Some blindingly obvious things to do in an amp build

[R.G.]

Upon reflection, I should have posted this here instead of "kit amps"

Some blindingly obvious things to do in an amp build

Here are some things that at least to me are incredibly obvious things that need included in an amp build that hardly ever are.
1. One fuse per winding on the power transformer. The AC power inlet fuse is there to prevent fires. It's a happy accident if it protects the power transformer. Put a fuse in the heater winding(s) and one in each side of the centertapped high voltage if so provided.
2. Install silicon safety diodes in series with a tube rectifer, one in series with each anode (not cathode like I first wrote 8-0 ). Then if the tube shorts, the silicon picks up the slack and does not let your power transformer and filter caps die.
3. Put bleeder resistors across the first filter cap. This ensures that it will bleed down in some small time to save YOU from shocks when you service it. Pick how much current you can allocate to them, or how much power you can waste, then figure the resistors. For instance, if you have a 400V supply that goes to 500V if there are no tubes in, and you don't want to waste more than half a watt, calculate P = Vsquared/R = 0.5 and get R = 500K, and rated at 1W. A 470K 1W would do, as would two 220K half-watts in series, or two 1M half watts in parallel.
4. Build it with a 3-wire AC cord and safety grounded chassis. There is no two-wire magic.

There are other things, but these scream to be done on any new build. Build it so it is harder to kill, and easier to service.

It's a bit of a PITA to site the fuses. I like side-by-side strips with several fuses in them. They screw or pop-rivet right to the chassis. Probably better to use a 2X block for the HV separate from any heater fuses. Note that the heater current is large, so get fuse holders with plenty of current rating for it. Maybe 10A or more.

Another thing I would say is a must is that the IEC pin terminals should be shrink wrapped, this is the only place I have ever been shocked. Amp turned off forgetting that its plugged into the wall.

As an embellishment, how about soldering on one of the neon power indicators to the back of the IEC so it's always lit when there's power coming in? I put LEDs in series with the bleeder resistors on the Workhorse line as an indicator that there was voltage on the filter caps and as a debugging aid. No light, voltage stopped before here.

One more thing to add. The light-bulb limiter. Build one and always do your initial power up with it. It will save you a lot of agony later.

[tubeswell]

Thanks for those timely reminders R.G. What about a varistor (or 3) across the PT primary? - I seem to recall that in one of your articles from somewhere.

[Chuck H]

How about a 220R resistor @ 2x amps watage rating across the secondary leads. Shouldn't effect impedance notably but probably enough load to save the OT from an open load spike.

... Thinking about it more I'm considering that many speakers have a pretty high impedance @ certain frequencies. There probably is a measurable effect. But I doubt you would hear it.

[NorCalTuna]

How about the heaters getting referenced to ground through 100 ohm resistors at least? This come up a lot...

[R.G.]

Thanks for those timely reminders R.G. What about a varistor (or 3) across the PT primary? - I seem to recall that in one of your articles from somewhere.

Well, some kind of transient clamping is needed. I used varistors, two of them in series to get a higher breakover voltage. Either this or the diode string from ground, which is for the same thing.

How about a 220R resistor @ 2x amps watage rating across the secondary leads. Shouldn't effect impedance notably but probably enough load to save the OT from an open load spike.
... Thinking about it more I'm considering that many speakers have a pretty high impedance @ certain frequencies. There probably is a measurable effect. But I doubt you would hear it.

Another good one, and for a similar issue. You get into practicality issues pretty quickly though if you have to do 2x the amp's wattage rating and you have a 100W head you're building. 8-0
Some kind of protection for open speaker lines is A Really Good Idea.

How about the heaters getting referenced to ground through 100 ohm resistors at least? This come up a lot...

That works, and is often more accurate a center tap than a winding centertap. I don't know if it's needed for every amp; maybe we should have a list of "things you really ought to consider" as well as a list of things you really ought to do.

