[JoshK]

Jeff, also check the bias voltages across the cathode resistors.  This will tell you if they are biased where they are suppose to be.

[andrewbee]

Jeff,

You keep changing the schematic. You now have a 470K grid R instead of a 220K behind the volume control as you had earlier, remember that grid R is in parallel with the 100K volume control and the value of the R combination is going to change as you change the volume.
Double check the schematic.
Most amp problems are voltage related.
Check the heaters and verify voltages.
This is what I would do, go back to basics.
Lose both the ccs's and just use a plate resistor.
remove the volume control behind the 6SN7.
Measure all the resistors that you have installed with a meter to verify value and the capacitors as well.
Fix the output tube as triode (easier) and eliminate the switch.
Double check the tube operating conditions, while they look "okay" at first blush you need to verify them.
Do you really need a 10k grid stopper on the 6SN7, 3K3 on the 6550 and 1K on the 6H30?

First thing to check is the heaters.
What would be easy is to replace the ccs on the 6H30 with a plate R. Check the tube curves, you just want ~15mA draw on the 6H30.
Lose the first stage.
Triode wire the 6550, use a ~100R resistor of a watt or two.
Double check all components and your wiring.
When its powered up measure all the operating points, remember to put a speaker or some load  (~10R ~ 20W resistor) across the OPT secondary.
All you are trying to do at this stage is verify you can get something working.
Is this your design? Its pretty elaborate to drive a 6550.

[JoshK]

Do you really need a 10k grid stopper on the 6SN7, 3K3 on the 6550 and 1K on the 6H30?

Yes indeed, that is strange.  It looks as though he has transposed the grid leaks and grid stoppers.  The 10K is normally the value for the grid leak and the 470 is closer to a grid stopper, for example.  Actually, most use 100-200 ohm grid stoppers (carbon comp). 

[jeffdavison]

After someone on another forum mentioned something about too much gain, I went back to my schematic and compared it with some original source designs.

The Audionote Kit 2 uses a 6SN7, then volume pot, then a 12AX7 in srpp. In the older shematic, what is currently the values in the amp are modeled after the SimpleSE and the Mikael. I put the values from the Audionot Kit two into the schematic as it is the closest to mine in layout, except for me using 1/2 of a 6H30 instead of a 12AX7 in SRPP.

http://www.audionotekits.com/PDF/Schematics/Kit2_AudioCircuit.pdf

Heater voltage looks fine. The 6550 uses 6.3 vac with a 100r off each wire and connected together at ground.
Heater voltage for the 6SN7 and 6H3o is also at 6.3 vac but the ground is elevated by about +80 v through a voltage divider from the B+.

There is plenty of current for the tubes as each channels 6.3 is rated to 5 amps.


JD

[JoshK]

I responded to you on AA too.

Quick comments.  The AN schematic is not ideal IMHO.  The 12AX7, even in SRPP does not make a good driver.  Your 6H30 is much better in this case.  #2 they are using feedback back to the driver which lowers gain (don't know how much in this case).  #3 they have a resistive load on the 6SN7 which results in lower gain then your actively loaded 6SN7. 

[andrewbee]

Jeff,

So this is a "cut and paste" circuit?
Thats not the way to do it, unless you completely understand what you are "cut and pasting", from whence it came and doing what job.
I know I said it before but lose the first stage,  on the 6H30 try a B+ ~ 300V, no CCS, use a 10K plate resistor, 300R cathode resistor. You can bypass the cathode if you want, try it both ways, without the cathode cap for less gain (probably don't need the bypass).
Those operating points should give you a ~120v on the plate, 20 mA cathode current and ~ 2.5W dissipation per side, well within limits.
Review the 6550 pin connections and make sure they are correct, connect it as a pseudo triode with a resistor. It should work and may surprise you.
What does your p/s look like?

[jeffdavison]

Here's a comment / suggestion about using or losing the CCS on the 6H30 I recieved in another forum...

>> >> on using a 10K plate R and a 300R on the cathode

You need a B+ of 410V to pull this off and
it will give you no headroom to operate.
The 6H30 and the 6550 will clip about the same time -
not a good idea.
Stick with the CCS here.


