Matchless Spitfire Build Troubleshooting

[silendt]

After a few years of modding VJrs (many thanks to this forum), I went ahead and embarked on my first scratch build. Always wanted a Spitfire, and it looked simple enough, so that's what I chose.

I bought a board from Watts, and used Ken's layout supplied here:
http://www.tubeamplifierparts.com/layouts/matchless_spitfire_layout.jpg

With the exception of using a choke instead of the resistor.

schematic here:
http://elektrotanya.com/PREVIEWS/63463243/23432455/egyeb/matchless_spitfire.pdf_1.png

Trannies are the actual OEM Spitfire models from TDS. It's worth noting that I had a stellar experience with TDS. Had a great hour long conversation with the owner, and his product came at an incredibly reasonable price. I would recommend them to anyone and everyone looking to do a Matchless build.

I've got it all together now, double and triple checked my connections. When I went to test it, it responded to playing by going from weak/dull with light playing, to harsh/splatty distortion when played hard. Increasing the gain control only made the harsh distortion worse. I started (safely) measuring voltages, and while troubleshooting I'm noticing that the (preamp) volume and tone controls seem to affect just about every measurable DC voltage, as well as the general bias of the amp itself.

With volume and tone at noon, the power tubes glow brightly, produces weaker sound, and DC voltages read at significantly lower levels. However, if I turn the tone control all the way down, the tubes cool off, overall volume increases significantly along with cleaner tone with smoother breakup, and DC voltages get much closer to normal.

The same thing happens to the voltages when I turn the volume control all the way down. Which seems to make sense as it sends the signal to the same audio ground point, the only difference being the .01uf cap at Tone lug 2 (of course, being that turning down the volume control mutes the signal, I can't speak to any tonal variation). What I can't figure out is why either of these controls would have this type of unintended effect, and what I can do to correct it.

As I've mentioned, I'm good at following directions, but I'm no electrical engineer or amp guru, so any help would be much appreciated.

I'd gladly provide voltages. However, given my problem, I'm not sure at which tonal configuration to provide those voltages  :???:

I've been slaving over this goofy project for months, so any input that could help get this thing up and running would be a big help, not to mention a huge morale boost, haha!

[sluckey]

If turning the volume or tone knobs is affecting voltage readings then I highly suspect a wiring error.

[HotBluePlates]

Are both speaker jack mono with no switch?

Did you use 100% new parts in this amp, or did you recycle some caps?

[silendt]

If turning the volume or tone knobs is affecting voltage readings then I highly suspect a wiring error.

Oh, I know, I'm almost certain it is. Is there any particular problem area I should take a closer look at? Based on the wiring/location of those controls, is there any likely culprit?

[Paul1453]

Ditto on the wiring error.

If the blue wire that goes to your pots on the layout diagram was not on the cap but on the 22K 2W resistor next to it that might do that.
Or that resistor is also tied to your cap somehow, solder blob/accidental metal or wire touching?

[silendt]

Are both speaker jack mono with no switch?

Did you use 100% new parts in this amp, or did you recycle some caps?

All three speaker jacks, as well as Input 1, are Switchcraft mono open-frame TS jacks; Input 2 is same, just TRS. I don't have the resistor matrix for the line-out built yet, but I assumed this wouldn't effect the signal coming from the 8ohm tap I have my speaker plugged into.

All parts are brand new. Board came pre-wired from Watts Tube Audio, as well as most of the hardware. Chassis from Mouser.

[Willabe]

Oh, I know, I'm almost certain it is. Is there any particular problem area I should take a closer look at?

Click on the link at the bottom of my reply for information on a great way to check your wiring on a new build.

[silendt]

Ditto on the wiring error.

If the blue wire that goes to your pots on the layout diagram was not on the cap but on the 22K 2W resistor next to it that might do that.
Or that resistor is also tied to your cap somehow, solder blob/accidental metal or wire touching?

