Zener clamp on effects loop send

[Gaz]

I was reading the effects loop chapter from Merlin Blencowe's preamp book and came across this section analyzing different commercial loop drivers. In the Peavey 5150 example he talks about a potential high voltage at the output while the coupling cap to the output send charges at startup. Here'is the section:




If I put two opposing zener diodes (say 9V) on the output, after the cap to clamp the spike at startup, how can I determine the rating? Do i need to add some series resistance? I've been looking at general electronic sites/calculators, but am confused about calculating for this particular applications. B+ at this point in the circuit is around 350vdc.

Thanks!

[Enzo]

I have great respect for Merlin, but Since the 5150 came out I have never seen a voltage spike problem on the FX send. That 100k to ground at the output keeps teh cap from sending charging currents through.

[Greg Robinson]

With deference to Enzo, I have seen this happen, with a brief pulse of around 90V when switching standby.
With most pedals this is of little consequence, as they should all feature some degree of input protection - some series resistance (say 1k or more) on the input to an opamp, bjt or jfet input or series resistance and zener clamping for mosfet should be more than adequate.
But it's a simple thing to avoid the potential problem, I fit 5V1 zeners back-to-back, standard 1/2W parts. This clamps that pulse to 5.7V (reverse voltage + forward voltage of the two zeners). This will protect any 9V powered pedal but will not clip the output of the 5150 send.
The benefit may be more-or-less imagined, but it certainly won't hurt.

[uneumann]

With deference to Enzo, I have seen this happen, with a brief pulse of around 90V when switching standby.

Just curious as to why this is mentioned as specific to the 5150. Doesn't every cap-coupled stage produce a power-on/off spike - especially when used with a standby? I can see the concern for the pedal, but doesn't the same problem (if it is one) occur for any cap-coupled FX output?

[Greg Robinson]

Just curious as to why this is mentioned as specific to the 5150. Doesn't every cap-coupled stage produce a power-on/off spike - especially when used with a standby? I can see the concern for the pedal, but doesn't the same problem (if it is one) occur for any cap-coupled FX output?

I can't speak for Enzo but I for one thought you were discussing the Peavey 5150 specifically, though on re-reading it's clear that you were only referencing the text which used the 5150 for its example. I was also specifically addressing the 5150 in my post.

Does the same problem occur for every cap-coupled FX output?
No, if the problem is defined as being only if the voltage surge is above some agreed "safe" level (9V is as good a place to call it as any).
Of the examples Merlin shows in section 12.7 (as well as other examples such as fig 12.3), all are AC-coupled, but only figures 12.5 (Boogie) and 12.7 (Peavey) have this issue.

[Enzo]

It is what the 5150 is an example OF that matters here. That is a cathode follower, so there is close to B+ voltage in the circuit at the cathode. Other amps are not driving the FX send with a CF.

Perhaps I was unclear. I have never seen a problem from this. I don't recall ever having a 5150 killing FX pedals.

[Gaz]

Enzo, thanks for the empirical evidence as always. I meant to ask if this was a known problem (I could not turn up any results online about the 5150's zapping pedals).

Greg, that's interesting that you noticed it (but I guess you have Merlin's book too!), and added a clamp. Even though it's probably not something to be concerned about, I am still curious about how to add a clamp for overvoltage protection in case in might come in handy for other circuits. In the case of an cap charging, do I need to have series resistance in line with the diodes?

[Greg Robinson]

Greg, that's interesting that you noticed it (but I guess you have Merlin's book too!), and added a clamp. Even though it's probably not something to be concerned about, I am still curious about how to add a clamp for overvoltage protection in case in might come in handy for other circuits. In the case of an cap charging, do I need to have series resistance in line with the diodes?

First, sorry, I wasn't paying attention when uneumann replied above, thought it was you Gaz. Apologies to both of you.

No need to worry about adding extra series resistance, this will only worsen output impedance and is not required. The pulse is very brief and the cathode source impedance is high enough to keep the max dissipation in the zeners very low even during that brief period. Back-to-back 5V1 zeners directly across the output is fine. Standard 1/2W parts (or even lower rated, but these are less common) are fine.

EDIT: Like Enzo, I've never known a 5150 to zap a pedal, but if I remember correctly I discovered this issue before reading about it in Merlin's book. I once had a concerned customer bring me one after getting a tingle when switching standby and holding the jack of an unconnected send lead. They likely had slightly damp hands and must have just been holding the tip of the jack. The available current should not be a direct danger, but the surprise could lead to jumping back and injury by knocking something, so I've routinely added clamping zeners to such amps since.

[Gaz]

Greg, thanks, that perfectly answers my question about the diode rating and that the ~600R source impedance is enough. Is there simple math or calculator I can use to figure that out. I would love to know how to calculate the dissipation of zeners.

That's an unusual circumstance, with the the customer holding the send cable and getting a shock, but makes sense! I after that experience your better safe than sorry approach is reasonable and I imagine the diodes are easy enough to retrofit.