Further work on my KT120 DLM (with pics)

[strelok]

Been doing some work on the amp I posted about a while back. It had a lot of things that still needed touching up.

The power supply still was making a small but noticeable amount of hum even with 140uf of mains filtering. Originally I just had a 40uf axial in there on the power supply board and added a dual 50uf can in parallel. Still didn't do it. I always knew SE amps needed more filtering, I just never realized how much more till I built this monster.

I tried turning the can into a resistive pi filter but by the time the resistance got high enough to kill the hum completely it was getting way too hot, even with a 10w resistor. So I turned it into an inductive pi filter with a 3H 200ma choke (thanks again to Colossal for supplying me with this) and that killed the hum completely. I was even able to take out the 40uf can and just have 50uf on either side.

The preamp took considerably more work. This is my first Trainwreck style preamp and it was certainly a learning experience as far as stability goes.

Originally it was an express preamp with an AC coupled cathode follower driving the KT120. It didn't have much gain and I later realized the cathode follower was unnecessary, so I converted it to essentially half the PI stage.

The additional gain pushed it well into instability region and it was very much theremin-wreck. I'm sure the fact that its a steel chassis didn't help any either. I added shielded wire to all the signal wires where it didn't already have it, as well as 1k grid stoppers on V1a V2a and V2b.

It was still oscillating horribly whenever I would touch the gain knob at certain settings (but stable otherwise), and after several hours of moving wires and poking around to no avail. I realized after probing the pot with a grounded test lead that the body of the pot wasn't making good contact through the paint on the chassis. Dremel with a grinding stone took care of that, and the amp was completely stable. Of course once I got it home I realized I should have run bus wire across all the pots as they are now having the same problem. Different interference at different locations or just the shifting in the car possibly. Though I'm just happy to know what the issue was and its an easy fix.

The final-ish version of the preamp is a pretty standard express affair for the first 3 gain stages. The final stage I ended up with a 1k top cathode resistor and 10k for the tail (half a long tailed pair). Grid leak I settled on 332k as 1M was just way to much gain. Additionally the transition from being mildly distorted to just right to seriously over compressed overkill happened on much too short of a range on the pot.

The only issue really left to solve with it, is what might be some parasitic oscillation. There's a weird buzzy sound that happens on the top of the notes, just present enough to annoy the crap out of me. However it seems to get less noticeable as I turn the volume up, so I'm not really sure what the issue is.

As for the cabinets, I built them last summer out of hardware store pine and Minwax. I think they look alright considering I have done basically no woodworking before and all I had was a circular and jig saw haha. The amp itself also isn't nearly as neat as I'd like it to be, its been rebuilt many times over the last decade, and just in the last year I've had to disconnect sockets and lift the boards up more times than I can count. Next time I do a prototype build I'm using all PTP with terminal strips haha.

Alright now onto the pics. Sorry for my sloppy hand drawn schematic, its all I have at the moment. I hope it's legible. The guitar I built about 8 years ago during my second year of college. It still needs some work but its my favorite to play right now.

[img2976]http://i.imgur.com/Qn3Zwze.jpg[/img]
[img2976]http://i.imgur.com/TDyE1IF.jpg[/img]
[img2976]http://i.imgur.com/xaFM9kX.jpg[/img]
[img2976]http://i.imgur.com/ZSKhBWG.jpg[/img]
[img2976]http://i.imgur.com/xtY8yn9.jpg[/img]
[img2976]http://i.imgur.com/U6NmE7n.jpg[/img]
[img2976]http://i.imgur.com/KmXe5he.jpg[/img]
[img2976]http://i.imgur.com/gx1RlQm.jpg[/img]
[img2976]http://i.imgur.com/jOjQ1La.jpg[/img]
[img2976]http://i.imgur.com/rYpinfx.jpg[/img]
[img2976]http://i.imgur.com/MouhLLb.jpg[/img]
[img2232]http://i.imgur.com/mD0ruSB.jpg[/img]
[img2976]http://i.imgur.com/V6ix37v.jpg[/img]
[img2976]http://i.imgur.com/VnsQekz.jpg[/img]
[img2976]http://i.imgur.com/V3IzuIT.jpg[/img]
[img2976]http://i.imgur.com/8tKIT57.jpg[/img]
[img2976]http://i.imgur.com/yGtq08n.jpg[/img]
[img2976]http://i.imgur.com/rLVTiLt.jpg[/img]
[img2976]http://i.imgur.com/T5YuLXW.jpg[/img]
[img2976]http://i.imgur.com/lyacUv0.jpg[/img]
[img2976]http://i.imgur.com/i9rnngr.jpg[/img]

[strelok]

Does it look that bad? lol. Or are things just slow because of the malware issues?

