Virgin Voyage - VTA ST-120 kit build

This is what your workbench will look like a year from now ...

I will admit my garage bench is a bit of a mess most of the time. My table saw fits under it when not in use.

The basement bench is a bit of a catch all but never abused.

(Long time since posting...)

The only additional advice I can add to the wise posts above is to recognize when you are tired and call to it a day.  You know all the reasons why.

As much as you will enjoy your beautiful creation you may be a little sad when the journey is over.

Finished it tonight.
Passed all start up tests.
All the magic smoke is still on the inside.
Dead Silent with no input signal.
Doing a bit of low volume listening; wife is asleep.

Once I feel this was worth the effort I'll have to look at the preamp and phono-preamp. I am not a fan of the case for the preamp. That cannot come in to my house looking like that.

Nice while it lasted. Just blew the slow blow fuse after about 1 hour of low volume listening.

To the manual. Hope I can figure out what I did wrong. I also have to figure out where I can get more of these fuses.

dalemurray wrote:Nice while it lasted. Just blew the slow blow fuse after about 1 hour of low volume listening.

To the manual. Hope I can figure out what I did wrong. I also have to figure out where I can get more of these fuses.

How did you blew it ? What was the sequence of events ?
A power off/on sequence could easily result in a broken fuse.

dalemurray wrote:Nice while it lasted. Just blew the slow blow fuse after about 1 hour of low volume listening.

To the manual. Hope I can figure out what I did wrong. I also have to figure out where I can get more of these fuses.

First some math:

You will need to calculate (roughly) how much current your amp *should* be using.

KT88 x 4 makes:

4 x 1.6A x 6.3 V = 41 watts (all results-figures rounded up) on the filaments
(35 + 6) x 4 = 164 watts dissipated in music and/or heat
Each of your small-signal tubes dissipates about 2 watts as above.
5AR4s dissipate about 200 watts at peak.

Makes a total of 251 watts or so. BUT, you ask, what about the 5AR4? - the current it dissipates is what is being used down-line.

Pull all the tubes, plug it in, turn it on. Read the current. That will give you transformer losses - but, generally, I look at about 10 watts per-each to be conservative. Mostly, and with Bob's transformers, losses will be under 10%, but let's use 30 watts. And as we are rounding, make the total to be consumed at 281 watts. Make your wallplate voltage at 120 VAC. Makes for 2.4 A THAT!! is the amount of actual current required by your amp at peak. I believe that the nameplate on these amps is 275 watts. So, close-enough.

Now, on to fuses:  There are two kinds of "slow-blow" fuses. One consists of a spiral element would around a thin ceramic rod. The point of this sort of fuse is to tolerate an overload for some period of time before failing. Such fuses are designed to protect real-estate, not equipment. Typically they will fail for a catastrophic short, before the fire actually starts. However: They will tolerate a 100% overload for very nearly 30 seconds. And a 20% overload very nearly indefinitely. These fuses are dangerous in the same general way that a Brown Recluse spider is dangerous. Look them up.

Then, there is the Dual Element fuse: A device that tolerates up to a 300% overload for a very brief period of time, then behaves as a normal fuse thereafter. Perfect for motor starts and cold-filament starts. By calculation you could use, safely, a 2.5A Dual-Element fuse. Writing for myself, and based on the fact that these fuses come in fractional values from 1/10 A through 10A, I would use a 2.75A. However, and for the record, I would let the amp sit on an ammeter for several hours of actual use before making the final decision. Which would be Average Actual Use + 10%. And it would be a dual-element device.

Long-winded way of getting around to:  What is the fuse value? What kind is it? And did it blow in use, or after a short-cycle?  And. Do you have an Ammeter and the means to test the actual current used by your amp?

peterh wrote:
How did you blew it ? What was the sequence of events ?
A power off/on sequence could easily result in a broken fuse.

Amp had been on for about an hour and playing at very low volume (low enough not to wake my wife at the end of the hall). The tone of the music changed and faded out over the course of a couple seconds.

I passed all the start up tests without issue.

dalemurray wrote:

peterh wrote:
How did you blew it ? What was the sequence of events ?
A power off/on sequence could easily result in a broken fuse.

Amp had been on for about an hour and playing at very low volume (low enough not to wake my wife at the end of the hall). The tone of the music changed and faded out over the course of a couple seconds.