Now this gets to be fun with different cooks looking at the recipe


I want to keep the 6SN7 as it's basically being used as an output stage for my standalone phono section and DAC. So what if we look at the entirety as an "preamp" (with a 10K volume pot) feeding and "amp" with a 6H30 driven SET 6550?

Jeff

Jeff,
 on the 6H30 try a B+ ~ 300V, no CCS, use a 10K plate resistor, 300R cathode resistor. You can bypass the cathode if you want, try it both ways, without the cathode cap for less gain (probably don't need the bypass).
Those operating points should give you a ~120v on the plate, 20 mA cathode current and ~ 2.5W dissipation per side, well within limits.
Review the 6550 pin connections and make sure they are correct, connect it as a pseudo triode with a resistor. It should work and may surprise you.
What does your p/s look like?

[andrewbee]

Well Jeff,

You can try it my way or any other cooks' way, its your amp and you need to make the decision.
I am only trying to help you get it going and I reiterate not as drawn.

quote]All you are trying to do at this stage is verify you can get something working[/quote] 

There are lots of experts over at AA (I saw your post over there) so I am sure you will gets lots of advice there.
Good luck with the amp.

[JoshK]

I still think you should loose the first stage until you get it working right and then add in the stage as you wish.  It would be as easy as jumpering it out. 

[nodiak]

DaveC113, what cathode bias resistor value are you using with the TS 6550's?
Don

[jrebman]

Don,

Unless Dave changed the value, we had all purchased 560 ohm 5 watt Kiwames for these.  I chose that value as it would give enough flexibility in operating points to use KT-88/6550 as well as EL34.  I may eventually change this to further optimize for KT-88/6550 only, but that will wait until I get a chance to hear the EL34s.  I had an EL34 amp before (a Sophia) and it was very nice in a lot of ways, but the bass was just a bit too mushy for my tastes.  I'll see how it does in this amp though, as this is an entirely differnt thing.

Still waiting for Miss Carina to make her appearance...

-- Jim

[nodiak]

Jim, I am considering a 3 way switch as discussed in this thread, posts #5 and 6 -
http://www.diyaudio.com/forums/showthread.php?s=&threadid=112666&highlight=SimpleSE
Wasn't sure where 6550 fit in. Same value as KT88? If so then the values in the thread will work.
Don
"Still waiting for Miss Carina to make her appearance..."
Is it in route? Will be an interesting comparison.

[jrebman]

Don,

Yup, KT-88/6550 would use the same cathode resistor.

Is it en route?  I dunno. :confused:

-- Jim

[nodiak]

Thanks Jim. Making another surge here. Chassis almost done, switched wood to some local Tan Oak. Ordering TS 6550's.
Hopefully Carina will arrive soon. Worth the wait I'm sure.
Don

[jrebman]

Oh, I'm sure too -- just that my Horns are sitting here with nothing to do right now .

-- Jim

[JoshK]

Some small progress was made...

the top plate with some renassaince wax...


a dryfit with the wood chassis...


its my first attempt at a box joint, and far from perfect, but nothing some sanding can't make presentable.


You can't quite see the whole contrast.  I should probably take it outside for natural light pictures, but I'll wait till I get further along.  These are just some preview pix. 

By the way, the wood is all cherry from the same batch.  The front piece is a lot lighter than the sides intentionally.  Its all dark cherry but the front piece was only dark half way through and light on the other side which I choose to face outwards to give some contrast. 

[alexg_ht]

Been reading on Georges web site (TubeLab) and his links to CFB..... I'm thinking of instead of just switching in and out of feedback from the 8 ohm tap, I'd put a 3 way switch to have the option of feddback from the 8 ohm, the 4 ohm, and no feedback.... For reason he says are valid, that I couldn't find reference to anywhere else, is that George  reverses the polarity  of the outputs of the OPT's,, that is he wires common as 8 , and 8 as common. He says different transformers may or may not be wound in opposite directions. He does  reference other OPT's ( http://www.tubelab.com/Output_Transformers.htm ) but doesn't elaborate further about the reverse output polarity... or did I miss something? 
My OPTS (Silk S-325's  http://www.sacthailand.com/Transformer_SilkS-325.html ) have  4 and 8 outs.


