Picture of Isolation transformer upgrade for old guitar amps
Save your skin! Upgrade that scary old amp with an isolation transformer.

Quite a few old amplifiers (and radios) back in the day drew power by directly rectifying the household "mains" wiring. This is an inherently unsafe practice.

Most guitars connect the bridge and strings to the ground (shield) wire on the guitar cord, essentially using the player as a "noise shield." In transformer-less amps, the Neutral wire of the mains is often used as the "ground." With a two-prong cord, Neutral and Hot can be switched (which could place the amp's ground on the Hot wire!) In other words, playing a guitar amp without an isolating transformer could be like sticking a fork in a wall outlet.

Isolation transformers limit the amount of current that can be supplied to the amp (and consequently to the guitar player) if any shock hazards arise, and eliminate possible "hot" ground issues.

In addition, we'll install a three-prong cord, so the amp has a proper earth ground. And a fuse, too. The earth ground and fuse help to maintain a sane ground reference, and protection from shorts.

And we'll incorporate the changes on a small "module," so as to change the original as little as possible. If someone is crazy enough to revert to the original setup...they can do that.

This mod works with radios, too. In fact, many of these amps were called "radio tube" amps, or "AC/DC amps"--like their radio counterparts, a transformer-less amp could be plugged directly into a DC or battery power supply without modification. A decently-sized bank of batteries were required (over 100V), but that was once commonplace.
1-40 of 73Next »

Hello, I just joined Instructables because of this article. Thanks for the great info. My questions is, (and please take into account that I'm an electronics newb) couldn't I bypass the ss rectifier and use a more robust iso transformer in order to take care of the half-wave rectifier issues?

gmoon (author)  gaiacidemusic1 month ago

Yeah, that is an option. It's essentially like running the amp on the mains without the rectifier upgrade.

So long as it doesn't mess up any other devices in your home, with noise, etc...

RodO1 made it!7 months ago

Thanks Doug!!

I have a Canadian made Pine/Pepco 801 ARC amp with 50C5, 35W4, 12AX7.

I have followed this excellent instructable and it works perfectly!

Barely any perceptible hum.

I'll upload the redrawn schematic for the original and the modified.

Thanks for the PDF drawing. I have a similar "stencil" amp, Chassis 801, CSA LR19952. I've looked at your PDF drawing and my intentions are to apply the same mods to my little amp. However I am worried to connect the Green and Secondary center white tap to the chassis of the amp as there is 39 volts on the chassis (please see attached schematic); I guess this would result in a huge hum through the speaker. Your help is very much appreciated.

Schéma Camaro.jpg
gmoon (author)  RodO17 months ago

Awesome, glad the ible helped!

Newslogger44.2 months ago

Today at 10:55 AM


I read your article about rectifiers with great interest and wonder if you can solve a similar problem.

I own a vintage radio (see link below including schematic) which includes two half-wave rectifier tubes--namely the 2X3--long since obsolete and having no direct substitutes.


The original 2X3s in the radio eventually burned out, so the closest tubes I found with similar characteristics were 2W3s which worked fine up until recently when they began performing erratically, going intermittently cold and then hot--clearly malfunctioning.

The 2W3s tested okay on my vintage Jackson 648 Dynamic tester, however such tests can be deceiving. When they work at all, these all-metal 2W3 tubes run very hot inside the radio--virtually untouchable--although apparently this is common with rectifiers.

Specs: 2X3 2.5 volts 2 amps 4X base diagram
2W3 2.5 volts 1.5 amps 4X base diagram

Seeking a better solution, when I typed "solid state rectifier tube replacement" into the Google Image search bar I saw a variety of solid state options made by Sovtek and others, so I am wondering if you are aware of an existing solid state replacement for the 2X3 or if Sovtek or another manufacturer would actually need to "clone" this obsolete product--if indeed that would be economically practical on their part, although since such rectifiers are of a pretty basic design, perhaps my idea has merit after all.

Yes, a few online tube dealers do have some old stock 2X3 tubes for sale but they are expensive, unlikely to be available in stock forever due to their obsolescence, and in the long run such tubes would have a relatively short life span whereas a solid state replacement would presumably solve that issue forever.

Your thoughts on this and potential solutions would be very much appreciated.


gmoon (author)  Newslogger44.2 months ago

Hi Newslogger44--sorry for the delay, I didn't check my I'bles account for a couple days.

