My 6L6GC amp build thread

[GordonW]

Could you substitute 12SN7s and 12SL7s for the preamp tubes? Put all four 6L6s on the 6.3v heater (that's less than 4 amps), and the four input tubes on the 12.6v heater?

The nice thing about that, is that 12SN7s are pretty cheap, compared to 6SN7s... for example: http://thetubestore.com/nos-12sn7.html . Same goes for 12SL7s, AFAIK...

Regards,
Gordon.

 

[Maxxarcade]

Could you substitute 12SN7s and 12SL7s for the preamp tubes? Put all four 6L6s on the 6.3v heater (that's less than 4 amps), and the four input tubes on the 12.6v heater?

The nice thing about that, is that 12SN7s are pretty cheap, compared to 6SN7s... for example: http://thetubestore.com/nos-12sn7.html . Same goes for 12SL7s, AFAIK...

Regards,
Gordon.

I called Antek, and they are sending me a filament transformer for free. It's one more thing I'll need to put under the chassis, but whatever works I guess.

I'm not too happy about having to pull out a bunch of wiring that is now at the bottom layer of components. It's going to end up looking really sloppy, after all the work I did to keep it organized.

 

[Maxxarcade]

OMG it lives

All I had to do was reverse the OPT primaries and it sounds great. I had to use my bench power supply to run the heaters for the 6SN7's until my other transformer arrives.

Though there are still a couple issues. The 10uf, 350v filter caps on the two lower B+ supplies are going to need to be changed to something for higher voltage, because they get pretty warm.

I also need to add some bleeder resistors, because those two B+ lines are really high before the tubes warm up. Way too much for the 350v caps that are on them. In fact, I think I saw it go over 500 volts before I backed off the variac

So after I fix these little problems, I can move on to building the pre-amp circuits.

So how should I benchmark this thing? Square wave response and wattage? I'm planning on hooking up the scope to it tomorrow to see how it performs.

 

[Maxxarcade]

Did some basic testing tonight. Here are some specs I was able to get:

Main B+ = 450v
B+ #2 = 320v
B+ #3 = 260v

With my audio generator at full output of 1.5 volts RMS, 1Khz, I get 10.5 volts out into 8 ohms. No clipping at all. Need more input voltage...

20Hz produces a perfect sine wave with no voltage reduction. Generator won't go lower.

HF roll-off does not occur until somewhere around 25Khz. 28Khz only drops output about 1.5 volts.

I'll try a PC based audio generator to do more precise testing later.

Now off to build the much needed preamp :thmbsp:

 

[thefragger]

Yikes! Sounds great!!

 

[Maxxarcade]

Main B+ = 450v (Plates)
B+ #2 = 320v (G2)
B+ #3 = 260v (6SN7/6SL7 B+)

Just occured to me... Shouldn't my G2 voltage be higher, like around 350v?

With only 320v, wouldn't the output power decrease? I thought it seemed a bit weak when I tested it :scratch2:

The original HK250 had 395v plates and 310v G2. That's an 85V difference vs the 130v with my amp.

Maybe I need to change the series resistor? I used a 6.8k.

 

[GordonW]

I'd think that the screens and the 6SN7s/6SL7s could use a bit more voltage. But, you probably don't want to go over 350V on the screens. 450V plate/ 350V screen is one of the textbook alignments for the 6L6GC... here's that exact alignment, taken from nj7p:

Push Pull Class AB1 Amplifier
Plate Voltage ................................. 450 V
Grid No. 2 Voltage ............................ 350 V
Grid No. 1 Voltage ............................ -30 V
Peak Grid No. 1 Voltage........................ 60 V
Plate Current (Zero Signal) ................... 95 mA
Plate Current (Maximum Signal) ................ 194 mA
Grid No. 2 Current (Zero Signal) .............. 3.4 mA
Grid No. 2 Current (Maximum Signal) ........... 19.2 mA
Load Resistance ............................... 6.0K Ω
Power Output (approx) ......................... 50 W

Maybe go to a 5.6K or 4.7K? Since you're running cathode bias, you would want to keep an eye on the idle plate current (no more than 55ma or so per tube, or 110ma for a pair is a good place to start- that gives around 24.5w idle dissipation per tube)... you may need to tweak your screen voltage to get it in that range...

