[Goobley]

Hi all,
 
I have just purchased parts to build this amp (12au7 variant) after discussing it at some length with KimLaroux in the dedicated thread.
 
This is going to be my build log. I'm going to do a few things slightly differently with making of this amp:
 
The left and right channels will be split.
The PS is not the original design and will be contained inside the case.
 
My soldering work is going to be built on stripboard and the case will be wood (oak and walnut). I haven't yet split the stripboard into channels, it should be very easy to do now I have it drawn out, but I wondered if someone could look over my design, I'm fairly sure that it's correct but it'd be good to know.
 
 

If you download the image it can be viewed in all its high-res glory
(Schematic can be found here: http://diyforums.org/SSMH/SSMHvariants.php )
 
 
My parts will be arriving this week, however one thing I didn't buy was wire, I already own a fair quantity of 0.75mm2 (18awg) copper automobile wire and the same gauge tinned copper wire for test leads along with some standard 24awg prototyping wire. I was planning on using either the automobile wire or the tinned copper, is there any reason not to? They are both fairly high quality wires, as is the prototyping wire (I know the prototyping wire does not have a major effect on sound quality as it made up my headphones cable while I waited for new wire from BTG to arrive ). Is there any particular ones of these wires to choose? (tinned versus pure copper?). Obviously any HT PS wiring will be done with mains flex salvaged from old appliances.
 
Thanks for any advice on this subject, looking forwards to getting started!

 

[Goobley]

Well I actually finished the build last weekend (pictures incoming, although the top of the wooden box hasn't been sanded and nothing has been waxed yet).  However I've had some noise problems. Connecting the amp to my computers sound cad and using some analysing software has allowed me to ascertain that the noise is on the ground (L/R is almost as clean as my Marantz PM6004 (5dB more noise iirc)). It would therefore appear that I'm struggling to get a clean ground. I seperated the headphone jack ground from the stripboard and connected it to the RCA and then tried connecting it to v-, ground and mains neutral. The last gives the cleanest sound however there is always a mid range hum present. The longer I leave the amp on the louder this hum gets (heat?).
 
I'm currently stuck trying to clean these things up so some advice would be great.
 
I also noticed that with my dmm, the potential difference between v+ and ground is 0V as is the diff between v- and ground. :S
The static doesn't change if I approach my hand to the tubes, however if I move my hand towards the wire that carries the input to the tube (2 & 7) a large amount of noise ensues.
 
If anybody has any ideas about quieting this amp I'd be grateful

 

[KimLaroux]

I also had noise issues with my build. The original was in a plastic enclosure, and it picked up so much noise it was not usable. I had to rebuild it inside an aluminum enclosure and ground everything, and add a 100 ohms resistor between earth and ground (0V rail). I suspect you're having the same problems with your wooden enclosure. Here's a test you can do: place the amplifier inside a metal case, storage box, tool box... anything that's made out of metal, ground it, and see if it helps. When I did this test, the difference was night and day.
 
Note that even with an aluminum enclosure, the amplifier was still noisy. The tubes themselves pickup and amplify EMI because they are not shielded. I can't use my MSSH with my Shure, but it's usable with my AKG.

 

[Goobley]

I tried putting the amp in a grounded metal box (~3mm steel) as well as running it in the open air with all electronics in the house off (wifi etc. I live in the country so am not picking up interference from others). The difference was at most minimal. I may be picking up interference from my own transformer as it is in the same box. I'll look into that possible issue.
 
However one thing struck me that could be quite a good way of cleaning up the power: the transformer I have can be set all the way up to 53V, if I was to set it to say 50V or 52V then add a secondary power smoothing board and a virtual ground so that the circuit is effectively running off 48V I may solve noisy power issues. Any opinions on this?
 
 
Thanks
Chris

 

[KimLaroux]

By transformer, you mean switching power supply? And what do you mean by "smoothing board"?
 
