SVT-CL misbehaving

[Prairie Dawg]

Can anyone give me a succinct summary of how the fault detection system works in these things? I've got one that's giving me fits and I'm getting really tired of schlepping it around.

Here's what I've got so far.

The original complaint was that it'd sometimes throw the amp off line right away, and sometimes after a couple hours. I'd replaced the screen resistors (220 ohms) a while ago and that seemed to cure the problem for a while.

So when I got it here like a dummy I didn't power it up to watch it and see what it did. I know, I know, I broke the first law of diagnosing problems but there it is.

I opened it up and checked the tubes on my matcher and found a 11 ma split lowest to highest. So there was that.

I did two fault detection desensitizing mods, putting clamping diodes across R8 and a .1/50v cap across one of the ICs.

Went to power it up and it wouldn't come out of standby at all-solid red light.

So I figure "I screwed something up" Went and rechecked all my wiring connectors and they're correct according to the assembly diagram but no joy.

I removed it again and demodded it but haven't tested it yet so I have a few ideas for tomorrow but it acts like either the relay's not getting power or it's not getting signal. I did read in the owner's manual that if the light stays red there's no high voltage.

I guess what I need is a succinct yet precise description of the fault detection system and how it works, and from that, what are likely suspects.

Be easy on me.

As an aside someone could do very well for themselves by writing "the Unofficial Guide to the Care and Feeding of SVT Amps." They'd probably have a good chance at canonization.

[Enzo]

First, MAKE SURE the small tubes in the power amp are in the correct sockets. Look at the V numbers in the schematic, and the socket numbers in the layout. There are 12AX7s and 12AU7s. I see them mixed up all the time. And sometimes I see someone has replaced an AU7 with an AX7.

The most common failure I see in those amps is bad 12Au7 driver tubes.

A wrong or defective 12Au7 can make the amp about impossible to get the bias right, and it can interfere with the protection.

An 11ma range in power tubes is no big deal.

AS far as I know without looking it up, those mods are about preventing the amp from shutting off speakers with large signals. Not related to idle setup.

Don't misinterpret the no high voltage red light thing. The relay turns on the high voltage. If teh relay is off due to protect, then that kills the high voltage. Not the reverse. A loss of high voltage won't kill the relay.

Look on the power amp schematic, not the power tube board. There is that 16 pin connector from the power tubes though. Pins 1,3,5 are the + side cathodes, and pins 7,9,11 are the - side cathodes. Now look at the lower right quarter of the power amp drawing, all those op amps and LEDs and stuff. Each cathode return goes to ground through a 10 ohm resistor. R35-40. Just like the proverbial 1 ohm resistor in any tube amp, these 10 ohm jobs are just there to sample tube current. The more current, the higher the voltage across the 10 ohms.

That goes two places. First, each of those is sampled through a 2.2k resistor over to U2b. If that voltage is high enough, it toggles U2b, which turns on protect and drops the relay, and turns the fault lite on in the preamp.

Note also below U2b, a positive short to the filament lines FIL1, FIL2, will also trigger the thing.

Those 10 ohm samples also trail off to the right to the array of op amps for bias adjusting. 6 pairs of op amps, one pair per tube, two groups of three pairs, one for each polarity of the output tubes. Pick one group, say the upper one. Note at the bottom of the group are three resistors, R71,73,76, that form a voltage divider off the +15v rail. Those are reference voltages. R76 drops the bulk of the voltage, then the two 150 ohm ones divide it further. The voltage at R71 will be half that at R73. The R73 voltage is the reference for the right column of three op amps, and then R71 provides the reference for the left column. The 10 ohm sample voltage is fed to both columns. If the voltage exceeds one reference, it turns on the RED LED, and if it is less than the second reference, it turns off the green LED.

SO the bias current, adn thus teh 10 ohm voltages must fall within the range set by those references to get the green light without the red.

[Prairie Dawg]

Thanks for taking the time to share the knowledge, Enzo. I do appreciate it, and if you're ever in Des Moines I'm buying the coffee. If you write the definitive manual on the care and feeding of SVTs I'll buy the first copy.

[Prairie Dawg]

After some study of the matter I decided to replace the TL072 op amp and the RTD14005 relay that sit between the cathode reference and the J12 lead out to the power supply board. I checked and rechecked all my connections and verified that all my connections were properly done. I then bypassed the relay, set the bias, let it cook for half an hour or so and shut down.