Anyway I was hoping you guys might help me sort out this strange noise issue I've been having. I'm not sure if its parasitic oscillation or what but its annoying as hell. I've gotten the instability issues worked out. Its fully stable at any control setting. But I haven't been able to get rid of this annoying buzz that happens on top of the notes. It seems a lot more noticeable at lower volume settings. Past quarter on the master and its not nearly as bad but at that point I'm pissing off my roommates haha.

Here's a sound clip. Its pretty lousy quality as I don't have much recording gear and its pretty low volume. Just a 57 into the computer. But you can hear the buzz I'm talking about.

[Colossal]

Cool build. Glad the choke worked out. I like the chewiness on the low E string in your clip; it would be nice to have heard more of that. There is a nice throatiness to it. What speaker are you using? I wonder if you are getting some switching noise from the 1N5408 diodes. You might consider bypassing each with a 1kV 10nF ceramic cap.

[strelok]

Thanks man! Yeah I really like that chewiness as well. It's tuned to drop D which always seems to give it more of that. There's more to that clip, I just chose a small section that highlighted the noise issue. I'll post more of it tonight. Speaker is a Scumback G12M-65 in a 1x12 closed back cabinet.

Switching noise eh? I hadn't thought of that. I'll see if I can dig up some caps at work and install them tonight. Also forgot to mention that the noise gets worse as the gain goes up as you probably noticed.

[billyz]

might try grounding the preamp up by the input jack. I have had strange issues when things are grounded together like it appears you have.

[Colossal]

might try grounding the preamp up by the input jack. I have had strange issues when things are grounded together like it appears you have.

Completely agree with this.

[rp]

What's your plate voltage? Hopefully it's something insane.

[strelok]

Plate Voltage are pretty normal:

V1a: 210
V1b: 180
V2a: 263
V2b: 264

V3: ~365
Screen: ~340

Didn't have time to try the snubbers while at work today, but I did find a couple. The owner of the company I work for says its probably a grounding issue as well. He suggested tying the grid leaks to the same point as the cathode ground for each stage and running each of those to the buss. So I've got a few things to try.

Here's another section of the clip I promised Colossal. Its just a goofy off time thing for the first minute or so then gets more normal, but has lots of low end chugging as you requested haha. Had to put it in mp3 as the wav was too big.

[Colossal]

Cool! Nice and crunchy. Seems like there's good clarity and separation.

[strelok]

Thanks! Yeah its starting to sound pretty good, save for that noise that sits on top of the notes haha. I pulled the board out yesterday and started working on the grounding scheme. Going to move some things around and generally try to clean the wiring up as well. Probably will move the second preamp tube over to that empty hole so its further away from the output section and transformer wires too.

[strelok]

So I had the day off and spent all day reworking the grounding on the amp and various other things. Its now a star grounded setup with each stage's grid leak and cathode tied together and run to a common lug. It took most of the day and made no noticeable difference in the problem lol. At first I tried it with both the preamp and poweramp grounds going to the same lug under the tone stack, then I moved the poweramp ground to a separate lug on the other side of the chassis as was suggested here. Also made no difference. I may yet try moving the preamp ground connection even further to near the input jack, as was also suggested, but I'm starting to suspect something else is the issue.

I then thought I'd give the snubbers a try but the fat legs of the 1N5408's didn't leave any room for the leads to go. We had some 1N4007's sitting around so I decided to give those a try to no avail.

Earlier in the week I hooked the amp up to a signal generator and scope and saw a strange looking waveform coming out of the coupling capacitor of V2b. I didn't get a chance to get a picture of it so I drew one up to the best of my ability.

[img2976]http://i.imgur.com/70id4v4.jpg[/img]

As you can see when the preamp is driven to hard clipping it has this weird exponential decay on the negative going side of the square wave followed by a sudden jump down to where you'd think it should be. I think this is most likely the high frequency noise that seems to sit on top of the notes. It gets more noticeable as the frequency is decreased. At 1khz you cant really see it. As you get closer to 100hz is starts to deform and look like what I drew in the picture. However I have no idea what's causing this.

