Looking for feedback on Transformer Saturation Box schematic

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
On the note of clipping. Say if I was going to use an LED instead of regular diode, would I still need to use a current limiting resistor?
 
Just wire them as in your schematic. The output of the transformer exists between the two pins of the secondary - one is NOT inverted relative to the other they just float. Just connect one of them to gound and use the other as the output signal.

Cheers

Ian
 
I had another go at this this morning, here’s a quick Audio demo A/Bing bypassed and affected with some different synth sounds. Things are set up pretty much like in my schematic.

Essentially it’s TL072 adding some gain > 10k pot > 600:6.25k Transformer > 100k pot > TL072 as a buffer > 10k:10k transformer.

From my experiments yesterday I suspect most of the change in sound is coming from clipping the TL072. I would have used a 10k pot instead of the transformer but I only have one currently. I think I probably need to reduce the gain a bit at the input since I cant have the 100k pot fully open without it sounding nasty.

 
Another quick test. 2nd TL072 bypassed completely. First sound is dry, second sound just the TL072, third is with transformers. Tried to roughly match the levels to make the test a bit more fair.

 
I just wanted to share the schematic on here to see if anyone can spot anything clearly wrong with my design or has any suggestions.

Some comments. First, I assume this is for e-guitar and ~ 100mV nominal and ~ 400mVbefore hard clipping sets in.

Driving a transformer with low impedance will limit HD at normal levels and cause a fairly sharp magnetic overload.

What I would suggest is to use a 620 Ohm series resistor to drive the 600Ohm primary and load the transformer secondary with an inverting Op-Amp set up as active attenuator with a 6K2 resistor loading the secondary, wire transformer to invert polarity to offset this inversion.

This will give a much more soft distortion/compression from the transformer without causing excessive bandwidth limitation.

The driving Op-Amp should have a gain setting ideally with a stepped attenuator from unity gain to whatever you think you will need to make the transformer scream "no more". With +/-18V rails and rail-to-rail Op-Amp's we can drive around 12V into 600 Ohm. That should be plenty.

With 400mV input before hard overload of the input Op-Amp maximum gain is ~ 30 (30dB)

Maybe 3dB steps and a 2-deck 11 position switch, giving 0/3/6/9/12/15/18/21/24/27/30dB Gain in the frontend.

The feedback resistor for the output Op-Amp should be switched along with a second deck matching the gain setting in reverse, to keep gain (excluding compression effects from the transformer) input to output at unity.

As the Transformer is 600:6250 and loaded at 6.2k net gain will be 1.613, so if the input gain is unity, the output Op-Amp should operate at a gain of of 0.62, so 6.2k input resistor and 3.9k feedback resistor (may need adjusting for transformer DC loss).

Bypass switch simply uses first Op-Amp as unity gain buffer into the output transformer (if used), so needs tow poles 2 throw (DPDT).

Voila, now you will hear transformer overload, pretty much solely if your Op-Amp's are clean enough. TI OPA2156 or OPA1656 (or OPA1652) would be my choice. And you will be able to turn up the transformer drive while keeping the output level the same.

I think it's an interesting option. I'd be tempted to add a tube driving the transformer and wire it 6.25k:600 instead.

Thor
 
Some comments. First, I assume this is for e-guitar and ~ 100mV nominal and ~ 400mVbefore hard clipping sets in.

Driving a transformer with low impedance will limit HD at normal levels and cause a fairly sharp magnetic overload.

What I would suggest is to use a 620 Ohm series resistor to drive the 600Ohm primary and load the transformer secondary with an inverting Op-Amp set up as active attenuator with a 6K2 resistor loading the secondary, wire transformer to invert polarity to offset this inversion.

This will give a much more soft distortion/compression from the transformer without causing excessive bandwidth limitation.

The driving Op-Amp should have a gain setting ideally with a stepped attenuator from unity gain to whatever you think you will need to make the transformer scream "no more". With +/-18V rails and rail-to-rail Op-Amp's we can drive around 12V into 600 Ohm. That should be plenty.

With 400mV input before hard overload of the input Op-Amp maximum gain is ~ 30 (30dB)

Maybe 3dB steps and a 2-deck 11 position switch, giving 0/3/6/9/12/15/18/21/24/27/30dB Gain in the frontend.

The feedback resistor for the output Op-Amp should be switched along with a second deck matching the gain setting in reverse, to keep gain (excluding compression effects from the transformer) input to output at unity.

As the Transformer is 600:6250 and loaded at 6.2k net gain will be 1.613, so if the input gain is unity, the output Op-Amp should operate at a gain of of 0.62, so 6.2k input resistor and 3.9k feedback resistor (may need adjusting for transformer DC loss).

Bypass switch simply uses first Op-Amp as unity gain buffer into the output transformer (if used), so needs tow poles 2 throw (DPDT).

Voila, now you will hear transformer overload, pretty much solely if your Op-Amp's are clean enough. TI OPA2156 or OPA1656 (or OPA1652) would be my choice. And you will be able to turn up the transformer drive while keeping the output level the same.

I think it's an interesting option. I'd be tempted to add a tube driving the transformer and wire it 6.25k:600 instead.

Thor

Interesting thoughts here thank you, I’ll have to come back to your post when I have the time to digest what you’ve said properly.

I’m using it with Eurorack which is 10vpp. Eventually I’d want to adapt the circuit to work with line and maybe instrument levels as well.

I’d love to use tubes but the high voltage scares me. Maybe in the future.
 
Interesting thoughts here thank you, I’ll have to come back to your post when I have the time to digest what you’ve said properly.

If I find time, I'll draw it out.

I’m using it with Eurorack which is 10vpp.

10V PP = 3.53V RMS = 13dBu

That needs ~ +/- 9V rails, or could use higher rails (+/-18V ?) and a step-down output Transformer 2:1.