You can always add capacitors to the output of the power supply, see if it helps. You could also try to build a capacitor-resistor-capacitor filter, and push the power supply voltage higher to make up for the voltage drop across the resistor. The only linear regulator I know that can work at these voltages need a 25 volt difference between in and out, so you can't use it in your application.
 
What power supply did you use? Do you have pictures of the build? I'm concerned about the noise radiating from the power supply itself. If this is the case, you can try to move the power supply around, even turn it upside down. Placing a sheet of grounded metal between the PSU and the rest of the amplifier might help.

 

[Goobley]

Unfortunately I'm not in the same location as my amp during the week. However I have managed to get pictures sorted now.
I'm using a TDK-Lanmda LS50-48 for my psu (switching) and the positioning is as shown:
As you can see I still have plenty of stripboard left on which to install a filter of some kind or an active ground channel.
I have tried moving things around inside the box and so far it doesnt seem to make much difference (there's not much wiggle room though). Possibly because the entire casing of the psu is grounded mesh. 
 
One thing that I'm possibly worried about is my back panel layout:
 
 

 
If anybody has ideas about cleaning the background noise up from this amp I'd be very grateful, I've enjoyed the sound I've got out of it so far but it'd be much better if it was cleaner. However I'm sure if it just can't be done that I can find another way to use this kit. 
 
Don't hesitate to ask if I missed necessary information.
 
Kim, you said that your amp doesn't work with your srh440s, well maybe that's part of the issue I'm having too? Are they just too low impedance?
 
Thanks Chris

 

[KimLaroux]

Is the PSU getting hot? What happens if you leave the amplifier ON for half an hour? If it doesn't get too hot to touch, I'd recommend flipping it upside down. I'm guessing the backside is solid sheet metal without holes. You may get less noise that way.
 
Two things come to mind from the pictures: Wooden enclosure and long wires. My personal experience told me a grounded metal enclosure is necessary for this amplifier. What really does not help in your case is all the long wires that twist and turn all over the place, especially the ones that passes over the PSU. I can see your input signal wires being way too long, not shielded, and passing right over the PSU. Same thing for the input ground wires. First thing I'd do if I was you is clean up all this wiring. Use shielded cables for the input signal, or braid them like in the following picture. Place the power and signal wires as far away from each others as possible. The only thing I can see wrong about your back panel is how it's wired inside: nothing to shield the input signal from the AC radiation.
 

 
But then again the problem is probably a combination of many things: wooden enclosure, messy and unshielded wiring, cheap switching power supply, tubes picking up RF, and amplifier design inherently noisy... I think I did everything possible to make this amplifier quiet in my build but it's still too noisy to be used with my Shure. They are low impedance headphones, but I don't think it's all the really matters. They also provide a very good seal from outside (The fire alarm once rang and I didn't hear it), which creates a quiet environment in which noise in the audio can more easily be heard.
 
Other things to try: place grounded aluminum cans over the tubes and see if it lowers the noise to a reasonable level. Most of the noise may be picked directly from the tubes. Using a sheet metal, like the picture above, as the top cover may help shield the tubes from the power supply and AC line. It could also shield the other way around, protecting the wires inside from noise from above the amplifier. If you conclude that the power out of the PSU is too noisy, you could build a capacitor-resistor-capacitor filter network.

 

[the_equalizer]

Lots of things are causing your noise problems here but very very little can be attributed to the design of the amplifier, contrary to what you  Single ended triode amplifiers have been built by many decades without much problem; my Woo Audio 6 is one such design and it is DEAD QUIET even with closed, sensitive headphones.
 
I have built several variations of the SSMH, more than one in an ABS enclosure and at leat two with no case at all (wired on a protoboard) and I haven't had bad noise issues.
 
While shielding the tubes and using a conductive enclosure are obviously good ideas, there are still a couple things that you need to consider.
 
1.- Build the amp over a groundplane. Check out Mr. Millett's original build. This will be particularly helpful when using a non-conductive enclosure. If you want to zap out the last remaining noise, cover the enclosure inside in conductive material (paint, copper tape, alumminum foil) and connect to the circuit's ground at JUST ONE POINT. A builder in the regular SSMH thread reported good results using this technique about a year ago.
 