Oh! You know what, there's a wire that comes pre wired on the underside of the board (indicated by a red dotted line in the layout) that goes right past the bottom of that turret. It's hidden from view, since the board is mounted in the chassis at the moment... But let me check, that could be it!

[Paul1453]

Board came pre-wired from Watts Tube Audio.

That doesn't mean that it is right, now does it?
If it is hand wired there could be a mistake.  We are human and I know I make lots of mistakes.   

[Willabe]

Trannies are the actual OEM Spitfire models from TDS.

TDS?   

[silendt]

Trannies are the actual OEM Spitfire models from TDS.

TDS?   

Transformer Design and Supply
https://goo.gl/maps/oigsHdpLZDL2

They're a father and son company who specialize in AC style amp trannies and have been the exclusive supplier to Matchless since the company's inception. A much debated subject, it seems, but they have letters and invoices to prove it. Apparently Rick Perrotta is a long time friend and sought them out when they first started making amps. I think the dad recently passed away though, IIRC.

[Willabe]

Thanks. 

[silendt]

Here's a pic of the tube wiring, fwiw. I can take a shot of the board too, if it would help.

edit: apparently it flipped the pic? Which I guess actually helps if you're comparing it to the layout...

[HotBluePlates]

If turning the volume or tone knobs is affecting voltage readings then I highly suspect a wiring error.

Oh, I know, I'm almost certain it is. Is there any particular problem area I should take a closer look at? Based on the wiring/location of those controls, is there any likely culprit?

Well, look at the Volume control wiring.

The wiper (middle lug) of the volume control is supposed to go to a coupling cap on the board (0.01uF). But that cap sits right next to a resistor (22kΩ 2w) which goes right to the power supply. So one turret over, you put high voltage on your pot and possibly get what you're describing.

[silendt]

I was able to flip the board over, and unfortunately it's a clean connection, no contact with anything else... Even confirmed it with a continuity test on my DMM.

Here's a pic of the board itself. Everything looks kosher when compared to the schematic.

If either of the preamp tubes are removed, the problem goes away, but there's of course no sound.

Also, if I have the tone knob at 0 and the volume knob at about 85-100%, there's a loud oscillation that sounds like it's physically present in the OT. As in, if I use a reactive load box, there's an actual buzzing/oscillating sound coming from the OT/preamp tube area itself (it also comes through the speaker when plugged in)... Kinda scary actually.

Also, just as an example, turning the tone knob from 0 to about 10% and beyond causes a 20V drop at that same spot in question (Point B on the schematic, where the two 22k resistors connect to the choke and 100R screen resistors). 353V down to 333V...

[Paul1453]

HBP comes to the same conclusion as I have.

Double check that jumper.  You can use your meter to test.

You should get 0 ohms where I made the Black line.

You should not get 0 ohms where I made the Red line.  If you do, there is your problem.

I extended the red line between the cap & resistor.  You should not get 0 ohms between the cap & resistor either.

[HotBluePlates]

Are all pot lugs at 0vdc?

If so, it sounds like it may just be oscillation due to long wires from the board to the tube sockets, and because your preamp tubes are right next to the output jacks.

Yeah, I know the layout diagram shows the jacks right next to the tubes. From my experience, I would know that the board & its components are the only thing being accurately shown to scale. The pots & sockets are shown in their positions on the chassis relative to the other pots & sockets, while the positions of everything else (jacks, switch, fuse) are just placed to make drawing easier. IOW, this "layout" isn't a "chassis picture" but a wiring diagram.

I'm thinking you'll need to move the sockets. Can you do that? Perhaps to the big open area between the board and the back apron (where the sockets & jack are now)? The images below are good examples of the relative positions of preamp tubes vs speaker jacks. The very bottom one is a Spitfire clone.

[sluckey]

Did you scrape the paint from the chassis at every point where a ground connection is made? This would include areas around screw holes for ground lugs, input jacks, etc.