[tubeswell]

Well, some kind of transient clamping is needed. I used varistors, two of them in series to get a higher breakover voltage. Either this or the diode string from ground, which is for the same thing.

Thanks again R.G. what does that look like in a schematic? (Got one handy?)

[R.G.]

I don't have an example to hand, but they're pretty common. There is a diode connected anode to ground, cathode to the plate of each output tube. The diode may actually be made of three or four diodes in series.

[pdf64]

some kind of transient clamping is needed. I used varistors, two of them in series to get a higher breakover voltage. Either this or the diode string from ground, which is for the same thing.

RG, I think that you may be thinking of the output transformer primary, whereas tubeswell may have been refering to the power/line transformer primary?

220R resistor @ 2x amps watage rating across the secondary leads

Chuck, in the case of the OT secondary being open circuit and the amp cranked, I don't think that it would put full power into the 220R protection resistor. Traynor used a 20 watt ceramic block 220 ohm resistor on the 50watt YB1A in that arrangement, see link. Pete.
http://www.0rigami.com/vb/traynor_ba...mkii_yba1b.pdf

[R.G.]

RG, I think that you may be thinking of the output transformer primary, whereas tubeswell may have been refering to the power/line transformer primary?

You know, it even says "PT" right there, and I just subbed in "OT" while reading. Doh!
Yes, you're right. Sorry for confusing things, guys.

@tubeswell: Yes, a MOV across the primary voltage is a great idea. For nominal 120V lines, use a 150Vac rated MOV, not the 130V ones, because MOVs drift a bit lower in voltage each time a transient breaks them over; a 150Vac rated one will last longer before it starts blowing fuses. Which brings up a point: most guitar amps don't include internal MOVs, so I don't mention it too often, but the way MOVs drift down in voltage can lead to the MOV being the cause of otherwise unexplained fuse blowing in equipment that contains them.

I think what set me off on the OT comments was that I *did* use MOVs for transient protection on the OTs on the Workhorse line. They work well for that as well, snubbing inductive flybacks above the normal operating voltages.

[Steve Conner]

I'm not sure of the need for MOVs. What are they protecting?

IMO, in an amp with a tube rectifier, absolutely nothing. So because of the above mentioned issues with the MOVs themselves, adding a MOV actually reduces the reliability of the amp.

In an amp with a SS rectifier there is more of a case for it. But if you arrange it so that the rectifier is always connected to the first filter cap, even in standby, then the filter cap helps to snub spikes from the mains. (the rectifier rectifies them, and feeds them to the capacitor which eats them)

The price paid is a higher voltage rating needed for the first filter cap, since it has to survive the full unloaded B+ indefinitely. So a more expensive part.

High powered bidirectional zeners called "TVS" or "Transzorbs" are available as an alternative to MOVs.

[pdf64]

I'm not sure of the need for MOVs. What are they protecting?
IMO, in an amp with a tube rectifier, absolutely nothing.

Agreed, though a case could be made for fitting them to protect overvoltage from generators at outdoor events / festivals etc.
My view is that it would be a 'nice to have' / value adding feature, rather than a blindingly obvious thing.

Safety Ground - is a transformer bolt suitable for this? I'm coming to the view that they aren't, as they often seem work loose over time. So, for a new build, a dedicated bolt, preferably with a star washer and nyloc nut, should be on the list.
However, with regard to this, for existing amps being serviced / repaired, what is required as best practice / good practice / minimum necessary to avoid negligence claims? I've seen Fenders which had their grounding tag under the PT nut and then also soldered on to the chassis; those seem very reliable. But then there are other methods which aren't so robust. Pete.

[guitarmike2107]

The EU standard says it should be reliably connected. I don't see anything that says it must have a screw all to itself? but if transformer bolts can rattle loose then that’s not reliable. Also if a transformer was replaced/removed and that connection was not restored then that not reliable either. A separate screw and nut with a nyloc nut or lock washer is easy to implement and cheap enough, should probably use some loctite too.

[R.G.]

I'm not sure of the need for MOVs. What are they protecting?

