New Express with VVR....and issues!

[ampgeek]

Iron and chassis from rj, caps from Allyn, VVR from Hall Amplification, PECs and cap cans from Antiquie ES, small componants from Mouser (per BOM for the most part) and...some fine pre-amp tubes from Blinkie. I also have a pair of NOS Siemens from Sanborn Electronics (ebay) but haven't plugged them in yet.

All in all....a very smooth build following the outstanding reference documents (thanks to Mr Worley!) and sage advice I have picked up here over the years. A sincere thank you to all!

But.....I have spent a few frustrating days trying to track down and fix an ever present distortion. Gut pictures and voltage chart attached. I was pretty meticulous with the lead dress and, generally, maintained high fidelity to the reference document's guidance and Francesca pics.

The amp is very stable and won't oscillate even with all of the knobs nearly dimed and no signal at the input. But...whenever I send it a signal I get a nasty/buzzy distortion on top of everything. Even when firing into a dead load I can hear it. Most prominently from the output transformer (13/16" wrench closed end to the ear and open end on OT body. Great way to hear abnormalities in bearings as well!). I don't even need to connect to a speaker to quickly troubleshoot/test the work that I am doing!

The only measurable anomaly with the equipment that I have is that V5 draws the same current (~180 ma) at all volume settings. V4 varies from ~60 ma to ~100 ma with a well struck A power chord. This difference follows the socket and is pretty consistent with the handful of well matched EL34 pairs that I have. So....I am convinced that it isn't a tube problem. The bias voltages changes the same (-33'ish to -55'ish) at both control grid pins when under load so it doesn't appear that one tube is being "told" to do anything different than the other.

Likewise, the screen grid current draw is the same between both sockets so it doesn't appear that this area is contributing to an imbalance in the overall current draw. I also lifted the diode plate-to-ground "bridge" at each socket to no avail.

I have checked parts value vs schematic (all resistors/pots checked before installation), poked, prodded, re-flowed, tightened connections, checked grounds, re-tubed (pre and power amp), replaced components, by-passed the VVR, yaddaa...yaddaa..ad nauseum and nothing at all has changed the character of the problem the slightest bit.

All that I can think of that is left to do is replace the power tube sockets and trying a different OT. I did measure the resistance of each side of the OT primary to the center tap and found that they are within an ohm or two of each other and determined that the offensive characteristics occurs at all three secondary windings.

Needless to say....I am at a real loss here. No amp has ever kicked my sorry ass like this!!

My amp stethescope won't help isolate it as I hear the distortion everywhere in the amp. Vestiges of it at V1 plate and worsening up to the OT center tap. Maybe some gets filtered out in the PS caps?

Any thoughts and/or guidance would be greatly appreciated.

The "singing" OT observation really twists my logic around as I can't clearly argue to myself whether it could be a cause or an effect.

TIA
Dave O

[Cliff Schecht]

It may be a bad or damaged output transformer.

Was the OT ever dropped by you or perhaps during shipping?

Have you accidentally run the amp with no load attached for any period of time?

How much tap testing have you done around the area of interest?

Have you tried the amp with just power tubes installed?

Does it still whine with the power tubes uninstalled?

Are your primary leads wired correctly? Secondary?

Sounds like V5 could definitely be your problem area as well (very likely since your iron is new).

Have you checked each socket pin for shorts both visually and with an Ohm meter?

Tried tugging each lead to make sure they are secure?

I can't see in detail but perhaps the socket is miswired? It looks alright from the pictures but I can't tell 100% if there is something that I'm maybe not seeing.. Hopefully somebody else can take a stab at this one and see something that I'm missing..

[ampgeek]

Thanks Cliff!
No physical damage to the OT and never operated without a load.

I should also mention that the character of the buzzy distortion occurs identically at each secondary tap.

Oh yes!! Lots and lots of poking, pulling and proding everywhere to try to induce a change that would give a clue. I also went back and carefully inspected and re-flowed all connections in the power tube sockets and PI sockets and componant connections (and.. systematically everywhere else as well). I stopped just short of replacing PI plate resistors until hearing suggestions here.

The gross imbalance in power tube draw has me perplexed. My guess is that they would never be the same given the likely differences in the waveform from each PI half. But...I am not skilled enough to even attempt to calculate/predict how different they might be normally.

Haven't tried with/without power or pre-amp tubes as you suggested. I will give that a go.

Thanks again,
Dave O.

[marcoloco961]

The only thing that stuck out right away is the wire running to the screens. It looks like 20, when I believe it is more common to use an 18.

I don't know much at all about the VVR, but I seem to remember reading about an adjustment to align the double potentiometer on the VVR units meant for an adjustable bias. I think there are some small pots on the actual VVR board. I don't know if these not being set would cause a symptom like this. Maybe someone who has installed a VVR could comment. Maybe double check instructions for VVR?? Just guessing here. Everything looked good from what could be seen.

[ampgeek]

Thanks Marco!

You are correct. I did use 20 ga on the screen wiring. I am getting another OT from rj and will change that during the test out process.

I would be interested in hearing about the aligning of the VVR pots. The VVR B+ and bias voltage appear to track each other very nicely but I didn't stop to calculate an actual variance along the way. I did set the VVR bias pot to max so that all power tube biasing can be done from the normal pot.

I didn't install the blocking caps on V1 a&b yet. Since I won't be using this at bedroom volume levels, I wanted to see how scratchy things got in the voltage ranges I will be operating at.

Alot of my troubleshooting has been done with the VVR bypassed (both B+ and bias) so it doesn't appear to be a major contributor to the issue.

