SVT Dies on the Bench

[clyde1]

Damn I hate these things. Actually it's on the floor because I can't lift it onto the bench. I tested all tubes, matched them, tested everything on the output board, cleaned all necessary contacts and put it back together after replacing a faulty Stdby. switch. The customer complained of poor tone but it seemed to sound fine after I biased it up (3 ma difference between sides). It just quit and wiggling the AC cord in the socket seemed to restart it temporarily-Aha, bad socket! Well, not so because I test 122v everywhere, bypassed the AC switch, no dice. I suspect it's the AC terminal board where the protection circuit is. Replace the board, or...?
Thanks.

[drewl]

Yeah, they're a pain, best to remove the parts from the shell and work on it.

So, is it an old original one?
The connector plug on the rear left side usually goes bad or has loose contacts from years of use.

[nevetslab]

SVT's do require a mental commitment to tackle. What you've found by wiggling the power cable and getting the amp to cut on and off is where the IEC-320 connector mounts to the AC Mains input PCB....you have solder fractures where the connector is soldered to the PCB traces that are then routed thru the relay on that board (assuming this is an SVT-CL,AV or VR model). Every one of these amps never leaves my bench without my pulling the Output connector PCB & AC Mains PCB out for inspection (as well as all the other PCB's). Solder fractures on the IEC connector is VERY COMMON. De-solder & re-solder is the order of the day on those, as is on all of the rear panel parts that support the main power amp board, as well as on the preamp board.

There's so much weight involved with these Amp's PT, OT & HT.....whose mounting hardware is perpetually loose. That also includes the 5 chassis mtg screws thru the case. Two of those mtg screws on the OT are underneath the main PCB, so there's an hour lost just getting to them. The SVT amp is, in one sense, a built-in shaker table while in transit. Allow the transformers to go unchecked, and the vibration transmitted by those shake apart all other PCB components being used to support the Main power amp PCB and other PCB's. Ampeg doesn't believe in supporting PCB's with hardware.....mistaken belief in the solder connections of panel components are good enough.

I've never had to replace any of the PCB's in these amps. I just have finished servicing 16 of the 19 SVT's in our rental inventory. Many had never made it into the shop in the 5 years I've been here, so those were automatically torn down completely to deal with all the solder fractures, loose or lost hardware, tightening up the Xfmr mtg screws, PCB standoffs (one of which is your main GROUND path to the chassis....if loose, it's hum city), replacing the stiff clear vinyl harness housings with Tech Flex, and even the large Xfmr's core bolts end up loose on the later ones.

Not a trivial amp to service.

[clyde1]

Yes it was indeed as you said, broken solder joints on the IEC socket/board connections. It's not quite biasing out all the hum with 2 LED's lit and I have a very distorted, mushy tone at any volume. As I said before, all tubes tested good, 3ma difference between the sides, output board all good. At a loss again.

[nevetslab]

When you say '2 LED's lit', do you mean both LED pairs are GRN, or GRN & RED? (SVT-CL or AV) Or is this an SVT-VR? If SVT-VR, then, as you know, you can see what the combined cathode currents of each group of power tubes can be measured across the 1 ohm cathode resistors (R40, R48) on the bottom of the main PCB. That won't tell you what the individual power tubes' current is, which is how I normally group the tubes....getting the best match out of the six. A bit more tedious and time consuming, but it does yield better long term results. And yes, with however they are grouped, there is a null that we hear in the residual hum.

On the mushy and distorted aspects, if this IS a SVT-VR, I also assume it's Ch 1 being referred to. On the preamp, V1 or V2 are the most likely suspects to that character, even if the tubes 'check out ok'. Not sure what your criteria is, but if you have some stash of workable 12AX7's to select from, I'd try changing V1 and/or V2 first. V4 is the output tube, a 12AU7, and one side is the mixing tube for Ch 1 & Ch 2, with the other half being the cathode follower that feeds the power amp. Ch 2 on the VR.....I've never cared for the sound of that channel, and haven't spent the time to revise it to yield more of the character we get from Ch 1, less it's Midrange EQ circuit. Ch 2 is always hard and lifeless to my ears.

If this is an SVT-CL, then V1 is usually the tube most responsible for distorted character.

You also can jack into the Power Amp directly with your bass. Not as much gain, but it will usually tell you sonically if that's where the mushy, distorted sound character is coming from. Thus far, I've only re-capped the power supply on one of the amps in our rental inventory, though I've always thought there's not enough capacitance in them for 300W.

Balancing the output tubes on the SVT-CL can be done by measuring the individual tubes' cathode resistors, with the power amp chassis standing up on it's side, as shown here:



You need to refer to the PCB layout of the power amp to show you where the 6 cathode resistors are. V1, V2, V3 & V5's resistors are along the edge of the board, with V4 & V6's resistors set back up a bit and rotated 90 deg. In these photos, I had tack-soldered in test points on the cathode resistor's solder pad so I could add a scope probe (for a different test setup photo). Then, it's a data recording, tube marking and juggling act while moving the tubes around and assessing the results until you get the best balance of the two sets of tubes, and tweaking the bias pots for the balance between the upper and lower halves. There's a lot more bias range 'in the GRN' than you have with a SVT-VR.

Doing this type of measurement on the SVT-VR, it has 5 ohm 5W plate resistors on the power tube PCB and that's a bit of a dangerous operation as you're probing at 660VDC. I've made a special insulated plate allowing access to the plate resistors and a few other test points. But, it can be done with two tubes at a time across the cathode resistors.

