So here’s a fun one (AD-100 clone woes)

[Mr_bibbles]

Chuck I really appreciate your sage advice, let me try and relay that Info as best I can:

1. There’s no choke that I can see.
2. The screens are fed B+ from 2x paralleled 100uf caps. Two leads connect to pin 6 from each half of the power section, and an additional lead
connects two separate tubes together at pin 6 (in case of failure?)
3. Each of the 8 screen resistors connected from pin 6 —> 4 is a 1.5k 5w cement.
4. The transformer leads feeding the screen supply connect to the center of a 1n4007 bridge, one half leading to the standby, the other half leading to to secondary fuse and then the aforementioned filters.
5. Screen Voltages:
A. At transformer leads: 194/189VDC
B. After bridge (filter caps/screen resistor pin 6): 400vdc
C. Pin 4 after screen resistor- 400vdc

[g1]

200uF at the screen node is really stiff. I'd drop that to the point where he complains about it.
I've seen screen nodes as low as 10uF to basically function as screen current limiting.

[Helmholtz]

As I said above, the amplifier has 4.7k 10w resistors from ground to the speaker +. Since you suggests 220, is it possible these dummy loads are so large that they’re causing issues??

I think that a base load of 4.7k is too high resistance to provide any protection. 220R will at least protect the OT in case of load-drop. But not the tube screens. Even doubling nominal speaker impedance typically causes screen dissipation to exceed its limiit with large signals.

...and I supplied the speaker cable.

To me this is maybe the most important statement/hidden advice in Mick Bailey's post.

[Mr_bibbles]

Word! If I were to disconnect one of the parallel caps, would that drop the current to a more acceptable point? Is there any reading you’d recommend to help me understand the cap’s effect on a screen node?

[Helmholtz]

If I were to disconnect one of the parallel caps, would that drop the current to a more acceptable point?

Most probably not. Can you post a drawing of the complete screen supply?

With a screen supply voltage of around 400V and 1.5k screen resistors I don't think that screen dissipation is an issue with nominal plate load.

[Chuck H]

Most probably not. Can you post a drawing of the complete screen supply?

With a screen supply voltage of around 400V and 1.5k screen resistors I don't think that screen dissipation is an issue with nominal plate load.

But what is the plate load? And are we sure this amp isn't sometimes plugged into too high an impedance? If there are two jacks on the back of the amp for speaker cabinets, how are they wired? Are they being used or if two cabinets are plugged in is the player using in/out hacks on the speakers themselves? How are those wired?

[Helmholtz]

But what is the plate load? And are we sure this amp isn't sometimes plugged into too high an impedance? If there are two jacks on the back of the amp for speaker cabinets, how are they wired? Are they being used or if two cabinets are plugged in is the player using in/out hacks on the speakers themselves? How are those wired?

All valid questions. But I am not the person who could give answers.
I think I gave the relevant hints, though.
Too high an output impedance is risky for the screens in most amps.

[Chuck H]

Oh I know. I wasn't asking you. I was just using your post as a segue for the questions. Usually I cut and paste the quote to a general response but I guess I got lazy.

[g1]

A couple sets of schematics here, are either close to what you have?

AD200B MKIII (only one pair power tubes shown) https://music-electronics-forum.com/...l=1#post541579

AD200 https://music-electronics-forum.com/...l=1#post541731

[Mick Bailey]

One thing to look out for with KT88 tubes is how pin 1 is connected. With many tubes this connects to the shell - even if the spec sheet doesn't show it or refers to it as NC. It's best to measure and check to find out. Many custom builders will then use pin 1 as a tie point for the grid stopper. So, now all the tube shells are at bias potential and there's a metal ring hanging off each grid. I don't have any evidence to support this, but my conjecture is that this can cause stability issues. I've had mysterious faults where removing the grid circuit off the shells has resolved problems. My fix is to install new grid stoppers without tying them to pin 1 and then connect this pin to ground.

Another issue with having the shells connected to the grids is that anything that shorts any one of the shells to ground immediately kills the bias to all tubes. A spring retainer could make contact with the shell, for example. Ok, it kills the drive as well but the tubes will very quickly red plate.

[Mr_bibbles]

Thank you all for your continued interest! Tomorrow I will sketch the screen supply, and give more detail
About the plate circuit.

G1:

For now I can say it’s closer to AD200 (4 tubes). But the grids are 2.2k’s and the screen filters are 200uf rather than the 110 total of two 220’s in series.

Chuck:

The 2 shorting jacks are paralleled to the OT selector with 2 additional speakon connectors. The 1/4 have a 4.7k 10w to ground from the shorting terminal, not he positive as I previously stated.

Mick:

Pins 1 and 8 are tied together on each tube with 1 ohm 2w MO to ground, with a 2afuse in between.

[Mr_bibbles]

Apologies for my lack of artistic vision:


This is the screen supply- note that the 6a slo blow fuse is usually Ok, it’s the mains that blow.

I only drew the first set of tubes but they’re tied pin to pin in a line, with the screen filters hooked directly to 1.5k’s.

[Chuck H]

While I think g1 is right about the screen filter, I have seen high value screens before in bass amps. Granted, they weren't typically overdriven, but I still have to go with the possibilities of:

1) An intermittent fault in the circuit causing a loss of bias or load or a condition of shorted load

2) Wrong load, no load or a shorted load on occasion due to user error or a bad speaker cable

3) The preamp design can deliver too much signal to the power tube grids and the owner is hitting the grids too hard with signal

Or some combination of the above. I think installation of 220ohm safety resistors is a good start. Also, clip your meter to each grid pin and smack the amp to see if bias is intermittent. Monitor bias at each grid for a time while the amp is running just to see if anything weird happens. Test the output jack while smacking the amp too. This time with the amp off while monitoring a cable plugged into the output jack. Any changes in resistance?

NOTE: Checking bias at the bias supply is not sufficient. You need to know what's happening at each and every tube socket under simulated playing conditions. That is, amp is on for a time and there are vibrations.

Get as much info as possible from the customer about the failure. How does the amp behave before a failure? Crackling? Cutting in and out" Gets quiet? Louder? More distorted? etc.

[Mr_bibbles]

Chuck, I will do all of those vibration tests today. One question, could a loss of bias cause a heater to plate short though?
Here’s a picture of the speaker jacks, are these already installed dummy load resistors you’re suggesting?

2. The PI caps are .1, I could try dropping hem to .047 if I don’t find a vibration fault?

3. What changes would you recommend to the filters? Just disconnect one of the 100’s?

[Helmholtz]

note that the 6a slo blow fuse is usually Ok, it’s the mains that blow.

I think the 6A HT fuse is rated too high. A Marshall Model 2001, 375W, 8x6550 bass amp has a 3A HT fuse. Typically the HT fuse should blow before the mains fuse. Otherwise the HT fuse doesn't provide additional protection.

A. At transformer leads: 194/189VDC

Do you mean AC?


BTW, 1N 4007s are only rated at 1A continuous average current. Not sufficient for outputs >200W.

Did you clean and retension power tube socket contacts?