Did you use a 1N400x? What is the last number on the diode? I always use 1N4007's because there is no reason to go smaller and they are pretty hard t o ruin in a 400V circuit..
Try swapping it out with another one. I've never overheated one myself but it's usually good to check the obvious stuff first. Pull out that diode and meter it. Have you tried measuring for AC after the diode? A DC meter will average the AC and give you a "DC" reading that is pretty much the average of the 60 cycle AC..
Check the resistance across each diode both ways, by swapping probe positions. Obviously without power and the caps drained. If it's conducting one way and not the other (which is what it should be doing), you'll get a reading one way and zero the other.
Thanks for the testing tips! I went ahead and replaced the diode using little alligator clamps for heatsinks. While doing this I noticed that I had left out the ground wire that attaches immediately after the 10uf caps.
I'm going to bump this thread in lieu of starting a new one.
For the bypass caps on the bias supply, is there any merit to sticking with the 10uF 100V caps as opposed to going up in value. The only caps I have on hand are 22uF 63V and 100uF 100V. I doubt it would make any audible difference and I can always change these out later but I wanted to see what others thoughts/experiences here were. I think I subbed out values on my last Express without issues FWIW.
I think it'll work fine. The more stable the bias, the better, and it should not be audible whatsoever (unless your values are so small that you end up with a tremolo effect).
One possible drawback is that larger caps mean a longer time for the bias voltage to reach the desired level at start-up, and this might result in under-biased operation for a short time. With 10uF's, about 4 seconds will be required. The 100 uF's will require 10x that or about 40 seconds. I think it would be best if the bias voltage came in well ahead of the cathodes reaching operating temperature, in case somebody hits the DC switch early.
Good point, that I hadn't thought about at the time.
Having a look at it in a SPICE model I came up with the following settling times for a pair of 10uF, 22uF and 100uF caps:
10uF: 6s
22uF: 14s
100uF: 50s
I retract my statement about 100uF caps being fine!
So once fully charged the voltage or current supply won't decline?
It's hard for me sometimes to wrap my head around this time constant stuff.
Like on the power supply or coupling caps.
When I built my Dumbleator, I used some large filter caps.
Funny thing is, I can turn the standby switch to standby and play the guitar for probably 3-4 minutes before those caps drain down and the volume changes and fizzles out, so I guess I can say they are holding a pretty good charge.
Structo wrote:So once fully charged the voltage or current supply won't decline?
It will, but it will be dependent on the current drawn by the load and the internal resistance of the caps themselves, and there's no significant load on the bias circuit at all.
Structo wrote:When I built my Dumbleator, I used some large filter caps.
Funny thing is, I can turn the standby switch to standby and play the guitar for probably 3-4 minutes before those caps drain down and the volume changes and fizzles out, so I guess I can say they are holding a pretty good charge.
The Dumbleator consists of only two triodes operating in class A, so it's drawing a constant current from the power supply, which will only be in the order of a few mA.
Add in the fact that we know starved-plate designs work down to pretty low voltages as Dana's VVR has demonstrated so well, and it's not surprising that it will carry on playing for some time with no B+ as long as the heaters are on.
Yes, I didn't even consider that the heaters were still on, that makes a lot of sense.
I just never experienced that before since most amps mute pretty quickly when put into standby.
Thanks!