2.- Unless you are VERY VERY CAREFUL with the layout of the AC lines, transformer and the audio components, an amp with an outboard switching supply (as the original SSMH design) will be much less noisy than one with an on-board linear supply. Laying out a tube amplifier with an on-board linear supply is no easy feat and will requiere the use of shielded cable for all the audio cabling and/or keeping the transformer and AC line AS FAR AWAY AS POSSIBLE from audio lines. A grounded mesh around the AC transformer will help a little but for true noise zapping you'll need to FULLY enclose the transformer (again, take a look at the Woo amps)
 
Don't get me wrong, I like the work you've done with your builds, I like it a lot; and I think adding a linear supply to this little amp is a great idea, but I wanted to make clear that it's not the amplifier design that's noisy here, it's the AC supply that's inducing most of that noise.
 
For some detailed talk and recommendations about amplifier layout you should read Morgan Jones' "Valve Amplifiers" and "Building Valve Amplifiers", given the work that you have done I'm sure you'll love every bit of them.
 
cheers and keep up that work in your amps !

 

[Goobley]

I got back to work in my barn today...
I started by removing the PS and extending my amps power cables. I then put the amp in a grounded box (2mm steel) whilst leaving the psu outside the box. There was a significant improvement in noise. However there was still more noise than I really wanted. So I grabbed the capacitors I had to hand and tried to build a crc (emphasise on the C) network. Here's what I came up with:
 

I have a gut feeling that I'm gonna need some bigger caps... In this instance earth and V- are shorted, it just sounded considerably better...
 
I should hopefully be able to find out how the PS actually works on wednesday as I'm taking it to my schools electronics lab. I say hopefully as they're being very finnicky about letting me use any equipment (I don't know what harm they think me capable of doing to an oscilloscope). I'll also be taking another project of mine which also has an issue related to, you guessed it, power smoothing! This seems to be a recurring trend.
 
Any comments or ideas about improving my cap network? (I know it's probably terrible but there seems to be a large lack of resources, so I decided to take action anyway).
 
Taking these steps has certainly improved the sound however it's not quite where I'd like it to be yet. I'm going to cover the inside of most of the box in tin foil and then separate off an area using 0.8mm steel sheet for the PS. 
 
One thing that occurred to me earlier is the fact that maybe one strip of stripboard is just not a good enough ground and is possibly too close to live strips (they are only 2.54mm wide...). Any thoughts on this?
 
Cheers
Chris

 

[nikongod]

3 things. Maybe more. You get what I can come up with. Will not go back and edit. I'm a little scared. 
 
1: WHY does your layout look like that? Long leads EVERYWHERE! The SSMH should be nice and tight not loose and sloppy. There should be maybe 3" (10cm) of wire between any 2 connected components. no, no, 3" is the absolute max. Wiring is like women's swimwear - less is more. 
 
2: PSU In the same chassis as the amp - icky. At best its "doable" but everyone who ever runs remote PSU la-la-loves it. Sure it costs a second enclosure and a bit of cable but it makes amps with long leads possible, and amps with short leads absolutely awesome. 
 
3: Your modified filter thingy probably wont do much for the reasons above. If you do want something between the PS & the tubes & FETs (I have a philosophical aversion to switchers so I would too) your resistor selection is inadequate and may fail suddenly. It may also run long enough to give you endless headaches when it mysteriously fails. You have what amounts to a 11-12ohm1W resistor passing ~300mA which is almost exactly 1W leaving you with basically no safety factor. EEK! This is poor design before you have even considered the turn on currents of the tubes. If you do add this filter or something similar please find a resistor rated for at least 3 times the average dissipation. 5W resistors aren't hard to find, just do that - if a little safety factor is good imagine how good more would be.
 
4: Just wire it point to point. Nobody listens to this. 
 
