I have a couple of dynaco mk4 mono blocks. One blows the 2A fuse when I switch it off for 5 to 10 seconds and then switch it on again. It has a JJ brand EL34, the other has some old stock type. This seems to be happening when the tube is still hot, but B+ has sagged significantly during the power down, causing a large in rush to the PS input capacitor (good and correct value). When this occurs there is arcing within the EL34.
Is this a problem with JJ EL34, or a problem with doing what I should not. This happened with a second JJ tube also. Have others had this issue?
Is this a problem with JJ EL34, or a problem with doing what I should not. This happened with a second JJ tube also. Have others had this issue?
The process you are describing is called a "Hot Start".
Has the Dynaco MK4 been modified in any way?
Specifically, has the Fixed Adjustable Bias circuit been modified?
Extra filtering (larger capacitance) or a cap on the wiper of the bias pot could cause the negative bias to come up slowly. The hot cathode of the EL34 would draw excessive current until the bias supply voltage stabilized to its normal voltage.
Also, check the coupling caps that connect to the grids of the EL34s.
If they are leaky when the voltage is high (driver DC plate voltage, and the Bias voltage actually add across the coupling cap, and they may leak there.
A few volts of leakage reduces the negative bias on the EL34, that plus the already low bias voltage during a hot start might be enough to cause the problem.
Did someone replace the coupling caps with a larger value, or the grid resistors with a larger value, or both?
A longer charge time, and a 'hot start' do not go together very well.
If the coupling caps are the old plastic cased "Black Beauty" replace them.
I would have to look at the schematic to get any more ideas.
Has the Dynaco MK4 been modified in any way?
Specifically, has the Fixed Adjustable Bias circuit been modified?
Extra filtering (larger capacitance) or a cap on the wiper of the bias pot could cause the negative bias to come up slowly. The hot cathode of the EL34 would draw excessive current until the bias supply voltage stabilized to its normal voltage.
Also, check the coupling caps that connect to the grids of the EL34s.
If they are leaky when the voltage is high (driver DC plate voltage, and the Bias voltage actually add across the coupling cap, and they may leak there.
A few volts of leakage reduces the negative bias on the EL34, that plus the already low bias voltage during a hot start might be enough to cause the problem.
Did someone replace the coupling caps with a larger value, or the grid resistors with a larger value, or both?
A longer charge time, and a 'hot start' do not go together very well.
If the coupling caps are the old plastic cased "Black Beauty" replace them.
I would have to look at the schematic to get any more ideas.
Last edited:
No mods, caps (black beauties) were replaced with equivalents. Selenium rect replaced with silicon. I have been through it for faulty components and only replaced what was needed. They work and sound great.
Are you talking about the GZ34 rectifier tube? JJ GZ34's are known to be pretty weak compared to old stock tubes like Mulllards. Doing a hot start will likely fry a JJ. So don't do that!
jeff
I am suspicious of the JJ EL34 versus old stock. Could it be that the old stock current limits better and handles the current limit (technically overload) better than the JJ new manufactured.
The main problem is power recykling. Don't ever !!
The difference between tubes could be cathodes with larger thermal mass,
or a larger current capacity ( better imission) which causes to much current,
or it could be external differences such as capacitor differences. What ever is
the difference stop power cykling.
Yea, I figured that out, no power cycling. It's a bit like picking on a scab though, know you shouldn't, but just can't resist.
Thanks to all.
Thanks to all.
Do your silicon diodes give the same voltage as the selenium ones? Or perhaps more?
Regards, Gerrit
Regards, Gerrit
The silicon diode replacement is for the bias supply. A bit more voltage, but easily compensated with the bias adjustment pot. Can't remember how much it increased by, but a couple of volts would be expected. No problem as long as the filter caps can handle it. I suppose a voltage dropping resistor could be used if one felt that they wanted to keep the voltage the same.