Update :73 Silverface Twin

[Tom Phillips]

...The meter I used for this may not be true RMS...

Well the nuances related to this stuff go on and on. When you are measuring a clean sine wave your meter will give the same reading as a "true RMS" meter. Therefore, it doesn't mater that your meter may not be "true" RMS. There is probably some detailed information about this on meter manufacture's web sites.

[gtrplayr1976]

Well the nuances related to this stuff go on and on. When you are measuring a clean sine wave your meter will give the same reading as a "true RMS" meter. Therefore, it doesn't mater that your meter may not be "true" RMS. There is probably some detailed information about this on meter manufacture's web sites.

Makes sense. So, when I was seeing 20vrms on my meter when I was testing the amp. The scope should have been seeing roughly 56 Vpp ?

When I set the test equipment up I had the 100mv signal at the input , the output connected to the load resistor with the scope + ,and meter + connected to the + side of the resistor. I had the grounds from the meters connected to the ground bolt at the power xfmr. Does this sound correct ?

[JoeM]

Yes, 20vrms as mentioned previously is 56v P-P. (56.56V P-P) Where do you have the scopes vertical sensitivity set?

You don't have to ground the meter at the transformer bolt, just measure across the load resistor.

[Tom Phillips]

Makes sense. So, when I was seeing 20vrms on my meter when I was testing the amp. The scope should have been seeing roughly 56 Vpp ?

When I set the test equipment up I had the 100mv signal at the input , the output connected to the load resistor with the scope + ,and meter + connected to the + side of the resistor. I had the grounds from the meters connected to the ground bolt at the power xfmr. Does this sound correct ?

That all sounds correct. The transformer bolt ground will work for your amp but I would use the ground reference at the speaker output or right at the load resistor because a) It is cleaner, b) It will also work for a speaker output that isn't referenced to chassis ground. B doesn't apply to your case but by making the hookups at my load box it works for me in all cases.

Cheers,
Tom

[gtrplayr1976]

That all sounds correct. The transformer bolt ground will work for your amp but I would use the ground reference at the speaker output or right at the load resistor because a) It is cleaner, b) It will also work for a speaker output that isn't referenced to chassis ground. B doesn't apply to your case but by making the hookups at my load box it works for me in all cases.

Cheers,
Tom

So connect both meters across the resistor +/- ?

[Tom Phillips]

So connect both meters across the resistor +/- ?

Sure. Using that method you can measure the power output without even opening the amp under test. You are measuring the power into the load resistor after all. There is no need to connect to a separate ground inside the amp.

Edit: I'm not saying that other methods don't work. This is just what I consider best practice.

[gtrplayr1976]

Sure. Using that method you can measure the power output without even opening the amp under test. You are measuring the power into the load resistor after all. There is no need to connect to a separate ground inside the amp.

Edit: I'm not saying that other methods don't work. This is just what I consider best practice.

Hey Tom, Found out the probes I bought were junk. I found the one that came with the unit ,and started getting much better readings. According to the equation they seem to match pretty well. I'm still blowing the fuse ,but still using 2.25 amp. The pictures show the readings as I progressed with the volume with the Vpd set at 5. The first video I had switched the Vpd to 10 ,and started increasing the volume over 3.5. Didn't last long at that point. The second video is the resistance reading I got after the fuse blew.

Sorry about the ear piercing tone. I had the meter on AC.

[gtrplayr1976]

Second video.

The number was actually higher than this ,but I didn't get the camera started quickly enough. I think it was around 16 ohms. Mute the volume !

[g1]

This just get's more confusing. Video seems to be making things worse, much easier to see the pictures.
It appears that you are getting about 30V p-p on the scope when clipping starts? What is the AC voltage reading on your meter at this time?
When you are measuring the resistance of the load resistor, it must be disconnected from everything, the amp, other test equipment etc. What is it measuring when cold? Hot?
You said something about the loud tone heard having to do with the meter being set on AC, that should not happen, was the meter set to ohms?
Also, can you link to the probes you were having trouble with so we can all avoid them? Or is there any chance they are 100X probes?

[gtrplayr1976]

This just get's more confusing. Video seems to be making things worse, much easier to see the pictures
It appears that you are getting about 30V p-p on the scope when clipping starts? What is the AC voltage reading on your meter at this time?

In the video I had the Vpd set to 10 ,and the meter was reading 21.7 V

When you are measuring the resistance of the load resistor, it must be disconnected from everything, the amp, other test equipment etc. What is it measuring when cold? Hot?

Yes it was disconnected. Cold measurement is 4 ohms. Hot is what you see on the video. I will try this again to be sure everything was done correctly.

You said something about the loud tone heard having to do with the meter being set on AC, that should not happen, was the meter set to ohms?

It was set to AC when I was reading the resistance. I usually use the DC setting to avoid the noise.

Also, can you link to the probes you were having trouble with so we can all avoid them? Or is there any chance they are 100X probes?

I got these off Ebay. I will try to find out where. They are switchable 1X-10X.