The primary-side insulation. Leakage inductance can keep a fast transient on the primary side of a transformer, where it can break over PT insulation, or even the inside of a switch. Granted, many, many amps have survived this kind of thing. But there are not accurate statistics on which ones didn't.

It is true that every part added adds its own reliability issues to any design. However, the other parts of a tube amp fail so much more often that adding a MOV - or tranzorb, which I loosely class as the same thing - doesn't change the failure rate of the other stuff much. I did say that I think they're a great idea, which I think they are. They're probably in the class below "blindingly obvious".

Safety Ground - is a transformer bolt suitable for this? I'm coming to the view that they aren't, as they often seem work loose over time. So, for a new build, a dedicated bolt, preferably with a star washer and nyloc nut, should be on the list.
However, with regard to this, for existing amps being serviced / repaired, what is required as best practice / good practice / minimum necessary to avoid negligence claims? I've seen Fenders which had their grounding tag under the PT nut and then also soldered on to the chassis; those seem very reliable. But then there are other methods which aren't so robust.

I've been told by safety certification test engineers that bolting safety ground to the chassis by a transformer bolt was once considered passing, and no longer is; the reasoning was precisely that the mechanical loading could loosen the nut, and that replacement processes invited leaving it out or making the connection incorrectly during service. There was another issue, that being that all grounding with a transformer mounting bolt uniformly connected the safety tag above the transformer leg, not directly on the chassis. The thinking was that the whole point of safety ground was connecting to chassis, and the transformer leg may or may not make good connection to chassis. But I suspect that each safety lab may interpret it themselves. The practice of using a dedicated bolt, used for nothing else, with toothed washers between the ground terminal and chassis is open to less question about integrity. The guy I talked to said he'd have failed my unit if I'd put it on the transformer mounting bolt.

On the issue of liability during repairs, a repair guy could use the defense that he had put it back the way the maker made it, so that he contributed no additional safety hazard beyond what it always had. The counsel for the plaintiff would then argue that if the safety guy knew that a separate bolt was more modern safety practice then it was negligence to NOT change it over. I have no clue where that would fall, and it probably depends on who looked most righteous and pitiful to the jury, and maybe what they fed the jury for lunch.

[Wilder Amplification]

Another thing I will add pertains to a lot of the "kit builders" out there...mainly the "Marshall DIY kits". Back in the day before Marshalls used 3 prong power cords, it didn't matter which leg of the mains got the mains fuse on it. Thus Marshall would always hook them up to the neutral leg for some reason.

However, with a 3 prong power cord, fusing the neutral would protect ONLY on a mains hot short to neutral. If you ever end up with a mains hot short to the chassis, the ground leg of the mains bypasses the neutral so the mains fuse on the neutral leg would never see this overload current, hence would never blow.

Placing the mains fuse on the hot leg instead will allow the fuse to blow in the event of either or happening.

Yet the instructions in a lot of these DIY Marshall kits specifically state to place the fuse in the neutral leg. Why? Because it's a "period correct" thing. That's how Marshalls were built back in the day before they had 3 prong power cords. But the kits themselves are coming with 3 prong power cords so the fuse MUST be on the hot leg in order for it to do its job both in the event of a mains hot short to neutral or the chassis.

Before anyone asks...no it's not going to sound any different.

[R.G.]

Good one, Jon. This is one that could save someone's life.

I think that the original 2-wire setup had the switch in one side and the fuse in the other, because that makes wiring the power line in easy. There is a terminal on the switch and one on the fuse where one of each line cord wires go. Converting this to 3-wire requires finding somewhere else to put the neutral wire, which now can't have the fuse in line.

A good solution to this impasse is to change out the power switch for a double-pole switch instead of a single pole. The double pole switch gives you a terminal for each side of the AC line, and you now have a terminal to run all the wires as well as breaking both sides of the AC line. This is a good companion to wiring a neon indicator on the inside of the chassis so you always know there is power inside the box when there is.