5: Tubes on the top, boards on the bottom is bad. Nobody listens to this either.

 

[KimLaroux]

Thanks for all the input guys. I tried to pin-point the cause of the noise in my build today, and the results are interesting, to say the least.
 
First of, I just realized that the AC hum I get is not from the amplifier's PSU, but from the audio-gd NFB-12 under it! I have both of these stacked on top of each others:
 
 

 
 
When the NFB-12 is ON, the SSMH picks up AC hum, when I turn the NFB-12 OFF, the hum disappears. I really did not expect this, as the NFB-12 has a toroid transformer inside it's enclosure. There's at least 4mm of aluminum between the NFB's transformer and the wiring inside the SSMH.
 
I placed a metal can over SSMH the transformer, and grounded it. There is no diminution of noise with this added shielding around the transformer. This tells me the transformer is not to blame here. I do have a shielded transformer already. At first, here was some weird coupling between the transformer, the heat sinks and the tubes, but this got sorted with the Faraday cages I built around the tubes. This test confirmed they are doing their job well.
 
I even placed the amplifier away on the wooden bookcase behind in the picture, and connected it directly to the AC sockets in the wall. It didn't change anything for the white background noise. It did, however, get rid of the AC hum and lowered the WiFi interference. It is in my plans to get/build some sort of metal shelf I can place my laptop on, and place the NFB-12 and SSMH side by side under it. I'm expecting this to have a significant effect on the noise, and make the tube amp more functional.
 
So the white noise is from inside the case. I'll have to open it and add some aluminum plates between the AC and everything else. I'm just concerned that it might be the PSU itself that isn't clean enough. But since you guys are saying even an external SMPS will be silent... it's more likely that the noise is induced between AC and DC and signal wires. But it's not like my build was dirty:
 
 

 
The signal wires between the RCA, the volume pot and the terminal strip are shielded. The wires to the output jack are braided. The two small red wires with the brown resistors are for the tube LEDs. everything else is cut as short as possible. The RCA are shielded with an aluminum plate. The AC wires are tightly twisted together. I have an EMI filter on the AC line. The next thing to do will be to place sheet metal all around the AC wires, and ground it.

 

[the_equalizer]

Your build is a thing of beauty, really.
 
Some suggestions that might help:
 
1.- Try a couple grid stopper resistors on the tubes, R14 and R15 in the schematic, soldered DIRECTLY to the tube socket pin. Sometimes noise in tubes comes from oscillation, grid stoppers damp the oscillation and stop it. One user in the SSMH thread reported much improved noise levels when using the grid stoppers.  If you do have those resistors but not soldered directly to the tube pin, reconnect them so they are directly soldered to the pin. It does makes a difference. R3 and R9 are analogous and so the same goes for them, they should be soldered directly to the MOSFETs pin.
 
2.- Substitute the plate load resistors for constant current sources. I have just posted some details of this mod in the regular SSMH thread.
 
 
3.- Finally, and I'd try this as the icing on the cake, after implementing the other solutions, try different tubes. A different brand of 12A_7 might be quieter. Or maybe the 19W8.
 
Good luck with your noise zapping!
 
cheers!

 

[Goobley]

I agree that KimLaroux has a lovely looking bit of design there. I understand a lot of my errors now - then again this is the first amp I've built and I didn't want a PCB, the reason being so that I'd learn more. I was, however convinced by my Dad to use the stripboard in that layout which obviously wasn't the best of ideas.
 
I'm going to pretty much start again as I bought double most components, and I'm going to make two small boxes. In one I'm going to build a linear PS and then in the second the ssmh in point to point.
 
It would appear to me I probably also want some shielded cable for the audio paths, although is this very important if I put the PS in a seperate box? (Obviously I'll speperate the power from the sound in the amp box).
 
Given your comments the_equalizer's comments on the ccs, I'm certainly going to keep an eye on that mod.
 
I want to continue using wooden boxes, however I'm going to like them with either heavy duty foil or light steel. I'll look at stacking them and whether it creates too much noise.
 