[sluckey]

I just noticed that you used the wrong type input jack (at least the one on the right in your pic). You need Switchcraft type 12A for both the input jacks in order to work IAW the schematic.

Can you post some hi-rez pics (bigger than 640x480) showing the insides of the amp? Some Close-ups of areas and also a good overall pic that shows the entire chassis so we can get a sense of the overall layout.

I'm still suspicious of wiring errors and possible bad grounds due to that painted chassis.

[MFowler]

I used TDS custom wound transformers in an Marshall 18w tremolo build and they are very good.


Take the voltages with all pots turned to lowest position to keep noise down while on the bench if using a speaker not a resistive load.


Please take a high quality photo showing the entire chassis so we can view the complete amp.


I'm suspecting poor solder joints.


This appears to be a combo amp with tubes facing down.  I did not see the 100R screen resistors on the EL84's.


Mark

[Ugly Distortion]

silendt, did you get the iron and talk to TDS recently? I had feared they closed up like David West who sadly also died recently. TDS had a barely adequate web presence, a single home page and a slow moving eBay store, but that little bit has been down for several years. So, if someone calls them they are actually still there? Will sell/make one offs?

And yes, great transformers, essential for Matchless clones, but very likely darn good in anything you put them in. Someone might want to retail the TDS Matchless line, West had a stiff mark-up and was selling them around the planet I believe.

[Paul1453]

If turning the volume or tone knobs is affecting voltage readings then I highly suspect a wiring error.

This issue really needs to be addressed 1st IMO.
It doesn't make much sense to troubleshoot anything else with this symptom present.
You obviously have voltage going to these pots that shouldn't be.

Did you make those simple resistance checks I suggested?
You need to isolate and identify the mis-wiring or bad component that is sending your PS voltages to these pots.
It should not be too hard to find, we are trying to help.

Once this major issue is addressed, then you can move on to addressing any other issues if any still remain.   

[silendt]

If turning the volume or tone knobs is affecting voltage readings then I highly suspect a wiring error.

This issue really needs to be addressed 1st IMO.
It doesn't make much sense to troubleshoot anything else with this symptom present.
You obviously have voltage going to these pots that shouldn't be.

Did you make those simple resistance checks I suggested?
You need to isolate and identify the mis-wiring or bad component that is sending your PS voltages to these pots.
It should not be too hard to find, we are trying to help.

Once this major issue is addressed, then you can move on to addressing any other issues if any still remain.   

Sorry, I've been at work all day. I really appreciate all the help from everybody though!!

I did check those resistances last night. It all checks out, everything is as it should be in regards to that potential trouble spot. I've also done the tracing test with the layout several times.

Somebody over at Wattkins suggested something that actually might make sense. Due to my poor placement of certain components (outputs next to V1, OT primaries running alongside preamp section) I could be dealing with a parasitic oscillation problem. So if there's an oscillation happening at a frequency higher than I can hear, it could be loud enough that I'm slamming it into my power section, pushing it to the brink of failure (thus glowing tubes), and turning the tone knob all the way down rolls it off, thus relieving the power section; turning the volume knob would have the same effect. The question then becomes would an excess of signal have an effect on DC voltages?

[silendt]

Also, to clear up the layout questions: this whole project started as an experiment to see if I could build an amp into a 2U rack chassis. I have a Palmer PDI-03 and wanted to create a small, silent, self contained rig for practice and smaller gigs.
The biggest PITA of the layout was that tubes and transformers would have to be mounted into the back panel... Which it seems like is creating a world of problems. I'll post a link to a Dropbox with hi res pictures so you can get a feel for what I'm talking about.

[silendt]

https://www.dropbox.com/sh/i1t98mpbvhmgsvs/AACLTuHR8o1YPwIHb2PFglPna?dl=0

[Paul1453]

The question then becomes would an excess of signal have an effect on DC voltages?