[g1]

Well that sounds much better. So you were getting around 100watts. This is under the assumption that the load resistor is 4 ohms. I thought it might heat up to a few ohms more, but nowhere near what you are getting.
At 12 ohms, you would really only be getting around 35watts.
As far as the tone from the meter, I guess the AC button toggles continuity mode when set on resistance?
Sometimes continuity mode is not very accurate for measuring resistance.

[gtrplayr1976]

Well that sounds much better. So you were getting around 100watts. This is under the assumption that the load resistor is 4 ohms. I thought it might heat up to a few ohms more, but nowhere near what you are getting.
At 12 ohms, you would really only be getting around 35watts.
As far as the tone from the meter, I guess the AC button toggles continuity mode when set on resistance?
Sometimes continuity mode is not very accurate for measuring resistance.

I will check it again this evening. The fuse blows when I get much past 20 volts. The resistor shows very near 4 ohms when cold. Yes the AC button does show continuity.

[Tom Phillips]

Looking at the video “Silverface Twin blowing fuse” linked in post #142 I see what appears to be high frequency parasitic oscillation starting when you reach maximum signal level. This is one of the things a scope can show. As you learn how to interpret the scope display you will recognize these anomalies right away. A well behaved amp will just start to flatten out at the top and the bottom of the sine wave when the amp reached full power. The trace on the scope display should still look like a single line under these conditions. Yours, however, widens out into a fuzzy trace at the top and bottom. That is indicative of high frequency parasitic oscillation. If you expand that part if the waveform you will see the oscillation.

The fuse blowing could be caused by the fact that the amp starts drawing more current from the 120V power line when the oscillation starts. The scope trace wasn’t shown when you increased the volume control over 3.5. The amp was already putting out full power at that point and I suspect that as you cranked the volume control up even higher it caused the parasitic oscillation to increase dramatically. This caused the amp to draw even more line current until the fuse finally blew. That’s my best theory.

Here is a test you can do to verify the theory.
With the same setup you used to make the video, move your Fluke 36 Clamp Meter to monitor the AC current in the 120V power feed to the amp’s input. Do you understand how use the meter to do that? (Do you have the manual?) Now you can monitor the current draw as you crank up the amp and determine if the rate of rise accelerates when the parasitic oscillation (That you now know how to recognize) appears and as you continue to crank up the volume. The lowest AC Current scale on your Fluke 36 is 0-200A so I’m not sure if you will have the resolution to easily see the current draw. This test just verifies the situation. Your primary goal is to eliminate the oscillation. So read on...

What is causing the high frequency parasitic oscillation? Could be lots of things and a forum search will find lots of discussion. Look especially for wiring layout recommendations) However, you have a good start because you can see the parasitic oscillation. If I was troubleshooting I’d operate the amp with the dummy load until the parasitic oscillation just started to appear on the scope trace. Then I’d poke around with an insulated device (like the much discussed chopstick tool) and determine if changing the wiring dress affected or stopped the oscillation. It’s also possible that your test load and meter hookup wiring is causing the problem so work on that area first when the parasitic oscillation appears. You want to be sure that the output signal is not making it’s way back to the preamp via the unshielded test leads or other test load wires. If this stops the oscillation then you will have shown that it’s not even happening when the amp is in the cabinet with the speakers. If that does not solve the problem then we move on...

I looked at photo #1 in post 96 and I see that your power tube grid feed wires do cross the plate wires at right angles. That is the correct wiring dress. However, one of the grid wires is cable tied in parallel with the standby switch wires. That’s bad practice. Take the grid wire out of that bundle. The grid wires on one of the prime candidates that you will be moving. I’m not talking about drastic re-positioning. Just small movements to start with. Your amp has been upgraded to a glass epoxy turret parts board. However, it’s still got all the god awful silverface hookup wiring mess. Rather than take the time to improve the wiring Fender decided to add various high frequency suppression caps in the circuit. This included 2000pf caps from the power tube grid feed to chassis ground. It looks like those have been cut out of your amp. My approach is to correct the source of the problem rather than add band aids like that so I’m not suggesting that you add them back to the circuit. Your photo also shows various shielded wires cable tied in parallel with other wiring bundles. I’d cur those free too. The original shielded cable that Fender used had less than perfect shield. It works OK but I sill wouldn’t tie it to the other signal wires. If it is freed up you can also nudge it around if necessary during your troubleshooting

I have more ideas but gotta go now. The above will keep you busy for a while.

Tom

[gtrplayr1976]

Thanks Tom. I'll be working with this over the weekend. To check the current I just clamp around a single + AC wire. The 2000pf caps you speak of weren't in this amp to begin with. There are discrepancies throughout. I couldn't find a schematic ,or layout that matched this exactly. The trace did flatten out some toward the middle of the video as I increased the volume ,but still had the oscillation. This definitely gives me some things to do.

[Tom Phillips]

...To check the current I just clamp around a single + AC wire...

Yes, any of the section with sufficient slack between the line cord input and the PT primary as it makes its way through the fuse and On/OFF switch.