I passed all the start up tests without issue.

Do you have the means to test tubes?

Peter W. wrote:

dalemurray wrote:Nice while it lasted. Just blew the slow blow fuse after about 1 hour of low volume listening.

To the manual. Hope I can figure out what I did wrong. I also have to figure out where I can get more of these fuses.

First some math:

You will need to calculate (roughly) how much current your amp *should* be using.

KT88 x 4 makes:

4 x 1.6A x 6.3 V = 41 watts (all results-figures rounded up) on the filaments
(35 + 6) x 4 = 164 watts dissipated in music and/or heat
Each of your small-signal tubes dissipates about 2 watts as above.
5AR4s dissipate about 200 watts at peak.

Makes a total of 251 watts or so. BUT, you ask, what about the 5AR4? - the current it dissipates is what is being used down-line.

Pull all the tubes, plug it in, turn it on. Read the current. That will give you transformer losses - but, generally, I look at about 10 watts per-each to be conservative. Mostly, and with Bob's transformers, losses will be under 10%, but let's use 30 watts. And as we are rounding, make the total to be consumed at 281 watts. Make your wallplate voltage at 120 VAC. Makes for 2.4 A THAT!! is the amount of actual current required by your amp at peak. I believe that the nameplate on these amps is 275 watts. So, close-enough.

I have:
4x  Sovetek 6550WE
3x  RCA 12AU7
1x  5AR4

5amp slow blow fuse.

I will open up the amp and look for signs of arcing - the only area I could imaging is the Russian caps and the sockets in the front for checking bias. I removed all unused tabs from the sockets. I bent the functional tabs over 90* and wrapped in heat shrink just to be sure there was no metal to metal contact.

I have a voltmeter and clamp meter.
I can pickup one of those meters that plug in to the wall to read actual current used by an appliance while I am out.

It was late, I have not even looked at the tubes to see if they were damaged.

Peter W. wrote:
Do you have the means to test tubes?

I do not. Sadly, I gave a tube tester away to a friend a few months ago because "I would never need it". It was going into the garbage from an estate so I grabbed it for him. I could always borrow it for a bit, I am sure.

dalemurray wrote:

Peter W. wrote:
Do you have the means to test tubes?

I do not. Sadly, I gave a tube tester away to a friend a few months ago because "I would never need it". It was going into the garbage from an estate so I grabbed it for him. I could always borrow it for a bit, I am sure.

If you are anywhere near 19027 or 17859, I have a big Hickok. If the decay took more than a few seconds - that is, happened before the fuse blew, rather than as a result - it could be bad tube(s).

Peter W. wrote:
If you are anywhere near 19027 or 17859, I have a big Hickok. If the decay took more than a few seconds - that is, happened before the fuse blew, rather than as a result - it could be bad tube(s).

I appreciate the offer but I am a few states away. I just updated my profile to reflect this.

dalemurray wrote:

peterh wrote:
How did you blew it ? What was the sequence of events ?
A power off/on sequence could easily result in a broken fuse.

Amp had been on for about an hour and playing at very low volume (low enough not to wake my wife at the end of the hall). The tone of the music changed and faded out over the course of a couple seconds.

I passed all the start up tests without issue.

I suggest you use another fuse and power on again. Check bias , if it has changed
significantly you have your answer there. Let it be powered on , play some music
and watch for anomalies.

Where did you get your tubes from? I've had several rectifiers break right out of the box back when I was buying the cheapest tubes I could find. The yellow sheet mod can help but your 5AR4 may be shot.

I highly recommend Jim McShane for quality, tested tubes at a fair price.

Tubes came with the kit.

I just bought a new fuses and a meter between the outlet and amp.

No tubes installed:
16.3 Watts
.23 Amps

3x 12AU7 tubes installed:
23.6 Watts
.27 Amps

I put in the rest of the tubes, connected speakers, meter already on bias to see setting.
Bias was in the millivolt range; obviously not the .50 volts I set last night.

A few seconds later my 5AR4 tube ARCED!

So annoyed!
Now what?

corndog71 wrote:Where did you get your tubes from?  I've had several rectifiers break right out of the box back when I was buying the cheapest tubes I could find.  The yellow sheet mod can help but your 5AR4 may be shot.

I highly recommend Jim McShane for quality, tested tubes at a fair price.

Tube came with the kit.