With regards to the linear PS, and this is my most important question for the moment: I have an old guitar amplifier which is not bad quality, and I decided to open it to look at the coil. It's 78VA with 2*20.5V secondary windings, could I get this high enough to get a linear regulator in if I used a voltage multiplier?
I've never used one of these before, having only ever worked with 5V logic, but looking at the design of the CTH, the B+ goes to around 100V with the use of a 3x Voltage multiplier. Being 78VA it sounds to me like this transformer may well be powerful enough for my needs, which'd be great, as, when I was discussing PSs with KimLaroux in the main thread, we identified a few suitable transformers with were around 30$ before any other components.
 
I should hopefully be grabbing a few bits like a HV reg towards the end of the week and I'm looking forvards to stating again - and hopefully getting it right this time.
 
Thanks for all your help and support so far, I've learnt an awful lot on this journey
Chris
 

 

[the_equalizer]

Good that you've learned a lot with this, that's, at least for me, the best part of DIY audio. Aside from the points mentioned by NikonGod in his post here are some observations I have about your build and ideas.
 
1.- Working with the perfboard is most definitely doable, and has been done before. You just need a layout that avoids all those criss crossing of lines. Since the perfboard can only work as support for the components and not as a ground plane or shield, I'd suggest you use a metal enclosure.
 
2.- If you want to build in wood, that's also fine (has been done before too), but I'd then suggest you do point-to-point wiring over a groundplane (or cover the enclosure in heavy duty foil, as you mentioned). Doing point-to-point is really very easy and it works nicely. For layout just look at Mr. Millett's original "Starving Student".  I built mine based completely on his layout, point-to-point over a new PCB board as ground plane in an ABS plastic enclosure, no shielded cable. Very occasional noise problems (mostly with nearby cell-phones). Of course, shielded cable can only improve things but if your layout is tight and you use short lead runs it can be more a nuissance than a benefit.
 
You can take a look at the SSMH build gallery posted by Mr. Millett for layout/build ideas.
 
3.- Maybe you can build the amp in a metal enclosure, for shielding, and then cover that with wood for looks. Look in the regular SSMH thread and you'll find beautiful builds that were done this way.
 
But more than anything my best suggestion is for you to get the amp working nice and quiet with the originally specified Cisco 48V PSU. Once you have it working smoothly this way, experimenting with your own linear PSU design built in a separate enclosure will be trivial: simply unplug the Cisco PSU and plug in your own.
 
I'd suggest the same for the CCS mod (which I think REALLY improves the sound of the amp). Get the amp working smoothly as originally specified and from there, experiment with your own CCS and PSU. After all, if it works GREAT with plate load resistors and a switching PSU, it can only get better with a proper linear PSU and CCS loads!
 
I think going at it this way will make matters so much simpler for you as you'd have a clear area or component to debug, as opposed to now where there are lots of factors contributing to the noise problem.
 
Good luck and keep posting your progress, experiments, etc I think your and KimLaroux are adding a new dimension to this amp build.

 

[Goobley]

I'm back again, with a metamorphosed amp. Yes some of the cables are longer than they should be however, they are mostly and have all been braided (the ones that aren't braided are not for testing reasons.)
 
Well I have some good news and some bad news, however I hope someone may have answers for some of it.
 

As you can see it is entirely point to point over a ground plane (zinc and foil). All cables have been kept as short as possible.
My main issue at the moment is a high pitched noise (around 8kHz in the left channel). This is the side with the two coupling capacitors, I don't know if they could be considered the issue, but no signal flows directly next to them. I still have a fair amount of other background noise (even when in grounded metal box) which may (probably) be due to the smps as it's about the same in both ears. I could live with it if I managed to clear up the uneven and very wearing noise in the left channel.
 
Do you think that the capacitors are likely to be a the left channel problem? And if yes, how should I go about resolving the issue?
Thanks for all the help I've had so far,
Chris