• 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.

6L6GC AB2 Amp

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
tubelab.com said:
I have a few 45's that I have collected over the years. Most are well used and not "grade A" quality.

Yes, mine are all well used as well. I have one pair that may be new and another that very strong...all Raytheon. The rest are various odds and ends from eBay in varying condition. The TSE seems to work fine with the weaker tubes, but they work in class A of course.

Could make a serious Dynaco-on-steroids with a circuit like this.
 
My $.02 on power. If you are getting 75w out of a pair of KT88's, triode strapped in AB2, then there really isn't much point in going to pentode to extract more power unless you are driving B&W 801s or Apogees. 75w is plenty-o-power for more domestic situations even with inefficient 6.5" two ways.
 
My $.02 on power. If you are getting 75w out of a pair of KT88's, triode strapped in AB2, then there really isn't much point in going to pentode to extract more power unless you are driving B&W 801s or Apogees.

That is my opinion also. I wanted to try pentode mode. It got me about 2 db more power which didn't sound that much louder (if at all) on my 87db Yamahas. Once that little 7 inch cone has reached its Xmax more power only makes more distortion and possibly dead speakers. The speakers in the living room have 15 inch woofers and are 96 db they start shaking the walls with a simple SE using this same pair of KT88's.

I have these Plitron OPT's that are rated for 400 watts at 20 Hz. I got them almost 2 years ago when someone mentioned them on this forum. I have visions of a multi kilowatt monster amp using a big bunch of sweep tubes, but a project like that will likely never get built or used. I have a perfectly good working 845SE amp (40 WPC) that I dont use because it is big, heavy, and puts out lots of heat. It seems like a waste to use 400 watt transformers to build a 75 watt amp, but it is a bigger waste to have them sitting in the box.

It has been about 2 years since I have built an amp, any amp. I am almost done with the Tubelab SE that I started 2 years ago. Once that is done I really want to build something, so this may be my chance. Once the Simple P-P boards get here I will be busy with them.

The original idea:

http://www.diyaudio.com/forums/showthread.php?s=&threadid=100852&highlight=
 
When do you think the Simple P-P boards will be in and available? I know that you mentioned that you still have to write a manual and such, but was just curious. Still jonesin to build one!

Also I was wondering if the Simple P-P boards could be used to build monoblocks. Just a thought.

-Steve
 
When do you think the Simple P-P boards will be in and available?

The boards are scheduled to ship in about 2 weeks. Advanced has been known to ship early. I have a skeleton of the manual written but the time frame for completion is really up in the air.

The boards should arrive at about the same time we get our next pass IC chips back at work. The last chip led to 60+ hour work weeks for almost two months. If the boards arrive one weekend before the chips it could make a months difference in the time to get everything done.

I usually want to build at least one amp, and get a beginner to build one with the instruction manual before I start selling them. It really cuts down on the questions.

I already have two Simple P-P prototype boards up and running. I have not tried my usual mono block experiments with them yet. There are usually 3 ways to make a mono block out of a stereo amp. The eaisest way is simply to leave out one channel. I have done this with the Tubelab SE and the Simple SE, and I am sure that it would work on the Simple P-P. Another option is to parallel both channels for twice the power output. Some circuit changes could be used like wiring all 4 tubes up to one OPT. Another option is to use the amp as a two channel amp for bi-amping a single speaker.
 
diyAudio Chief Moderator
Joined 2002
Paid Member
tubelab.com said:
I think this is how I blasted the IRF820's. One of them blew when I switched ranged on the audio generator... I would try the 820's again except I inly have one left. I will get some more and some 10 or 12 volt zeners next time I order parts.

Maybe the addition of a diode also is even safer? Like with the classic Zener & diode protection combo for lateral Mosfet amplifiers?
 

Attachments

  • normalz-d.gif
    normalz-d.gif
    10.5 KB · Views: 1,950
I think in the circuit that we are contemplating the addition of the 1N4148 diode may be a bad idea. The gate can go up to 10.7 volts positive with respect to the source before being clamped by the zener (assuming a 10 volt zener). The addition of the diode will allow the gate to be pulled negative far in excess of the gate source breakdown rating. This may not have been possible in the solid state circuit shown, but it is certainly possible here.

A single zener will clamp the gate to 10.7 volts in the positive direction and 0.7 volts in the negative direction. Since the fet should have been cut off once the gate has gone below 3 to 6 volts (depending on the fet), clamping the gate at -0.7 volts should be OK.

I have noticed that many mosfets have internal zeners. This is what I have been using. They use two zeners in series in opposing directions. This clamps the gate by the same amount in each direction. Since this is what the mosfet manufacturers do, this is probably the most correct solution. I would choose a zener voltage 3 or 4 volts below the gate source breakdown rating.
 
