MOV, Gas Discharge tube, or TVS Diode protection for output transformer??

[SoulFetish]

Trobbins, thanks for the tip about connecting them in series. That actually helped me solve another issue I had run into on the mains side of my Power transform. I wanted the ability to switch means voltages so I designed it to have primary connections for 120, 220, and 240V. I wanted to be able to protect against transients for each of those taps so I initially had water protection across the primary windings in from each line to earth. Like this:



But I realized I can't do that because that MOV for 120V neutral side is all we switched in and would conduct to early when connected to the 220 and 240 taps. I realized I'm probably going to lose my line to ground mov protection unless I can find use another type of switch. But I can still have proper protection across each leg of the primary windings using appropriate series Connected MOVS bridging the taps. I don't see why this wouldn't work seeing that it will always maintain its voltage divider ratio. Does that male sense?

[trobbins]

In general, it would be better to alleviate any AC mains disturbances at or near a domestic main switchboard (or similar upstream point of coupling), and not within a particular item of equipment. Any MOV that is full-time connected to the AC mains becomes a local sinkhole for significant disturbances 24/7, such that high loop currents could circulate through the establishment to that item of equipment, and could cause neutral voltage disturbances on the way.

I can see the benefit in applying a MOV to the PT primary winding, as that could alleviate AC mains switch contact pitting, depending on the timing of switch on/off, and could alleviate damage to the PT if something bad did befall the mains. I guess you could put a MOV on each of the PT primary windings for that purpose, if you know the tapping is being manually changed.

[SoulFetish]

I can see the benefit in applying a MOV to the PT primary winding, as that could alleviate AC mains switch contact pitting, depending on the timing of switch on/off, and could alleviate damage to the PT if something bad did befall the mains. I guess you could put a MOV on each of the PT primary windings for that purpose, if you know the tapping is being manually changed.

It's really there to protect the PT in the event of a large transient spike/surge which could destroy it. The mains selector would half to be manually switched and the chassis pulled out of the head (or I may put in a small screw access panel).
This is the set up here:





Windings and MOV connection:

[KevinOConnor]

Hi Guys

I never use MOVs nor buy any of those "spike protector power bars" that have them because MOVs degrade as they absorb energy, and their ultimate point of zero-protection is unknown. The diode protection method for OTs seemed a bit pointless in some regards, too, as RG has commented. Instead of using active devices or reactive ones, I use resistors to damp the OT. A simple R from anode to ground on each side and a load resistor on the secondary is all one needs to impose a maximum resistance to the tubes if the load opens. The Rs absorb energy benignly and last the life of the amplifier. My amps have had no blown OTs since I began building nor has anyone bulding the designs shown in TUT3 or TUT5.

JF: Thirty ohms is a common value for pure resistance in the reactive load boxes built by various manufacturers. Thirty ohms dissipates a lot less power than a true 8R resistor, so one can use a smaller part and deal with less heat. The reactive components bring the nominal value to 8, 16, whatever,over most of the frequency range, but you see 30R even when no reactive elements are present, as in the Ultimate Attenuator you showed in post-18. That design was shown in TUT4 and was designed by a tech in Vancouver BC, named Ho. It changes the sound fed to the speaker because the EF stage has lower output impedance than most tube amps and the load is fully isolated from the tube amp.

Have fun

[trobbins]

I never use MOVs nor buy any of those "spike protector power bars" that have them because MOVs degrade as they absorb energy, and their ultimate point of zero-protection is unknown.

Kevin, if you checked out the lifetime and degradation performance literature of MOVs with any effort, the reality would be that there is no degradation effect for modern MOV's subject to less than 'moderate' impulse energy levels (compared to their impulse energy rating). Operating a MOV in an energy limited situation, and with normal margin for selection of parameters, is effectively as reliable as a resistor degrading with time.

Ciao, Tim

[Leo_Gnardo]

if you checked out the lifetime and degradation performance literature of MOVs with any effort, the reality would be that there is no degradation effect for modern MOV's subject to less than 'moderate' impulse energy levels (compared to their impulse energy rating). Operating a MOV in an energy limited situation, and with normal margin for selection of parameters, is effectively as reliable as a resistor degrading with time.

It's all well and good, but how's anyone to know whether an MOV has, umm how do we say this? "Used up its nine lives???"

[Enzo]

How do you know ANY part is on the verge of failure? That EL34, will it lose a screen grid tomorrow? That fuse that slightly bends with each powr up, will it fatique out tomorrow at the gig? That 82k resistor in your DeVille, will it open tonight? We use parts within reasonable parameters, we can expect reonable service from them. Are there any guarantees? No. Not for anything.

[trobbins]

Same situation with a resistor, a test is needed, it's just that the test is a little more onerous to perform (not dissimilar to testing for coupling cap leakage).

[nickb]

Hi Guys

I never use MOVs nor buy any of those "spike protector power bars" that have them because MOVs degrade as they absorb energy, and their ultimate point of zero-protection is unknown. The diode protection method for OTs seemed a bit pointless in some regards, too, as RG has commented. Instead of using active devices or reactive ones, I use resistors to damp the OT. A simple R from anode to ground on each side and a load resistor on the secondary is all one needs to impose a maximum resistance to the tubes if the load opens. The Rs absorb energy benignly and last the life of the amplifier. My amps have had no blown OTs since I began building nor has anyone bulding the designs shown in TUT3 or TUT5.

JF: Thirty ohms is a common value for pure resistance in the reactive load boxes built by various manufacturers. Thirty ohms dissipates a lot less power than a true 8R resistor, so one can use a smaller part and deal with less heat. The reactive components bring the nominal value to 8, 16, whatever,over most of the frequency range, but you see 30R even when no reactive elements are present, as in the Ultimate Attenuator you showed in post-18. That design was shown in TUT4 and was designed by a tech in Vancouver BC, named Ho. It changes the sound fed to the speaker because the EF stage has lower output impedance than most tube amps and the load is fully isolated from the tube amp.

Have fun

I assume that "from anode to ground" was as typo and what was meant was "from anode to OPT centre-tap".

That's interesting about the 30 ohm loads. If it's true that they are relying on L's and C's to lower the impedance then that reactive power is getting dissipated somewhere. It must be in the tubes if it's not the load resistor. It seems that might be a problem. You save on the resistor but bake the tubes?

[trobbins]

The simplest MOV "ok test" method is likely to use a digital insulation resistance megohm-meter as the voltage supply and 'current sense', and a DVM to measure the MOV voltage.

Luckily my megger (Digitech QM-1492 or Honeytek A60B) has a 1.4mA current limited dc supply at 250, 500 and 1kV nominal test level. For a 330Vdc 1mA min spec small MOV disk that I use for OT protection, a direct connection across the megger on 500V range shows a 'resistance' of 1.3Meg, and a parallel terminal voltage of 355Vdc, indicating a test current of about 0.3mA.

A MOV in an amp would need to have one terminal disconnected for testing, due to the parallel path through eg. an OT primary winding.

For valve amp restoration, I find an insulation resistance megohm-meter is a commonly used tool for checking magnetics and coupling caps, and 2nd hand units are pretty cheap on flea-bay.