Alles in ordnung? Not yet.

I just finished reverting it to the normal configuration and it's the same as before. It won't come out of standby. even for a split second.

Ideas?

[Enzo]

If you didn't follow my attemot at explaining the circuit, please say so and ask what yo will. If you did, then go through the protect circuit. If the relay won't budge, look at what controls it, then what controls that, and then what controls the thing that controls the that that controls the relay.

If a transistor controls teh relay, then something feeds its base, and that something is either off or on. Then what makes that something do its thing, etc etc.

[Prairie Dawg]

You did, you explained it quite well and I now have a pretty good idea how it works. Although they share a common source of information (voltage drop across 6 each 10 0hm resistors) the fault detection takes off in another direction, through an op amp, and over to the relay panel. I bypassed the relay (found instructions on how to do it) and was able to bring the amp on line, set the bias with the LEDs, and play my guitar through it for an hour or so. Using my cathode bias probes I was getting about 25 ma on one trio and 22 ma on the other trio of power tubes. That told me two things-that the tubes were not that far out to lunch, and that the bias set part of the system was working OK. So all that eliminated a lot of problems, like, did I get the wiring right? I did, because I rechecked every connection against the print.

So the game tomorrow is to do as you suggest which is to go through the fault sensing circuit up on the smallish control board as I already know it works down below on the power amp board.. It seems relatively uncomplicated, but the biggest part of that will be to find test points and assign values.

[km6xz]

Enzo was being polite with his suggestion but it was apparent what he meant...if you ask for information that is factual and concise in how to resolve your problem, and get it, yet ignore it, why waste his time?
The circuit is not complicated but has a number of variables, proper diagnostics would have pinpointed the problem very quickly with a 1/2 dozen test points yet you ignored it and started modding and bypassing protection devices.
My advice, which you will likely ignore also is to stop any more shotgunning and make a logical diagnosis of the protection comparators to figure out the real cause. The hardest part of that is putting it on the bench.
Are you charging money for this process? Were the parts you changed actually faulty and if so how did you determine that? The ethical conclusion would be to charge only for the time that it would have taken if approached properly, and none of the shotgunned parts. The amp is too valuable to be hacking up for no good reason.

[Prairie Dawg]

Sorry I'm not a rocket scientist, just trying to learn from people who have more experience than I. Maybe what you missed was that I haven't been in this line of work as long as some people and there's still a lot I have to learn-which is why I ask, because the only dumb question is the one you don't ask, right?

to tell you the truth, I feel a little silly talking about all this, but I do feel I have to justify my actions. You gotta learn to crawl before you walk before you run, and we all put our pants on one leg at a time.

What you're characterizing as shotgunning and hacking was no more than replacing the two parts in the system that I have no economic way of evaluating. I mean, I could probably dream up an elaborate way of testing a TL072 but at fifty cents and five minutes to change, that's not time well spent. Likewise the relay. They cost $2.31 each. Likewise temporarily bypassing it.

Now. You have the complete rationale as to why I've done what I've done so far. Step forward and show me the way.

What half dozen test points are you talking about? If you know all this teach me.

[Enzo]

Oh, let's all be friends...

So we have a new IC and relay, nothing changes so they must be OK. Your bias seems to work, so we can assume everything to the right of R47 is probably good. COnceivably an open R46, but I ain't betting on it.

The print says IC2b-7 has +14v normal or -14v fault. Start there. If it is stuck in fault mode there, then certainly the relay and associteed stuff on the AC board will follow. If normal - +14 - then what is at the base of the relay transistor? Or at 0least R3 there at its base?

This works indirectly. Normall on the power amp board, R43,44 send +15v over to the relay board to the base of Q1. This holds Q1 on all the time UNLESS turned off by the IC. Fault voltage at the IC pin 7 pulls down that +15v through D2.

+14v at the IC will turn on that Q1 thus energizing the relay, or more accurately allow it to turn on. The left end of R3 sees the +15v. R3 and R4 form a voltage divider against the -3v rail, so right at the base of Q1 I'd expect the +15v to be more like +1.5v. Got good base voltage but no relay turn on? Sounds like bad Q1. You can short E-C on Q1 and that should energize the relay. And for sure check R2 10 ohms for opens. This whole relay circuit is basic. So if a current path is available to the relay, then the transistor is either not doing its job, or it is OK and not being turned on.