When do you think the Simple P-P boards will be in and available?.......The boards should arrive at about the same time we get our next pass IC chips back at work. The last chip led to 60+ hour work weeks for almost two months. If the boards arrive one weekend before the chips it could make a months difference in the time to get everything done.

Well. IBM won the race in the shipping early department. I was handed 900 IC's today and told to test them all. I think most of my Tubelab time is gone for the next few weeks. I have squeezed in a few experiments during the last few days. There was no major new discoveries, but here is the summary:

Attempts to drive several sweep tubes in conventional grid drive did not work near as well as the KT88's. Since I am out of power supplies I had to run the output tubes in triode mode. This puts the screen grid well above their ratings. Screen drive will require a board modification.

I have had no success with cathode feedback in push pull in previous experiments. CFB requires a well balanced secondary since uneven feedback that varies with frequency will increase the distortion, destroy frequency response or both. Granted most of my P-P OPT's are low priced bargains. Could these Plitrons be different? Yes, they are. Limited testing revealed a reduction in distortion at all frequencies and power levels in pentode mode. There is a considerable loss of power output too. Maximum power output was redused from 130 watts to 105 watts. I think the driver was running out of headroom. There was a similar reduction in power and distortion in triode mode too. No listening tests were performed.
 
I have a bunch of old WWII vintage tubes in my warehouse. I have looked at a box full of funny looking beam pentodes several times without thinking twice. Today I looked up the specs and thought to myself that this amp would be the perfect place to try them.

What would you think about a tube that had a maximum plate voltage rating of 4000 volts (not a typo) a maximum screen voltage of 850 volts, a plate dissipation rating of 40 watts and a peak plate current rating of 10 AMPS!

There are even published triode curves to 1000 volts and 1 amp showing negative and positive grid voltages. Yeah there has to be a catch, out of the 5 that I have there are only two good ones and they have different type numbers and filament voltages. Worse, nobody lists them for sale.

I will reveal the type number after I find out if I can buy 4 similar ones for reasonable money.
 

Attachments

  • mysterytubecurve_a.jpg
    mysterytubecurve_a.jpg
    94.2 KB · Views: 1,158
Oh, I forgot to say that these things just blow away the KT88's. Power, I stopped cranking when my old Fluke power supply went into convulsions. I was over 100 watts with 450 volts in triode. Distortion at 75 watts was about 1.8%. I couldn't make the two good ones glow with my power supply even with the mega ASC cap across it. The bad ones either glow blue, run away or both.

Measured specs are one thing, but what about the sound. I got these things hooked up at about 7PM and I have been listening since then (3 hours). I have fed it everything from Enya, Diana Krall and Norah Jones to Metallica, Pink Floyd and Billy Idol. Right now I am playing Dire Straits. The dynamics and clarity are awesome.

If anyone recognizes these tubes please don't reveal the number until I can get 4 (if they are afordable). I would like to experiment further with these guys.
 

Attachments

  • mysterytube_a.jpg
    mysterytube_a.jpg
    93.7 KB · Views: 1,213
diyAudio Chief Moderator
Joined 2002
Paid Member
tubelab.com said:
The addition of the diode will allow the gate to be pulled negative far in excess of the gate source breakdown rating. This may not have been possible in the solid state circuit shown, but it is certainly possible here.

Maybe it just allows an additional 0.7V towards negative and that is it in that solid state implementation? Such laterals have under 1V Vgs. It seems weird to me why I see it all the time like this for +/- 14V Vgss lateral parts, if it allows full negative pull.

P.S. Fantastic find the tubes you drive in the pic.
 
I have another box of tubes. These are E130L or 7534. I have a few and so far they are all good. There are several flavors including Amperex, Mullard and Tungsram.

Maximum plate voltage is 2000 volts, good. Peak cathode current is 1.5 amps, good. Max dissipation is 27.5 watts, OK. Maximum screen voltage is 250 volts, bummer. Maximum plate voltage in triode is 250, how can I have any fun here? Typical operating characteristics show 60 watts output from a pair in pentode mode with a 300 volt supply, maybe there's hope. So what to do? Plug them in and crank them up.

Triode wired with a 1250 ohm load, 50 watts is pretty easy to get with a 275 volt supply. I cranked a pair of the Hungarian Tungsrams up tp 320 volts and saw the power meter read 80 watts at 5% distortion. OK, cool. I dialed back the supply to 300 volts and plugged in a pair of Mullards. Power is 70 watts at 5%.

I have been listening to them for a while now. They are not quite as punchy and dynamic as the Mystery Tube, but maybe better than the KT88's. Who knew that you could rock with only 300 volts!

I have now found 3 different setups that all sound good. I guess it's time to stop playing and build the other board so I can hear this thing in stereo. Then I need to think about a power supply. I have all sorts of transformers that can get me 450 volts, but 300 volts isn't easy.
 