JD

Hi JD, how did the silk OPTs perform?

Am thinking of trying these on my Mikael Abdellah's SE KT88 that I already built. Currently, am using a custom wound transformer from a friend who owns a tranformer factory here in the Philippines (pictures posted at diyaudioprojects.com). I also tried James, but quite expensive shipped from Taiwan.

Thanks.

I haven't tried cfb yet.... but judging from how good it sounds without it I doubt it will be an improvement.

[JoshK]

OK, so I can heartily recommend the Porter Cable jig.  There is no comparison between the Craftsman and the PC.  The PC is only a 12" fixture, instead of the 16" and didn't come with as many template guides.  However, the guides themselves are aluminum (craftsman was all plastic).  The fixture base is steel.

The PC jig has everything marked in such a way that it makes setup painless and nearly dummy proof.  The instructions are clear and understandable and once you read through them for a particular joint (they have seperate instructions for each!) you can easily go about it.  They have depth guides on the side for each kind of joint, so you place your router on the side and adjust depth till the bit is touching the depth guide and viola it comes out perfect. 

The instructions make it so no matter how wide your wood piece is the tails and pins come out symmetric. This is pretty neat because with the Craftsman everywhere I read said the wood had to be a multiple of the finger width.  This is not so with the PC.  Actually you could do the same conceivably with the other jigs but there were no such instructions on how to and the offset guide weren't as flexible to do it.

The quick release handles on the PC also are a huge convenience over the fixed clamp of the Craftsman.  Once you've set your wood and scraps in place you can quick release them to flip the piece to cut the other side and pieces. 

The PC really made the whole thing quick and painless.  I spent hours with the craftsman and still don't think I could figure it out.  I sold it on ebay.  I can give the thumbs up to the PC jig.  At $120, its a deal.

Oh and the PC guide bushings fit all the standard routers, none of the craftsmans bushings did.  The bushings also work with the templates they were intended for, the craftsmans didn't even fit the templates they were made for! 

[JoshK]

P.S. in the pictures above, the joint on the left isn't set tight, whereas the one of the right is a bit tighter.  This was due to a bad straight bit I first used.  The bit had two flutes and one had broken half the blade off.  I thought because it still had a whole flute on the other side it would still cut fine.  It didn't and made the joints not want to sit in fully.   Well I got a new bit and went over the joints again last night and it fits together like a glove. 

Live and learn.

[jeffdavison]

I've gotten my amp (which I've named after Ben) working and has been burning in over the last two weeks. Looks like changing some grid stop and grid leak resistors and eliminating the CCS and cathode capacitor on the 6SN7 (to reduce some gain which was helping in overdriving the 6550) did the trick.

After some experimenting and listening, I've settled in on Pentode with CFB. This is the most transparent and neutral amp I've ever heard since my audiophile habit started in the early 70's. This one I've permanently hardwired for Pentode and CFB.

The NOS tubes I'm using also help along.

Since I had all the work done up on CAD and etch my own boards (for the power supplies... point to point for the signal chain) and have an abundance of rosewood left from my luthier days, I'm just about finished with a second one.  Tweaked just a little bit for easier assembly but using Mundorf Silver/Gold/Oil for the coupling caps and tantalum resistors in the signal path... and some precious metal internal wiring. This one has an extra transformer under chassis for a switchable 12.6v heater so I can swap between 12SX7's and 6SN7's. Also switches for triode/pentode and pure/CFB so I can change my listening pleasures occasionally.

If any of you read up in the Audio Assylum Tube DIY forum, I've tried to great satisfaction, Jeff Medwins (drlomu) power supply suggestions that are a little counter to common prctice but the sonics have proved themselves to my ears.

There's a audio store near where I live who has auditioned it and was asking if I'd ever consider selling commercially....man... that's a lot of soldering....maybe I could farm it out to China  ;^)

Thanks guys for the help and also thanks for the Uber Amp  inspiration and spirit of this thread that was started some time ago.

Jeff