Eventually I’d want to adapt the circuit to work with line and maybe instrument levels as well.

+13dB is not bad for line levels
Instrument levels depend on the instrument (surprise, surprise)

I’d love to use tubes but the high voltage scares me. Maybe in the future.

The definition of "Safe Extra Low Voltage" is < 120V DC Ripple Free. Such a voltage is considered "safe" by regulators.

Back in the early 00's I designed a simple tube module with ECC88/6DJ8/6922/6N23 etc. with two channels, that basically worked under SELV directives in place then (it would need updating now).

It was designed for DIY use and in effect contained two tube Amplification section with up 30dB Gain and jumpers for configuration in a number of fixed gains.

It used a solid state anode load that also offered a "buffered" output, 90V HT generated with a doubler and powered by a custom mains transformer.

1681411207444.png

1681411375219.png

Some examples of use.

1681411627029.png

While it was never picked up by the Pro DIY crowd (Why? Too Expensive?) after adding a Mic-Transformer, Level Controls and a 4:1 output transformer this would have made a swell and EZ Tube mic pre.

This is now unavailable, but shows tubes need not be scary.

Thor
 
Last edited:
@Thor, you schematic seems to show a 90VAC power input which exceeds the 50VAC allowed by SELV or have I missed something?

I would change this now to use ~45V AC and a Voltage doubler to get the HT to be compliant and I noted that in my post. As said, this is long out of production (the last batch was made in 2010 IIRC).

Actually, these days I'd use a switcher to make 110V/10mA from 5V or 12V DC and another for heaters.

My point was that it is possible to run many tubes within SELV safety limits without starving the tubes.

Thor
 
I’m not sure I follow, are you saying I can use a voltage that’s considered safe say 48VDC and below then step that up in the circuit? Surely that still means you’re dealing with high voltage? Or does the fact that you’re not plugging straight into the mains mean you’re cutting down the risks?

Also on a similar note, what’s the standard way of prototyping high voltage circuits, using a breadboard seems a bit dodgy 😅
 
I’m not sure I follow, are you saying I can use a voltage that’s considered safe say 48VDC and below then step that up in the circuit? Surely that still means you’re dealing with high voltage?

First, 120V DC (no ripple) are legally and for agency compliance "(safe) extra low voltage".

A traditional tube circuit uses a resistor as "load". Commonly 1/2 or more of the powersupply voltage is burned up in this resistor.

So in (say) a circuit with 250V nominal +B the tube's anode voltage is often below 100V.

At the same time we usually see fairly moderate signal levels at the circuit output.

If we replace the resistor with solid state circuit that (for arguments sake) operates from safe 120V and places 90V on the tube anode, the tube operates basically the same as in the 250V circuit.

The solid state actually infers some additional benefits, including a very low impedance output which, depending on precise design can be made to drive 600Ohm directly and with ~ 60V P-P signal.

All that with a safe 120V DC supply voltage.

Mind you, some so-called "high perveance" Tubes can operate very well at 48V.

There is a white paper by Apex on a suitable circuit, mind you, I'd not use a 12AT7 in this circuit myself, but ECC88 or similar.

IMG_20230415_050527.jpg

From here:

Aphex Tubessence White Paper

I have seen an earlier circuit from Aphex (IIRC) that used just a TO220 PNP Bipolar transistor and the tube heater as ballast for the transistor.

Like so (Aphex Patent):

IMG_20230415_051600.jpg

I thought this circuit very interesting but never got around to trying.

I have designed products around variations and refinements of "Tubessence" which were well received.

E.g. this:

iFi iTube | Inner Magazines

Also on a similar note, what’s the standard way of prototyping high voltage circuits, using a breadboard seems a bit dodgy 😅

Terminal strips. Best do some research on the subject.

Thor
 
Last edited:
And when testing keep one hand in your pocket at all times.

Actually, best use isolation transformers everywhere.

In my Lab Setup there is a way to (safely) lift all earth. I have literally touched 500V rails on that bench and felt nothing.

It was mainly implemented to protect the lab gear, not myself. I seem impervious to electricity, give or take.

I still have the scars on my hand from where a 100W Marshall Head HT arced through dirt and my hand. Quite nasty burns and not superficial, but all through the meat in my hand.

Also had a fair few industrial accidents with old Lecky that saw me send to hospital and observed for 12 hours.

Can't say I enjoy electric shocks, but there is worse. Up to now, not much damage really.

Still, wear rubber sole shoes, marigolds (rubber gloves) if needed and never have both hands on the device or circuit you are testing AND avoid moisture as well as dust build-up are excellent safety suggestions.

And keeping voltages below 120V DC or 120V P-P AC is another.

Thor
 
I’m guessing you would use something like this to get 120VDC?

https://www.mouser.co.uk/datasheet/2/468/Rxx_B-1709947.pdf
What I can’t seem to work out is why there’s two pins for everything? It says it’s single output but is that just referring to the fact there’s no negative voltage?

When you say terminal strips, do you mean to use them in conjunction with soldered breadboard/strip board etc?
 

Ooh man this all sounds very exciting but I have no chance of deciphering what’s going on here. I’ve read a bit about tubes in the past and know a couple sources that I’m able to follow if I wanted to put together a standard, basic tube gain stage, but a lot of what’s going on here is going over my head.

Edit: I’ve found the patent and I think things are starting to click, sorry expect more questions later 😬
 
Last edited:
I annotated one of the Figures from the patent based on what I think was explained. Think it's starting to click. Very exciting, I've wanted to make a tube channel strip for ages but seemed like something way out of my depth.
 

Attachments

  • Reflected Plate Amplifier Fig 4.pdf
    436.4 KB
Back
Top