Box: 5AR4

Marking on the tube:
Electron Tube
made in china

SUPER ANNOYED.

Yikes!!

First, check all the bias resistors and make sure they are intact and connected properly. With a dental pick, check _ALL_ the connections to the tube sockets, boards, terminal strips and so forth. ALL of them. You are looking for a cold/broken/incomplete solder. Be systematic, and work in really good light.

Obtain a new 5AR4. The post-blight examples seem to be about as sensitive as our president's ego to any sort of rough treatment.

With inputs shorted and speakers connected:

Run the bias pots half-way up if below half now. This will be your starting point moving forward, but assure there is *some* C- on the output tubes.

(A+ = Filament voltages; B+ = operating/signal voltage; C- = Bias Voltage)

Now, turn it on - WITHOUT the 5AR4. Make sure all the other tube are lighting, and look even.

Measure power use with all tubes in place except for the 5AR4.  
Allow the entire system to cool for about an hour.
Install the new 5AR4 and turn on - watch the bias voltage come up like a hawk, and watch the B+ come up - power use will increase on the external meter as it comes up - also like a hawk. If it goes above 280 watts/2.4A, SHUT IT DOWN.

Keep the bias voltage at/about your target voltage. After +/- 15 minutes, it should be sufficiently stable to set permanently - but check again in one hour. Note that the bias voltage is not dependent on the 5AR4, and should be present very nearly at turn-on. In point-of-fact, you can set all this without the 5AR4 in anticipation of its arrival. It will change as the transformer is loaded with B+, but a good start.

Peter W. wrote:Yikes!!
First, check all the bias resistors and make sure they are intact and connected properly. With a dental pick, check _ALL_ the connections to the tube sockets, boards, terminal strips and so forth. ALL of them. You are looking for a cold/broken/incomplete solder. Be systematic, and work in really good light.

Obtain a new 5AR4. The post-blight examples seem to be about as sensitive as our president's ego to any sort of rough treatment.

I ordered two off Amazon a few minutes ago - $35 minimum order required to get them delivered today.
JJ GZ34 / 5AR4 <-- what I ordered.

I will probe everything.
I made a visual inspection prior to reinstalling tube and testing. I saw no signs of arcing, etc.

Peter W. wrote:
With inputs shorted and speakers connected:

Run the bias pots half-way up if below half now. This will be your starting point moving forward, but assure there is *some* C- on the output tubes.

(A+ = Filament voltages; B+ = operating/signal voltage; C- = Bias Voltage)

If I understand you want me to power it on again with a new 5AR4 installed?
I just arc'd one a few minutes ago. Inputs were not shorted but were connected to my power off preamp and speakers were connected.

Am I supposed to pull the output tubes and measure the sockets?



[quote="Peter W."]Now, turn it on - WITHOUT the 5AR4. Make sure all the other tube are lighting, and look even.

Measure power use with all tubes in place except for the 5AR4.  
Allow the entire system to cool for about an hour.
Install the new 5AR4 and turn on - watch the bias voltage come up like a hawk, and watch the B+ come up - power use will increase on the external meter as it comes up - also like a hawk. If it goes above 280 watts/2.4A, SHUT IT DOWN.[quote]

Again. Not sure what or where to measure. Remember, this is my first foray into tube audio.

Peter W. wrote:
Keep the bias voltage at/about your target voltage. After +/- 15 minutes, it should be sufficiently stable to set permanently - but check again in one hour. Note that the bias voltage is not dependent on the 5AR4, and should be present very nearly at turn-on. In point-of-fact, you can set all this without the 5AR4 in anticipation of its arrival. It will change as the transformer is loaded with B+, but a good start.

If I understand this:
I can power this on without 5ar4.
I can set bias, and recheck bias without 5ar4 - this is purely to check stability?

I am making a run to Fry's electronics to pick up a pair of IN4007 diodes for the yellow sheet mod.

Yes. Purely to check stability. The bias circuit is independent of A+ or B+ (except will run a bit high with the system unloaded otherwise). The idea is to *MAKE SURE* you are getting the correct bias on all tubes before applying B+.

Generally, when dealing with a vintage amp such as my ST70, or Scott LK150, I will let it sit without the rectifiers in place for several hours to make sure nothing is wonky before I will let it off the bench and into polite society. Call it "proof-of-concept".