Thanks again,
Dave O.

[rooster]

The OT is not the problem - count on it, and very nice of rj to offer support like that. What I think you forgot to do was solder the wire going to the junction of the 220K/.022 /bias lead wire that runs to V5. Do you see this?

If its not this, then replace that cap. Let us know.

[Cliff Schecht]

Looks soldered to me, or at least he hit the wire itself with an iron . I would check that joint though and the actual bias wire connection on the right side of the board (where the white wire goes to the turret and meets the under-board connection).

[paulster]

I agree that the grid wire to V5 doesn't look as solid as it might be, but it's really difficult to see with only one angle.

First thing I'd do is pop out the power tubes and measure the voltage on the grid to make sure you've got a stable bias.

I know it's there on your voltage chart but it's good to go back to basics and make sure it isn't intermittent for example.

The problem could well be a leaking coupling cap affecting your bias assuming there are no component value errors and everything is properly connected.

You could get an idea of whether this is the case fairly easily by temporarily disconnecting your negative feedback and then swapping the grid wires either at the power tube sockets or at the board.

Assuming the problem moves to the other tube then your problem lies on the V5 side of the phase inverter somewhere, but on the off-chance that it stays on the same tube then you've got a socket wiring issue or a possible OT problem.

I'd do that first as it's a pretty easy test to do and will give you a direction to focus your fault-finding in.

[ampgeek]

Thanks guys! I really appreciate hearing about ANYTHING that you think looks funky.

That joint is good as it is one of the ones that I reflowed at least once as evidenced by the melted insulation so I am confident it isn't the issue. But...I will hit it (and the underboard wire connection turret) again just to make sure!

I have also monitored bias voltage level at each power tube socket VERY closely while testing hoping to see something unusual that would lead me down the right path.

Could a leaky C13 coupling cap show up as a wavering bias voltage? I wasn't sure about that given the bias supply feed downstream of it.

I have even replaced the PI tube socket (ceramic now) as I thought that I heard an improvement when pushing hard on the original with a chopstick while banging out some power chords. Wishful thinking!!

I agree that it is HIGHLY unlikely that the OT is the problem but given the observations it must be evaluated.

So...the high V5 current draw at all volume settings isn't normal? I keep coming back to that (and the "singing" OT) as the only obvious things that indicate a problem.

I will disconnect NFB and swap the grid wires this evening and see what happens.

Thanks as always!
Dave O.

[sst4270]

If you were applying the VVR to the whole amp then you are missing the filter cap/1 meg resistors prior to Pin 7 of V1 and another place (I can't remember now off the top of my head).

[sst4270]

Where's the bias pot? I couldn't tell from the photo what that item is on the bias board.

Also am I missing the 1 Ohm resistors between your bias test points and ground or are they just out of view? Or is that them at the sockets?

Oh and I just remembered... I think the second filter set for the VVR goes between V1 Pin 2 and the Volume pot ( I think ?).

[ampgeek]

Thanks SST!

Yep. VVR is applied to whole amp but I left off the blocking cap/1 meg on V1 A&B until I got a feel for if they were really needed. That network is intended to keep DC voltage off of the guitar volume pots when B+ voltage is lowered below some threshold. Since I don't expect to ever play this amp at bedroom volume levels, I may not actually need it when all is said and done. I did leave provisions to easily install that network at each position if needed.

The 1 Ohm bias check resistors are mounted at the sockets and the leads are run over to the test points. Given all of the concerns over lead dress, layout and grounding points, I didn't like the idea of running 50'ish mA wires half way across the chassis.

The bias pot is in the usual place on the PS board. I procured the part suggested in the BOM. It is a little bitty thing but works just fine. I didn't like the idea of soldering it's little leads into a turret so I liberated some pins from a relay socket, soldered them into the turrets and now have a "push fit" bias pot!

I use that trick a lot in the pedals that I make. Makes swapping out componants a breeze to fine tune tone/response of pedal kits that use PCBs.

Thanks again,
Dave O.

[sst4270]

Some other things I've noticed.
It's hard to tell from the photo, but are the metal film resistors you're using for R7 & R8 the correct values? It looks like R8 has a green first stripe.

Also the junction of R16 & C13 doesn't look soldered.
Or rather the blue wire doesn't look soldered to the turret.

Just my observations. It's hard to tell without multiple angles.

[UR12]

Do you have an Oscilloscope? It kind of sounds to me like you may have an oscillation that is out of the range of hearing but may be causing the tube to draw a lot of current even with no input. You could see something like that on a scope.

[ampgeek]

Not sure what is up with the extra band on those metal film resistors. But...they measure spot on 150K and 10K as they should.

The V5 grid wire is and always has been well soldered.

Hoping to get access to (and...a lesson!!) an Oscope this weekend.

I disconnected the NFB from the selector switch and flipped the power tube grid wires. The high current condition switched to V4 so it looks like a PI issue. The nasty distortion and "singing" OT remained the same.

I popped out the power tubes, flipped the grid wires back, connected the NFB and fired up the trusty amp stethoscope. I "walked" around the circuit listening carefully to the quality and level of the signal. Sounds awesome up to the PI top half grid! Gets really nasty after that. Specifically the bottom PI half grid, PI cathode and both PI plates. Pretty much identical for the 3 known good PI tubes I tried. So...I am convinced that it is not a tube problem.

The V5/pin 5 signal is ugly but much stronger than V4/pin 5. Might that explain the high current draw condition?

What component(s) could be guilty? Does the voltage chart data point to something fishy with the PI bias condition?

Thanks as always!
Dave O.