Regards,
Gordon.

 

[Maxxarcade]

I'd think that the screens and the 6SN7s/6SL7s could use a bit more voltage. But, you probably don't want to go over 350V on the screens. 450V plate/ 350V screen is one of the textbook alignments for the 6L6GC... here's that exact alignment, taken from nj7p:



Maybe go to a 5.6K or 4.7K? Since you're running cathode bias, you would want to keep an eye on the idle plate current (no more than 55ma or so per tube, or 110ma for a pair is a good place to start- that gives around 24.5w idle dissipation per tube)... you may need to tweak your screen voltage to get it in that range...

Regards,
Gordon.

What would be the easiest way to check idle current/bias on this amp? My output circuits are basically identical to the HK250, aside from having 6SN7 instead of 6CG7.

The other day I turned the variac up faster than normal, and I thought I saw one of the 6L6GC's arc. I'll check on that this weekend when I get the 10uf caps changed over to 450v.

 

[GordonW]

What would be the easiest way to check idle current/bias on this amp? My output circuits are basically identical to the HK250, aside from having 6SN7 instead of 6CG7.

The cathode circuit on these, with the balance pot centered, is around 280 ohms, for the pair of output tubes. That means, at 100ma current, you'd have 28v at the cathodes.

Pretty much, you'd not want to see more than 30.8v on those cathodes. That'd be right at 110ma for a pair. If it's higher, you either will have to reduce B+, reduce the screen voltage, or increase the value of the cathode resistors... or, a combination of more than one of those measures. If it's less... you can always just bump up the screen voltage...

Regards,
Gordon.

 

[Maxxarcade]

The cathode circuit on these, with the balance pot centered, is around 280 ohms, for the pair of output tubes. That means, at 100ma current, you'd have 28v at the cathodes.

Pretty much, you'd not want to see more than 30.8v on those cathodes. That'd be right at 110ma for a pair. If it's higher, you either will have to reduce B+, reduce the screen voltage, or increase the value of the cathode resistors... or, a combination of more than one of those measures. If it's less... you can always just bump up the screen voltage...

Regards,
Gordon.

Thanks, I'll have a go at it this weekend and see what happens.

 

[Maxxarcade]

Well, the preamp circuit seems to have been a waste of time. Once the tubes warm up All I get is a loud 60Hz buzz, like you get when you touch an input lead with the volume up. It doesn't seem to pass any signal either. I didn't leave it on more than a second because the hum was just too loud to stand, and I didn't want to damage anything.

I checked everything more than once to make sure it was hooked up right, but obviously something's wrong.

Did I get the pinout correct for the 6SL7? The circuit originally called for a 6EU7. Also, R6 is partially bypassed by using two 470 ohm resistors in place of the 1K, with a 47uf cap across one of them to increase gain a bit. This was something GordonW suggested if I recall.

So, should I inspect everything again tomorrow, or is this circuit just crap? I'd like to troubleshoot it if I can. It's just a simple passive tone circuit with an amp on each end. I just don't know where to start I guess.

Aside from that, the power amp circuits are working pretty well. If I can get the preamp going I'll be all set.

 

[battradio]

Your pin to pin convesion from a 6EU7 to a 6SL7 looks good . A photo of the construction and where you are geting the B+ from would help .I don't see a volume control this would be full gain all the time , so one needs to be added .First does the tube fillament light . Second if you have a volt meter does the tube have B+ at pin 2 & 5 . Third if yes to 1&2 dissconnect C3 from R7 & C8 and connect the loose end to the output jact ,after removing C7 from the jack.See if this will pass audio .If so your problem is after this point , if not its before this point .

Mark

 

[PabloL]

Gounding?

Where did you ground the new stage? Just thought I'l throw something out there.

 

[Maxxarcade]

OK, this is odd... I hooked the preamp circuits up to the scope and they are both working perfect. Output at C7 is just like it should be. But when I connect it to the power amp, I get the loud hum, and if it's on long enough for the 6SL7's to warm up, it starts to squeal.

So basically, the preamps and power amps are working fine when separated, but not when connected together.

Ground continuity is good between all stages.

I know it's messy, some of it is just temporary for testing...