RE: solid-state replacements for rectifiers, I'm aware that they exist. The trouble with using one in this (my) application, is that for "radio tube" amplifiers, the tube filaments are all wired in series, directly on the mains AC. A plug-in SS replacement won't work, unless it has a resistive element (with the same current draw, at the correct voltage) for the filament pins.

Most high-end tube amps don't do that, they have a separate filament supply (usually 5 or 6 volts secondary on the power transformer) dedicated only to the rectifier tube. Since no other components are wired in series with the rectifier filament, those SS replacements simply forego the filament supply altogether--which in fact makes the PT run a little cooler.

This is actually discussed within the I'ble, mostly in the sections about the series AC filament...

It it possible to make your own solid state replacement--they aren't very complex. If you search for them, I'll bet you find a schematic. If I recall, they are generally just SS rectifiers with a current-limiting resistor in series to mimic "sag."

Otherwise, if you're using the 2A version in place of the 1.5A one, there's a good chance you'll damage the amp...

nathangill4 months ago

Thanks for this post!
Could you share what model/value rectifier you used?

gmoon (author)  nathangill4 months ago

Sorry, I missed this... For the bridge? For this proj I think it's a pretty standard 1A 400V rectifier.

chadsmiff gmoon4 months ago

Where would I find a 1A 400V bridge rectifier? The closest I've found is a 3A 400V from Antique Electronic Supply https://www.tubesandmore.com/products/P-QBR-34

Thanks, this is a great post btw.

gmoon (author)  chadsmiff3 months ago

The 3A would work just fine. Otherwise, you can fabricate a bridge rectifier with four diodes (1N4004 or 1N4007)...


rinomazzucco11 months ago


I've read your brilliant article about the iso transformer upgrade but I'm a bit confused about applying the concepts to the new Kay 703C amp I bought off of ebay recently... Attached is the schemo... it appears that only a part of the filaments are not isolated...?!?! Am I getting this right? Any particular guidelines to apply the concepts and not end up frying some component or myself in the process? I'm not sure how to go about this especially relative to rectifying this circuit adequately... I know a fair bit about electronics but not as much as you that's clear...Thanks.

Ian W.

gmoon (author)  Alphascorpius671 year ago

That is one really strange schematic... There isn't any isolation for most of that amp.

The only thing that's isolated on that amp is the 18V filament for the preamp pentode. The only purpose I can think of for that transformer is that the preamp tube filaments draw a different amount of current than the other tubes, so they couldn't be used in series. The preamp voltage is still AC, even with the iso trans, so it doesn't have a DC supply for the preamp filament (which would be one reason to separate that heater from the rest). The tube filament voltages look perfect to run in series, so it's puzzling unless there are differing current requirements.

The wall AC runs in series through the iso transformer primary winding, the power tube, a voltage dropping resistor (R9) and half of the rectifier filament--where it splits off through the other half of the rectifier filament, and then becomes the B+ for both the signal tubes. Meaning the B+ voltage is wired in series with half of the rect heater, so they figured the current draw of the plates and sundry would balance with the draw of the half filament...

Very weird. Maybe the current draws for all the tubes are different, and they creatively figured a way with the transformer and half of one filament to make it work. Sometimes products are designed around an existing stock of tubes (like they got a truckload cheap).

I'd add an isolation transformer for the whole thing, right at the cord. It's too complicated and weird to change the rectification. Get a bigger iso transformer if you're worried about the half wave issue...

Doug, I recently picked up a Gregory Mark I for $10 at an estate sale and I've just started to dig into it. Mine is slightly different from yours in that it has a tone control and inputs for three (yes three!!!) instruments, and the construction is slightly different. The output transformer, for one, is mounted to the speaker, and mine has a different filter cap than yours. I don't have your education on these matters, but I believe it is called a multi-section tubular electrolytic capacitor. Do you have any advice on replacing these. Are new ones still made, or do I replace this with three separate capacitors that match the three respective sections? Also, do you think that this capacitor indicates that my amp is even older than yours (I could fine no date of construction)? I've attached some photos, and thanks in advance.
better chassis.JPGcapacitor.JPGChassis mark I.JPGfront mark I.JPG
gmoon (author)  northsidesoxfan1 year ago
Hey, that's pretty cool. Funny that the output wires are routed through a hole that was clearly meant for a tube, at one time (do you think it every held one?)

My Mark one originally had a multi-cap, but I replaced that with individual caps. Yeah, you can certainly buy them still, although that particular style will be hard to find (with wires instead of posts, and mounted thusly). You can try Antique electronic supply, for one. What they sell will be new, too. Ebay would have "vintage" caps like it, but I wouldn't bet on them being viable...