If the fault lite is lit on the preamp board, my vote is that the relay board with relay is probably OK.

[km6xz]

I am sorry that you are taking this as a personal attack, it is not meant to be. A hobbyist working on his own gear is held at a much lower standard of competence than a commercial service provider.
Yes, I have read your rationale but it does not change the fact that you are swapping out parts for no reason, hoping that something will start working without identifying the mechanism of fault. That is shotgunning and is unfair to the the customer besides not being productive. Actually you are further away from affecting a repair by introducing new variables.
The IC you replaced has published specs. External conditions will produce predictable reactions that are easy to observe. That is what was meant by having a number of easy to verify test points.
I don't have any problem with people getting into their own gear or those of friends as long as there an understanding about implied professionalism or lack of it. Charging for skills sets not in evidence raises my hackles. That is the only reason for my less than patient response. The fact is you have lowered the value of the unit by adding mods and removing them, both actions, for no reason, plus removed parts and replaced them with unknown quantities, for no reason, plus you are going to charge for this work.

The statement that you had no economic way of evaluating the parts is strange, besides a scope or DVM, what would you need to determine if the fault circuit is locked up or is doing exactly what it was designed to do due to a circuit condition it monitors? I have no idea of how long you have been "in this line of work" but I would venture a guess that it was long enough to read every book on electronics theory in the library which would have prepared you to analyze the circuit in as much detail as needed to resolve the problem. Enzo has almost repaired it for you, yet you are still trying to justify shotgunning, and ignored his detailed explanation. You still have not determined if the fault detection circuit is working or not, even though all the sense and control levels are within easy reach and a scope probe reading away.
Please resist taking in units that you do not fully understand, it does not take experience to understand a unit, but it does take having a firm grasp on the fundamentals. Most professions require proof of proficiency, although this one doesn't but there is no less need for it before handling valuable property of others.

[Prairie Dawg]

Enzo, thanks again for making the complex simple.

@km6xz, No, I'm not taking it personally. You've got your opinion and I've got mine.

[Prairie Dawg]

Just as an update, thanks to our illustrious crew chief Enzo for pointing me in the right direction with a cogent and thorough explanation of how this system works and how it is tested, this one can go in the fixed column. Explaining how things work together is a gift that not many people have and even fewer are willing to share.

Next time I'm in Lansing I'll buy lunch for the entire crew.

[hodgy]

So Prairie Dawg, what was the final fix?

[Prairie Dawg]

There was a whisker of solder bridging two resistors on the back side of the relay board. It was only detected with a magnifying glass

[azatplayer]

More SVT issues for an old thread!

Trying to find any info, trawling thru the gold on this and other great forums.

I too have an svt cl that wont engage the relay.
I have the required +14V at IC2b. That passes thru the 4k7 resistor R51 to the junction of D2/D5 and also along to the preamp box.
As theres no 345V hitting the preamp, that 14V+ gets turned back around to keep the lamp lit solid red.

So back at the diodes D2/D5. D2 is reverse biased in the direction we want that 14V+ to flow, so it shall not pass. Above this is another 15V supply on top of R43.
Thats 68K bypassed with another diode in this amp(2007) then on to R44, 4k7, then to J12.
ALl i get at j12 is 500mv. approx. Pulling j12s cable, it creeps up to 750mv.

Ive bridged the transistor at the relay and the relay clicks on, so that circuit is fine, its simply not getting the 14V+ hitting it.

Q1 gets a negative voltage at its base of -2.7V, has +3V at the collector and -3V at the emitter, all looking like it should apart from that base V.

Im trying to understand how that +14V is sposed to get past the reversed diode D2. WHat am i missing here?

I have no tubes in this while trying to find this issue out. As far as i can see, the opamp ic2b is only going to fault with tube drawing too much, so not tubes should have no effect.

I have tested the opamp, fine, and have tested Q1 out of the amp and its fine. Tested all the diodes on the main board, no bad 10r resistors or burnt screens, had both of those before a few times so were the first things to look for.
Also had bad au7's, always the first thing to try before breaking out a screw driver!

Regardless of the repair, id love to understand how that +14V gets across to the base of Q1..

Any help would be too good