 

[Maxxarcade]

Figured it out. If I connect the earth ground to the signal ground, everything works fine. I still have an annoying 120Hz hum though. Maybe having an isolated signal ground bus won't work in this amp.

I'm also going to put a 10uf cap across the B+ right at each channel, instead of a 22uf on one side of the chassis.


But OMG is this thing loud! I was worried about the balance control attenuating it too much, but I think it might be ok after all. With the balance control bypassed and the volume half way up, it's probably near full output.

 

[PabloL]

Ground

You likely know this but I'll tell you how I ground just in case. I start with a long piece of solid copper wire, the stuff they sell to hang pictures actually (16 gauge I think) I start the run at the first PSU cap. All PSU grounds (Transformer CT, caps, bleeder resistors, etc...) attach to about the first inch of that wire. I run the bus to the power tubes, and then the input stage. So as the sensitivity of the stage increases the further it is from the PSU grounds. I then terminate the bus on the non-grounded lug on a 2 lug terminal strip. I then link the nongrounded lug to grounded lug through a pair of diodes (one in each direction) a .1 ceramic cap and a small (~20 ohm) resistor. So there is a sum total of one ground connection located between the RCA jacks (not counting the safety ground at the 3 prong plug). The only time I've ever had ground issues is when I deviated from this practice. YMMV of course.

 

[Maxxarcade]

You likely know this but I'll tell you how I ground just in case. I start with a long piece of solid copper wire, the stuff they sell to hang pictures actually (16 gauge I think) I start the run at the first PSU cap. All PSU grounds (Transformer CT, caps, bleeder resistors, etc...) attach to about the first inch of that wire. I run the bus to the power tubes, and then the input stage. So as the sensitivity of the stage increases the further it is from the PSU grounds. I then terminate the bus on the non-grounded lug on a 2 lug terminal strip. I then link the nongrounded lug to grounded lug through a pair of diodes (one in each direction) a .1 ceramic cap and a small (~20 ohm) resistor. So there is a sum total of one ground connection located between the RCA jacks (not counting the safety ground at the 3 prong plug). The only time I've ever had ground issues is when I deviated from this practice. YMMV of course.

That is the scheme I was trying for, but physical placement of things in the chassis made it hard to do.

I think some of the hum is coming from the filament transformer too. Would some of that copper shielding tape fix that?

90% of the hum seems to be coming through the preamp. With the power amp connected directly to the source, there is barely any hum. Most of what is there is in the right channel.

The pot housings are also at earth ground through the chassis. Not much I can do about that.

I tried simply disconnecting the earth ground from the chassis, and it only made things worse, so I know it's not a ground loop.

 

[Maxxarcade]

Looks like most of the noise is because of the shielded cables in the preamp not being shielded enough. When I move them around, the hum changes quite a bit. Next week when my new cable comes in, I'll see how it improves.

 

[PabloL]

That makes sense, I think that the nature of preamp sections make them sensitive to interference from other sources. Is the shielding grounded? I've never used shielded cable, but I think one side needs to be grounded. Actually I think I did once, it failed because it ended up picking up FM radio signals....

 

[Maxxarcade]

Ya, one side of the shielding is grounded to the signal ground.

My ground wires are mostly 22 AWG, except for the big filter cap, which I used 2 or 3 strands on.

Should I use a heavier gauge wire, and suspend it along the terminal strip lugs that are used for signal ground points, instead of running it along the chassis?

Also, the PT has two completely separate windings for the high voltage. If I tie the wrong pair of wires together to make the center tap, can that cause problems? I know that I used one wire from each winding, but maybe I used the wrong pair?

The amp is sounding really good aside from the hum, so I really want to get this figured out :scratch2:


EDIT: Another thing I just realized- The filaments for the 6SN7's and 6SL7's are on their own transformer that has two 6.3v windings. Neither of those windings has a center tap. I used one winding for the left channel and one for the right. Do I need to make a center tap with resistors for each channel's filament winding and ground it to the chassis?

Looking at the original HK250 power supply, there are two 68 ohm resistors across the filament lines, with the center point "x" going to the 6L6 cathodes. I don't fully understand the concept of that, but I know it is something to do with hum reduction. I used the same method on my amp. The two resistors are on the filament winding for the 6L6's only.