Some radio restorers like to hollow out the old multicaps and replace the guts with new electrolytics, which are quite small. I'm not sure that a Gregory Mark I will ever be a collectable, so it's up to you if you think it's worth the effort. Good luck!
Thanks for your response.

In regards to the multisection capacitor, I was going to replace it with individual caps, but I found a guy who will make a replica for $17 (http://hayseedhamfest.com/) plus $3 shipping, so I decided spend the extra few dollars and take the path of least resistance.

I found the empty tube slot funny as well, but I don't think this amp never had an extra tube. My guess is that to save production costs they used the same chassis for several different amps.

Where did you find a production date for yours? I don't see anything on mine.

In regards to value, I know it's not really worth anything. I bought it for $10 and I may spend more on parts than it's worth. I'm just using it as an educational tool and an entertaining hobby. If I get it working fairly well I'll probably swap out the speaker, so keeping it "original" is not an issue.

When I get to installing the isolation transformer, I may have a few more questions for you. Thanks.
gmoon (author)  northsidesoxfan1 year ago
The production date was stamped on the wooden frame of the cabinet.

I'm sure you're right about the chassis... I think they changed design on these when a new truckload of surplus parts came available...
Two more things.

1) Please excuse my typo.  I meant that I could not "find" a date of construction, not "fine" a day of construction.

2) I should have clarified that the amp has a sever hum, which leads me to believe that the capacitors should be the first thing to go.
slor2 years ago
For more than a few seconds, yes. Someone on a forum suggested wiring that particular transformer "backwards":

gmoon (author)  slor2 years ago
I found some info on your amp, and it's pretty similar--three tubes, nothing unusual. It should work.

If you're using "option A" you've added a fullwave rectifier, but not for the filaments. You should be well within the limits of the iso transformer. The Gregory is currently running the filaments on DC also, but it still works fine (and consumes more power with the DC heaters). The transformer gets a little hot, but not as hot as some of my vintage amp PTs.

I've never tried the "backwards wired" approach. It appears this lowers the voltage, and consequently the current draw. At a lower voltage it's probably more distorted--OK for guitar, but maybe not for a reverb driver...
slor2 years ago
Thanks so much for posting this! I implemented this on a surplus Graymark 509 to run the plate reverb in my studio. But even with "Option A" in place, the transformer runs hotter than I'd like. Any thoughts?
gmoon (author)  slor2 years ago
Sure, glad it helped.

How hot is hot? Is it too hot to touch?
Great Article!! I have a Gregory Mark V. Simiar to yours, but it has a tone knob and tremelo. Tubes are 50C5, 35W4, 12AU6 and 12AV6 (for the tremelo). I can easily install an isolation transformer, fuse and grounded AC cord. However, I need to figure out which transformer and fuse value. If I used a larger isolation transformer, could I keep the tube rectifier? Which fuse value would be best?
gmoon (author)  mysistersdiary3 years ago

Thanks! I upgraded the speaker on this one, and I use it all the time...

I think you'd do fine with a 100 VA transformer. A 0.5 amp fuse should work.
haz_mat4 years ago
Maybe I missed it, but what should the fuse be rated at?
gmoon (author)  haz_mat4 years ago
Depends on the amp. Use the lowest rated fuse that works.

To ballpark: amperage = watts / volts. A 40 watt amp (including 5% added for the iso transformer) is 40/120 = .333 amps. Might have to use a 0.5 amp fuse.

If the setup works witha .35A or a .25A fuse, then that's even better...
corbin5694 years ago
My amp uses 50eh5 tubes(pair) with no tube rectifier, only a single ss diode. Would I only need to replace it with bridge rectifier, or do something else??? thanks.
gmoon (author)  corbin5694 years ago
Since the SS diode doesn't have a filament, you can replace it completely with the bridge. No need to keep it there...
corbin569 gmoon4 years ago
Ok ,thanks for te help !!!
gmoon (author)  corbin5694 years ago

Just make sure the new ground comes from the bridge, not from the old wiring setup, which is the "neutral" side of the mains plug...
Dumb question maybe, but isn't a full wave rectifier enough isolation?

