwhitf

Member
2008-05-20 12:11 am
Hey Chris!

Don't mean to beat you like a horse(with questions) but you seem "to be the man" on this project. Having broke ground on this Project and having a working amplifier.

I'm in the parts collecting phase and would love to know the "hefty toroidal transformers" you picked for this project. HOW ABOUT A PICTURE?
Thanks in advance!
dwhitf/Darrell