getting started Debugging a buzz/hum (AB763 deluxe, new build)

[Enzo]

On cheap meters, DC measures OK, but if you set it to AC and try to read the ripple sitting on a 500v rail, it can't ignore it, and so reads WAY high. They lack an isolation cap at the input. AC is read on the DC function through an internal diode. The assumed calculation causes the reading.

I always suggest measuring a 9v battery on AC function. If it measures zero, then the meter is OK, if it measures like 12v or something, then the meter is not suited to read AC on top of DC.

[mikepukmel]

Have to ask,Mike,how are you connecting the leads to measure the ac on the filter caps?Even the cheapest of meters should get a reading.Dont quite understand what you are saying here "(meter is reading something like 800 on a 500 scale)"

Hi Stokes, all I did was flip the dial to AC, the max is 500 on this meter, then probe the hot side of the first filter cap lead with the red lead, and clamped the black lead to ground, same ground where the filter cap ground wire goes. The meter goes flooey. The numbers jump up and down 884 729 645 .... I thought I blew it out. I checked the AC wall voltage and it read about 122 steady, then set it to DC and got 410 or so on the same filter cap lead.

If you saw this meter, you'd laugh. Its a Kelvin 100 Made In Taiwan, probably close to 30 years old, cheap back when it was made. I took some audio courses at a school in Manhattan in the early 90's, they gave them out in one of the classes.

I have a line on a low end Fluke 115 on sale for 125.00 range, will order that soon.

[mikepukmel]

On cheap meters, DC measures OK, but if you set it to AC and try to read the ripple sitting on a 500v rail, it can't ignore it, and so reads WAY high. They lack an isolation cap at the input. AC is read on the DC function through an internal diode. The assumed calculation causes the reading.

I always suggest measuring a 9v battery on AC function. If it measures zero, then the meter is OK, if it measures like 12v or something, then the meter is not suited to read AC on top of DC.

Enzo, yeah, man, it went insane! I have to find a way to get something done this weekend, without a good meter (won't be here for a few days, but that puts me smack in the middle of a tough work week). Will see if any holiday sales on meters around here. I will try reading a 9v on AC but I think you pointed out the problems with the "free with 10 cereal box top" meters.

[mikepukmel]

Back to Stokes and RG's great posts:

I'm fairly sure I have two problems. The very low freq hum I think is 60hz or 120hz (hard to tell with overtones) is not normal for that amp, even with my terrible grounding scheme. I screwed up something. Id like to fix this first. I think something is broken here. If I pull out the first two preamp tubes it goes down a little and down a little more if I pull the phase inverter. So, I think something is building up, like a bad signal fed to each tube. (Will test for AC in the DC line when the new meter arrives)

Then, Id like to go back and figure out a layout that better fits the suggestions RG mentioned above. The Fender scheme of having the filter caps on top of the chassis and a bundle of wires running all over to get to and from the caps seems to cause some trouble here. I got the idea (mostly) about reducing noise propagation, but Im not sure how to physically rewire all of the grounds. E.g. how to rewire the preamp and PI tube cathode grounds that come off the cap and resistor, or the 1 ohm resistor grounded to the chassis from the output tubes.

There's an eyelet on that side of the eyelet board with the 3 (soon to be 5) cathode grounds. If I put a bus bar there just next to the eyelet board, and solder the cathode grounds there, should I wire one end of the bus bar to the - filter cap, i.e. run a wire through the chassis up to the cap? or can I make the physical connection at the bus bar itself?

As an aside ive seen photos of many home builds where all the 'ground' stuff, e.g. cathode wiring, input jacks, etc, go to a bus bar, but then the bus bar is wired right to the chassis at some point. This seems to be opposite of what RG suggested.

[mikepukmel]

Someone mentioned to avoid posting photos directly, but to host them someplace and put a link Here are some chassis photos on my flickr account. I put a link above but its not an eye catcher.


Here you can see the terrible mess of a grounding on the brass plate.
https://www.flickr.com/photos/145832...7689043482311/

[mikepukmel]

Cleaned up the solder joints, ready to plug in and give a listen. I bet that hum is gone now:

[stokes]

What is with that pot in that last pic?Did you see the pm I sent you?

[Justin Thomas]

Wow, how did you get a hold of that first guitar I ever modified, a Tokai TST-62?

Justin

[loudthud]

When a low cost meter gives you a strange reading on AC Volts or Ohms, reverse the Red and Black leads to see of you get the same reading. In the AC Volts case, the meter can't deal with the DC offset. In the Ohms case, there is probably some DC Voltage coming out of the circuit you are trying to test.

[mikepukmel]

What is with that pot in that last pic?Did you see the pm I sent you?

Just a bad joke, Stokes! Re your pm, yep, just replied.

[mikepukmel]

When a low cost meter gives you a strange reading on AC Volts or Ohms, reverse the Red and Black leads to see of you get the same reading. In the AC Volts case, the meter can't deal with the DC offset. In the Ohms case, there is probably some DC Voltage coming out of the circuit you are trying to test.

Thanks will do!

[Justin Thomas]

Just a bad joke...

I thought it quite good, actually... I put that pic up on Farcebook & got a few bewildered responses, thinking I actually did that...

Justin

[Chuck H]

All the good and experienced advice on grounding is a great thing to follow up on and idealize. But DO try other tubes as well. You mention that you "hope it's not the tubes because they are new". And I'm sorry that it's just not like that anymore. When I order tubes for a new project I buy extras on the likelihood that I'm going to get a couple of noisy preamp tubes and possibly even a set of power tubes with a bad one in there. Since you're questioning your own work that's the first thing you're going to suspect, but I can tell you from experience that a new, noisy preamp tube or even a questionable new power tube is always a possibility (for me it's actually been a probability). The first time it happened to me was about seven years ago and I spent no less than two days trying to troubleshoot my circuit when the hum was cased by a bad, new tube. Buzz from new tubes is also common. And a bad power tube can cause hum in the power amp. If such a power tube isn't bad in some obvious way (exploded, melted, doesn't light, etc.) you will need to take individual bias measurements of the power tubes to discover it.

[stokes]

Just a bad joke, Stokes! Re your pm, yep, just replied.

Ok,you got me,I was gonna give you a head slap.

[mikepukmel]

After cleaning out the gutters (most of them anyway), and searching for a roof leak, that I didn't find, I spent the afternoon removing the brass plate from my Franken-Deluxe. I had a whole bunch of ground points all over the place, mostly because i didn't want to solder to the back of the pots. I removed all of the grounds, put a piece of the biggest bus wire I have, and put most grounds to that, then grounded the end of the bus wire, and a few other grounds to a star like thing I put way on the other side of the amp from the power supply. Its part way there. The changes reduced the hum so that I can barely hear it now. Barely! It was so quiet that I practiced until after 9pm when my wife got home from her church meeting.

I now know how to make *excellent* ground loops!

Chuck, thanks for the info! Got lucky, in one of the dozen things I tried, I had swapped the 2 output tubes and it helped some. I was surprised that the bias measurements I took, the two tubes were off by so much. One was 6.45 watts the other 7.3, at idle. Thought matched tubes would be closer. The output trans resistance measurements may be a bigger factor than the tubes, though, I don't know what is good and what is bad. I measured 210 ohms on one side 199 ohms 0n the other. later on, after heating up 212ohms and 202 ohms.


Thanks to everyone again. Thanks to Randall for putting up photos of your awesome builds for me to check out how the grounding scheme should look like.

Since the standby voltage is right around 475v Im going to upgrade the first two filter caps to 500v.

Then will go back and do a more careful job finishing fixing up the grounding as per your suggestions.