If you were to run the power through the rectifier first, and feed the filaments DC with a series resistor to drop the extra voltage, do you need the extra bulk and expense of the isolation transformer?
gmoon (author)  TechnoWombat4 years ago
Just for the heck, let's compare the numbers:

-- Ohms law-- current = watts / voltage.
-- Rectified voltage is about 170 V (120 * 1.4)
-- This isolation transformer limits at 50 watts (50 VA, but close enough).
-- a 1n4007 diode bridge has a 1A limit (but it may take time to fail).

Iso transformer draw: 50 watts / 170V = 0.294 Amp max
1N4007 draw: 1 Amp max
Other rectifying diodes may draw 3 amps or more...
But the purpose of an isolation transformer isn't to limit current using core saturation, it's to dereference the ground using inductive coupling, like you would use optoisolators in a MIDI interface. If you are earth grounded, and grab the live (hot) side of an isolated supply you should be OK, as the supply has it's own isolated ground, not referenced to earth ground

The issue is, that as we then earth (or ground) the chassis anyway, and the circuit ground is generally tied to that, we have just re-referenced our de-referenced ground anyway, which makes me wonder why the heck I'm using an isolation transformer.

According to the math above, if I were to use a 200VA transformer, the amp is suddenly dangerous again? There's got to be something we're missing here?

gmoon (author)  TechnoWombat4 years ago
You're not missing anything, but limiting current IS one of the reasons to use an iso transformer.

Don't get me wrong--enough current can pass through a 50VA iso transformer to do considerable damage. It's hardly a harmless amount of current. But there's a LOT more potential current in the mains...with or without a rectifier.

When people are often partly insulated by clothing, shoes, etc., it can make a huge difference.

Earth and neutral are NOT interchangeable. They are usually close in relative potential, but they are not the same. In fact, I bet that's one of the 'ground faults' that ELCBs, GFIs and RCDs trigger on...

Where are you planning on getting the earth reference for the chassis if an earth / neutral connection is by definition a "fault?" (neutral being connected to ground through a full-wave bridge rectifier, or directly with a half-wave rectifier.)

Also those devices also introduce a level of complexity, with their own set of faults-- they can fail themselves. Simply wiring one up incorrectly can render it useless.

Personally, I would never depend on one alone--especially not when I'm holding the signal ground in my hand. I might use a GFI together with a transformer, though.

Not to mention that you're assuming that all wall sockets are wired correctly... And that other people's equipment is safely wired and referenced.

Nothing is set in stone...witness newer switching power supplies which forgo isolation (although all have sophisticated current sensing circuitry). But how many of those power supplies are used in applications where the the user is literally grabbing on to one pole of high voltage?
gmoon (author)  TechnoWombat4 years ago
Not a dumb question at all... I wouldn't depend on a rectifier (diodes) to limit current safely.

Transformers limit current by their nature. If you try to pull too much current, the core saturates and that's the limit.

Diodes have a current / voltage limit, but they can exceed that for a time before they fail. Too long a time to protect from electrocution. Diodes can also fail "closed," which would limit NO current. Could still result in a "hot" chassis in the worst case, if only one or two diodes in a rectifier fail closed (and other safety changes weren't done).

Of course, both can fail "closed"...transformers can short internally. And that's why we change the other stuff... Like adding the fuse (and the earth ground). But a transformer short would prevent any current passing beyond that short.

A fuse can't react quick enough to prevent a serious (fatal) shock. It's there to prevent damage to the amp and act as a fault indicator.
Ah, OK, my thought was to add the earth connection, unit is already fused, and have the AC to the power switch, and then, straight to the rectifier.

Maybe I should drop in an ELCB prior to the fuse?
gmoon (author)  TechnoWombat4 years ago
I'm not certain about the safety and efficacy of ELCBs replacing transformers entirely. You'll have to research that on your own.

Most of the amps that are missing power transformers are pretty small, so a small iso transformer isn't very expensive... This 50VA model was only about $12 USD. I think that's pretty reasonable.
mjsdiy5 years ago
I have an amp with the same 3 tube configuration (the only I.D. anywhere on it is "Model 5000" on the front panel). I'm a tube noobie, but I want to be safe, so I installed the iso transformer and 3-prong cord with true earth ground. Now I have a 60 cycle hum that was not there before. Any ideas? Oh, and any help identifying the amp would be appreciated as well. Thanks for a great walk-through with loads of great info. mjs
gmoon (author)  mjsdiy5 years ago
Have you played with the AC wire routing at all? Maybe that has an effect. And has the amp been recapped lately? Nice amp, BTW. It REALLY looks similar to Gregory amps. Maybe it is one, but missing the badge on the baffle.
1-40 of 73Next »