[clyde1]

Thank-you again, it's the AV model. I removed the tubes and put them in a Fender Chassis 2 at a time to measure bias using the shunt method after a considerable warm-up time. When all said and done I had 3ma difference between the triplets, left side vs. right side as I've heard it's not front-back configuration. Two green LED's lit, but a little more hum than I'd like. After I removed the AC terminal board and resoldered the cracked IEC joints, I put the amp back together (foolishly) thinking it would work. It doesn't start to distort immediately, taking a minute or so, making me think it's heat or voltage related. I subbed out the rear 12A_7 tubes one at a time and put Pro-Gold on the pins, no improvement. I'm just about to take it apart and sub the preamp tubes out and reinspect.

[nevetslab]

One mechanical PCB support stud is involved in overall hum....the output connector end of the main PCB...the rear mounting screw that threads into a standoff...that is the main ground connection of the power amp circuit to chassis. The standoff HAS TO BE totally tight at both ends, or you'll have hum. I've had some SVT's that didn't have the chassis hole for that standoff masked off prior to painting, and never found the hum source until further digging.

[clyde1]

I have it apart and the input jack is causing grief so out comes the preamp and we'll check the solder joints under there. I've seen bad jacks before-not grounding properly. As well when you flip the standby on initial power-up, there's a *huge* snap. Subsequent turn-ons (standby) don't elicit such an angry transient. I still hate these damn things but thanks very much for sharing your certainly hard-won knowledge.

Update: Neither input jack was shorting. This was easy to fix however, by inserting a u-shaped seal pick in the gap between the tabs and lifting on the pivot end while simultaneously pushing the contact end down with a small screwdriver. As well, -15db jack has cracked solder.

And did they have to use threadlock on the the preamp mounting screws? Anything that small doesn't stand up well to disassembly-rounded screw heads and one twisted off.

The gain, bass, treble and master pots have cracked joints-done and now we'll try a run without assembly, after checking that main ground. Still wondering about that loud transient on start-up. Maybe it'll be gone now-if only!

[nevetslab]

Often that input jack has a tension problem, no longer making contact with the Normal blade when empty. I have had some success in re-tensioning the upper leaf spring, though other times, it's a lost cause, requiring replacement jack. I normally remove the jack in trying to cure the mechanics. Cliff S4/BNB PC-A. Often no stock availble, and overpriced. I recently bought the solder terminal version of that, and cut the lugs to the shape of the PC terminals, as Newark had them available for about $2 ea, instead of $10 ea.

Cliff 1_4_INCH_S4_Jacks.pdf

On your initial scary sounding turn-on transient, does that happen if you dead-patch the Power Amp input jack (shorting plug)? I don't normally hear that transient on the SVT's, but often do on Fender Twin's. There's only film coupling caps between stages being charged up while all the tube circuits are being run up as the power supply caps are energized. Finding the source of that transient can be a challenge. Hopefully some other members have an effective trick in finding that.

[clyde1]

Transient still presents itself putting the shorting plug in Power Amp In. The distortion is improved somewhat, but increasing the volume and hitting the guitar strings with force elicits crackling. Onwards.

[clyde1]

Plugging the guitar straight into the power amp still gives a distorted tone. Only 1 red LED lights up when strumming or picking a note. Pushing down on the middle 12AU7 will make the the other red LED light up and the distortion goes away. Something loose somewhere yet I don't see any bad joints, which would by my first suspicion. Any brainwaves? Looks like I'm going to have to pull the main board for a better look. Shit.

[Jazz P Bass]

With the power supply discharged, you could try 'retensioning' the socket tabs.
And make sure the tube pins are nice & straight.

It may also help to measure the pins at the socket.
It is not uncommon to have a broken socket tab.
Without a tube you can measure the plate voltage.
Amp off & drained, you can verify that any pins that are connected to ground, through a resistor or direct connection have a solid connection.

[clyde1]

Thanks Jazz P, good advice. I did retension the tube sockets, although I've seen much looser that worked fine. I reworked some solder joints that looked rather suspicious, one in particular, the plate side of the coupling cap on one of the 12AU7's was definitely cracked, extremely hard to see but now the amp is fine. The amp(eg) gods have smiled on me. :-) This baby badly needed some work. Thanks for your help everyone.

[nevetslab]

Nice bit of diagnostics with the absence of the RED LED during signal drive, and pushing onto the middle 12AU7 to suddenly yield proper behavior. I am constantly amazed at how much time I spend with the surgical loupes & headlight scrutinizing solder joints, and where odd symptoms finally reveal themselves, as you and Jazz P Bass found. Great to hear all the effort has now yielded a solid working amp again. As you stated 'This baby badly needed some work"......it is so often true. I rarely find one that comes in for service, or even those sitting in their road cases ready to go out on rentals.....if they hadn't been serviced for a few years, there's just all sorts of ailments to find and resolve. Pleased to have helped out!

[audiopete]

I think next to blown tubes, cracked and/or cold solder joints are one of the most common issues I come across. I had a Doepfer MS-404 monosynth in recently that lost it's portamento and pitch control. Cracked solder joints between MIDI cpu and DAC chip that provides control voltage for the VCO/portamento (plus a couple spread contacts on the CPU PLCC44 socket).