Note: If I were running any tube amp with solid-state rectification, especially something high-powered, I would install a delay-relay, if only just for giggles. Solid-state rectifiers are unmerciful, hitting the entire system with full B+ before it has anywhere to go.

First things first ... run thru the checklist one more time, step by step, and closely inspect each connection. Make sure they're all soldered well, and that there's no shorts to chassis. Look for any arcing or discoloration.

-If you went with the Russian PIOs, make sure neither the leads or metal cases are shorted to ground.

- Check all around the driver board for long tails that could also cause shorts. Things are mighty tight in that area.

- Review your connections at the T/P switches. That bit can be a little confusing.

- I'd also test the bias resistors. You can do that with a VOM if you remove the resistors and test them out of circuit. That would of course require resoldering them if they test good, which is probably a good thing as cold joints there are a common problem.

- Double check the multi can cap. If you were able to power up and listen for a while, the connections should be correct, but I'd resolder them anyway, and make sure the can is tight and solid to the chassis.

- DO review all the test voltages listed in the instructions to make sure they're in range.

- Follow the initial startup instructions in the manual when you're ready to power it up again.

- Cross your fingers.

PS - you listened at low levels how? Keep in mind you do need to have a load on the speaker terminals that match the ohms you built it for.

Oh. As mentioned, a slow start board can work wonders if you go thru a lot of rectifiers, as it's the initial surge on startup that really hits em hard, especially if your wall voltage tends to run high. I strongly recommend going with a thermistor in line with the main hot - that ramps up over a few seconds and never goes completely open, so you get the slower start AND it should drop line voltage 1-2vac. Another option is the WS68 copper top solid state rectifier. That also has the slow start built in. And yes, I know Weber now recommends the WS1 for straight audio amps, but I take that with a pound of salt. I've run a WS68 here with nary a problem, and other folk have used them for years. I've since gone with a Mullard GZ37 big bottle, and that seems to be pretty much bulletproof, but those are getting hard to find and tend to cost some big bucks anymore.

** I've also had no issues using a CL90 thermistor. And yes, the amp DOES draw more than what that unit is rated for, but that's only under full load, and that's only intermittent at most. Never had any problems here, and I've run them on 10 amp loads. The recommended thermistor is a CL80, but that tends to run a lot hotter. In either case, make sure to allow max clearance to any other components as both need room to breathe.

sKiZo wrote:Double check the multi can cap. If you were able to power up and listen for a while, the connections should be correct, but I'd resolder them anyway, and make sure the can is tight and solid to the chassis.

And make sure the quad cap is hooked up correctly with 1, 2, 3, and 4 connected per the instructions.  I did it wrong on my ST-70 mod and had results similar to what you're experiencing.

dave

sKiZo wrote:Oh. As mentioned, a slow start board can work wonders if you go thru a lot of rectifiers, as it's the initial surge on startup that really hits em hard, especially if your wall voltage tends to run high. I strongly recommend going with a thermistor in line with the main hot - that ramps up over a few seconds and never goes completely open, so you get the slower start AND it should drop line voltage 1-2vac. Another option is the WS68 copper top solid state rectifier. That also has the slow start built in. And yes, I know Weber now recommends the WS1 for straight audio amps, but I take that with a pound of salt. I've run a WS68 here with nary a problem, and other folk have used them for years. I've since gone with a Mullard GZ37 big bottle, and that seems to be pretty much bulletproof, but those are getting hard to find and tend to cost some big bucks anymore.

** I've also had no issues using a CL90 thermistor. And yes, the amp DOES draw more than what that unit is rated for, but that's only under full load, and that's only intermittent at most. Never had any problems here, and I've run them on 10 amp loads. The recommended thermistor is a CL80, but that tends to run a lot hotter. In either case, make sure to allow max clearance to any other components as both need room to breathe.

Hey Skizo, what about putting CL90's in series (or CL80's for that matter)?  Would that not 'share the load' as far as amperage and heat?  Total heat and amps is still the same, but spread out more (between 2 devices)?  Or would they BOTH get as hot, lol?

And wouldn't this actually drop more voltage, by using 2 in series?

I'm about to build a ST-120 with solid-state rectifier, and am looking for a way to reduce line voltage a little without using a bucking transformer.  I'm going to use a TDR board, and use the Weber WS1 that has a thermistor built-in also.

By the way, another extremely helpful troubleshooting dialog......many, many thanks to all of you!