I would not expect it to, but I'll let someone more knowledgeable address that.

I was playing around with a number of simple two tube circuits on my breadboard and had quite a few with whistling issues.
High frequency feedback, but none of those whistling circuits had the voltage fluctuation issues you have described.   

[Willabe]

If an amp is oscillating it can/will be working very hard, the harder an amp works the more current it will draw from the power supply (PSU) , as it draws more and more current after a certain point the PSU can't supply it any more and the voltage will drop or call it sag.   

[HotBluePlates]

Also, to clear up the layout questions: this whole project started as an experiment to see if I could build an amp into a 2U rack chassis. ...

That explains it all.

Yes, you can build a tube amp in a 2U rack chassis, but probably not with a typical power transformer like you're using. A toroid transformer is almost required to jam a powe amp into that small height. Additionally (or maybe in lieu of the extra space gained by using a toroid), you'd probably have to mount the tube inside the chassis and have vents above (and possibly below) the tubes. On top of that, you'd then want to leave 1U of rack space immediately above & below your chassis to facilitate ventilation (or otherwise provide forced air ventilation with a blower fan).

But doing all that starts negating the space savings you though you were getting.

While I'd definitely not want to, if it were my amp I'd get a new chassis with typical guitar amp dimensions. But I also might take a go at removing the off-board terminal strips, mounting the socket near the board on right-angle brackets (or brackets holding a G10 board to which the tube sockets are mounted) and shorten-up all the wiring. Rectifier tube, & output jacks could stay where they are. This would probably alleviate a lot of the oscillation issues. Still would like those vents above & below the new tube location though.

... The question then becomes would an excess of signal have an effect on DC voltages?

The power supply has its own internal impedance, due to PT winding resistance, rectifier voltage drop and series dropping resistors. So yes, if enough current is drawn from the power supply (beyond brief small signal which the filter caps can immediately supply), then the B+ voltages sag downward.

Usually oscillation (positive feedback) will grow to be as big/loud as it can possibly be within the limits of the power supply and whatever series impedance restricts its growth. This is true whether the oscillation is audible or super-sonic. But you mentioned hearing oscillation, so that combined with the 1st preamp tube being right next to output jack wiring means the layout is very likely the source of your problem.

[Paul1453]

If an amp is oscillating it can/will be working very hard, the harder an amp works the more current it will draw from the power supply (PSU) , as it draws more and more current after a certain point the PSU can't supply it any more and the voltage will drop or call it sag.

I knew you guys would help us all understand.   

On my circuits I definitely saw the current ramp up, but since it was nowhere near the limits of what my bench PS could provide, that explains why I did not see any voltage sag.

Current nearly doubled on all circuits when oscillation began.  It is easy to understand how that would overstress a normal PS and cause voltage to sag.

Thanks!

It is so much simpler to be just a repair tech than a circuit designer/engineer. 

[HotBluePlates]

On my circuits I definitely saw the current ramp up, but since it was nowhere near the limits of what my bench PS could provide, that explains why I did not see any voltage sag.

Your bench supply is also regulated; the output voltage is artificially held constant (within sensible limits) as load current increases.

You might say the regulation reduces the apparent impedance of the bench supply. This will generally be true unless the load current well-exceeds the bench supply's rated output, eventually exceeding the regulator pass-device's ability to supply additional current or the supply's power transformer to deliver raw voltage/current. Drooping line voltage may also impact the bench supply's ability to regulate, but well-designed regulators account for reasonable, foreseeable line voltage variation.

[silendt]

Argh... Definitely not the conclusion I was hoping for, haha! Just for the sake of my own bull-headed determination, not to mention my fancy light up front panel emblem, I may have a go at that daughter board idea for the tubes... I'll keep everyone posted with my findings, haha.

[MFowler]

You need to use shielded wire (coax) and the oscillation issue will be gone.


You ignored my post for some reason?


Where are your screen resistors?


Mark