Attachments

  • e130l_a.jpg
    e130l_a.jpg
    76.1 KB · Views: 1,086
It's easy to make an amp sound good with a large by huge by expensive OPT. Low impedance tubes are usually good candidates for using a power (mains) toroid for an OPT. So into my dumpster collection we go. Yes I got 4 or 5 of these from the dumpster at the warehouse complex. They came out of some kind of printer. They were powering some stepper motors. Do they rock? Do my neighbors like to hear Money for Nothing?

As with any mains toroid careful balancing, or deliberate misbalancing of the cathode currents is required. I applied a 40 Hz tone cranked it up to 30 watts and adjusted the cathode currents for lowest distortion. In this case an imbalance is required. One tube is at 42 mA and the other is 47 mA. Frequency response is 11 hZ to 77 Khz at 10 watts. 50 watts can be obtained from 31 Hz to 57 Khz with a distortion below 5%. Saturation causes the distortion to rise below 31Hz and an attempt to crank it up at 5 Hz really upset the power supply.

It sounds good too. Now I can save the BIG OPT's for a big amp, and there is no reason not to make one of these. Now I just need to find a dumpster power transformer and I will have a very low cost amp.
 

Attachments

  • dumpstertoroid_a.jpg
    dumpstertoroid_a.jpg
    94.7 KB · Views: 1,062
P.S. Fantastic find the tubes you drive in the pic.

I called my favorite tube supplier (ESRC) and he had 5 of the Mystery Tubes. I got them all for a reasonable price, so I didn't look any further. He has none left.

I will likely use them in my amp with the big OPT's. Their desire for about 450 volts is easilly satisfied by one of several toroidal power transformers that I have. This leaves me with some E130L's and some dumpster OPT's. I guess this means two amps. I have been looking for an excuse to make some P-P amps. With the Simple P-P coming soon, this will be my year for P-P.

The Mystery Tube has two numbers. Some tube suppliers often use both numbers interchangeably. They are NOT the same. The heater pinout is different. With a jumper from pin 2 to pin 7 you can use either tube interchangeably with a 6.3 volt heater supply.

The tube on the left in post #332 is a 3D21WB. These have a brown phenolic base. The heater is 12.6 volts with a 6.3 volt tap so that they could be used on either voltage. I have Bendix, Tung Sol and Cetron versions. Some were made with getters, some were not. These were all used, and all bad (gassy). The only good one is a Chatham, USAF. As you can see in the picture it is well used.

The tube on the right is a 7403. These have a metal base and the heater is 6.3 volts only. The only one that I have is a Tung Sol. It looks to be new or lightly used. The ones at ESRC have no brand name at all, we will see what shows up.

These tubes are listed on the internet for from $30 to $50 each, but I do not know if they are actually in stock anywhere. These are NOT the same tube as a 3D21, 3D21A, 3D21B or 3D21WA. Those all have 15 watt plates. Only the 3D21WB and the 7408 have the 40 watt plates.

I have some 3D21A's. They are also well used, and I have not tried any of them yet. They may work at a lower power level.
 
I was curious if 6146 tubes could take the challenge. Do you have any ?

I have some 6146 tubes. There is only a few of them, and I don't know their condition. I have not worked with the 6146 in recent years, but these tubes should be approached with caution. The original Ampeg SVT bass guitar amp used 6 X 6146 tubes to produce over 300 watts. These amps quickly gained a reputation for fiery explosive deconstruction. There was a very public example during a televised performance by the Rolling Stones (around 1970) where the amp burst into flames on stage!

I witnessed an SVT deconstruction at an outdoor concert in Miami, also around 1970. I heard the speakers failing for several minutes before the meltdown, so I believed that the speakers went open circuit at full power taking out the amp.

With this, the reputation of the 6146 in the audio world was tarnished forever. Ampeg redesigned the SVT to use 6550's and it was a very successful amp.

I had a big bunch of surplus 6146's about 30 years ago, and I made some screamin' loud guitar amps. Guess what, some of them fried! I understand the electronics behind tubes much better today than I did 30 years ago. I believe the root cause for all of these failures is the fragile screen grid in these tubes. The usual guitar amp method of powering the screen grid (a resistor and a cap from B+) doesn't work here. The 250 volt screen limit is real.

The limit for the screen grid is 250 volts in pentode mode. I wouldn't abuse this limit. Some data sheets show that the tube can handle 400 volts in triode mode. No limit for UL operation is specified. I have not tested or worked with the 6146 in a long time, so I don't know how real this 400 volt rating is. My suspicion is that it varies from tube to tube.

Most of my testing with the octal driver board has been in triode mode. The monster Plitron OPT's that I have do not have UL taps, and I have never been a big fan of the pentode with GNFB sound. Maybe I just haven't made a good pentode amp yet though. I am in the process of building a two channel test amp to explore the output tube / OPT choices a bit further. I have a big list of things to try, so I can't say if the 6146 will ever get tested.

The Simple P-P boards have arrived, so these experiments have already been put on the shelf for the near future.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.