I did some spectral analysis in audacity and it looks like its occurring at around 7khz or so. So I'm thinking if all else fails I could add a small cap (assuming my calculations are correct 250pf would put the knee at around 6.3khz) across the plate resistor of V2b to try and remove some of that. However it feels a bit like cheating. I'd like to know what the source of the problem actually is. Just as an experiment I took the clip that I originally posted and Low pass filtered out everything from about 6.3khz up and it made quite a difference.

[strelok]

So I tried a few more things this week:

Moved the preamp ground to the far side of the chassis under the input jack.

Added a 332pf cap across the plate resistor of V2b. (Didn't have 250pf laying around.)

Added 10nf snubbers across each diode.

The snubbers seemed to help a little. I think. Maybe. Its hard to tell. The noise is still there though after doing the aforementioned things. The grounding change and the 332pf cap didn't change anything. Though I'm liking what the cap did for the tone so I might experiment with that more.

The snubbers where also making a rather unsettling buzzing/arching noise, despite being rated for 1KV, but seemed fine otherwise. Is that normal? I haven't ever had to use snubbers before but it doesn't seem right. Also I read somewhere that they need to be in series with a resistance in order to work right. Is that the case? I know it will shift the filter some so maybe I need to experiment with the values till it starts catching the switching noise? If that's indeed the problem.

I also took some pictures of the scope at a couple places in the amp if that helps any, all where taken with tone and gain controls maxed and the volume set to zero.

Output of V2b's coupling cap at ~1khz:
[img2976]http://i.imgur.com/OkMxoGY.jpg[/img]

Same place at ~100hz:
[img2976]http://i.imgur.com/gyJxxWl.jpg[/img]

Same again at maybe 250hz or so:
[img2976]http://i.imgur.com/gVuopr7.jpg[/img]

And finally V2a coupling cap at ~100hz
[img2976]http://i.imgur.com/jyu6w41.jpg[/img]

The problem is I haven't bothered to scope out a bunch frequencies before as I've never had the need to, so I can't tell if this is normal or not, I'm guessing its not. Anyway hopefully it will help some.

[rp]

Not sure about the details of a KT120 and the OT but given the modest plate voltage why not rewire the output stage (maybe change the load with a different tap on the OT) for a 6V6 6L6 or EL34, if you have, just to get the amp tested out? No warranty on that KT120?

[rp]

hmm I posted to the last post:

PostPosted: Fri Jul 10, 2015 12:03 pm Post subject:

Ugh...

But wound up above it. Weird. Maybe strelok you were editing while I was posting? More TAG bugs?

[strelok]

Ugh.

So I put in a 150R resistor in series with the bypass caps. Didn't make any difference. So I decided to say fuck it for a bit and just play some guitar before I needed to leave.

After about 10 minutes the KT120 shits itself. Arc over failure. I shut the amp off immediately. Let things cool down a bit. Pull the tube out to probe the pins to see what shorted. Couldn't find anything. Plugged it into the tube tester. Doesn't show any shorts or leakage, but heaters are also not lighting up. Might just be because the tester can't supply the current needed.

Put the tube back into the amp, probably against better judgement but I wanted to see how it failed. Noticed that the heaters weren't lighting up nearly as bright as they usually do. Popped the standby for a quick second and turned it off again. In that time I saw two arcs traveling inside the inner structure of the tube. Looks like a short between the screen and control grid. Not sure what happened to the heater. Pulled the tube again, hit the standby and started checking voltages. The good news is everything seems to have survived. +1 for overrated transformers.

Sadly though that was my only 120 and $50 that I'll never see again.

Trying to figure out now what caused that to happen. Maybe it was just a bad tube. The only other things I've changed in the power section recently was the bias and screen resistor. The bias I lowered to down to 200R since the choke caused the B+ to drop some and that helped compensate for it, though it was still only at 45 watts. Screen resistor I had been experimenting with a while back and left a 1k8 in there. Changed it back to 1k recently. Think that value was too low? Its hard to imagine that would have done it, given what this tube was rated for and the voltages I was running it at....who knows...maybe that was the source of the noise all along, bad QC and something rattling around in there finally came loose. I can dream lol.