Best effect loop circuitry for tube amp

[Thomeeque]

 Hi!

I'm planning to revitalize (well, practically completely rebuild) tube amp I've built many years ago. I have quite clear vision about preamp and power amp sections, but I don't know how to implement effect loop yet and I'm looking for some inspiration. Is there some tube amp well known for perfectly implemented effect loop, let's say etalon of how should be effect loop implemented, supporting everything what one could expect and degrading signal passing thru as little as possible? After all I won't probably need all possible features (bypass-able serial loop supporting standard guitar stompboxes will be probably all I'll need), but at this moment I'm open for everything

Thanks, T.

[kurtlives]

I am no expert on effects loops but I have built a few Dumble ODS clones as well as the external effects loop the Dumblator. I really like the Dumblator's design, simple and effective.

It does attenuate the highs a bit, this is actually a nice thing as it can really smooth out an amp. To compensate for the loose in highs there are "bright switches" though.

Anyways Id check it out, there is a schem on schematic heaven.

[fpaul]

Metroamp has a kit, kind of expensive ($90) but I'm guessing it is great because those guys are pretty passionate about tone.  Wish they posted the schematic but they don't.  The metroamp forum would be a good place to look for more info about other types.  I don't have one in my Marshall clone yet.  I've heard the simple ones can kill the tone and I'm too cheap to spend the 90.

[Ripthorn]

I would check out some Mesa schematics.  You could do a tube buffered effects loop.  Other than that, you could go unbuffered or do simple solid state (opamps) buffering as long as you have the headroom for it.  You might also consider something like the adjusticator from Geofex.

[Thomeeque]

 Hi guys, thanks for your replies!

I am no expert on effects loops but I have built a few Dumble ODS clones as well as the external effects loop the Dumblator. I really like the Dumblator's design, simple and effective.

It does attenuate the highs a bit, this is actually a nice thing as it can really smooth out an amp. To compensate for the loose in highs there are "bright switches" though.

Anyways Id check it out, there is a schem on schematic heaven.

1) I did check http://www.schematicheaven.com/dumble.htm, but none of ODS there contain FX loop..?

2) Dumblator really looks interesting, but where it goes when used? I mean, does it fit between preamp output and PI input?

Metroamp has a kit, kind of expensive ($90) but I'm guessing it is great because those guys are pretty passionate about tone.  Wish they posted the schematic but they don't.  The metroamp forum would be a good place to look for more info about other types.  I don't have one in my Marshall clone yet.  I've heard the simple ones can kill the tone and I'm too cheap to spend the 90.

Wow, that really drives my curiosity! According to their description it is FX loop I'm dreaming about But damn, it *is* expensive (plus final cost would probably get significantly higher on it's way from Michigan to my door) and they seem to be pretty careful to keep their secret - but I'll try to search their forum anyway, maybe I'll be lucky..

I would check out some Mesa schematics.

I'll do that!

You could do a tube buffered effects loop.  Other than that, you could go unbuffered or do simple solid state (opamps) buffering as long as you have the headroom for it.  You might also consider something like the adjusticator from Geofex.

Well, at this moment I'm more interested in some already proven solutions, I don't want to experiment too much in this case - but thanks anyway of course..

Cheers, T.

[captntasty]

Here's a link for a couple schematics - I can't say they are good or bad as I have not implemented any into my amps - soon...

http://www.blueguitar.org/schems.htm  not quite halfway down under misc amp mods - look for Kurt's FX loop.

[lunchie]

I made a little bit different effects loop into my hi octane, but here is nice schematics: http://www.freewebs.com/valvewizard/accf.html
I also included bypass switch to bypass entire effects loop circuitry.

I actually made a combination of cathode follower from this schematics for fx-send including variable level with 100k pot (entire schematics somehow didn't work, so I improvized a little bit )


and gain stage from schematics on valvevizard for fx-return and instead of 12au7 I am using 12ax7, which I feel works better for the sound.

[zombiwoof]

I'm pretty sure that Weber Speakers also has a kit for an add-on loop.  Check on their amp kits page for it. 

Al

[fpaul]

Did a web search and found diy info on the Kleinulator.  The Kleinulator is a 9v buffer/recovery box using j201 that plugs into a simple insert loop. Think I'll try it when I get some time.  Maybe not "the best" but probably "good enough for me".  We'll see...

[Thomeeque]

Did a web search and found diy info on the Kleinulator.  The Kleinulator is a 9v buffer/recovery box using j201 that plugs into a simple insert loop. Think I'll try it when I get some time.  Maybe not "the best" but probably "good enough for me".  We'll see...

Hi fpaul! Thanks! But I would not believe 9V circuit to do the job properly - you need like 80V peek-to-peek signals pass to power-amp input (=phase-inverter input) for full drive..

I'm pretty sure that Weber Speakers also has a kit for an add-on loop.  Check on their amp kits page for it.  

Al

Hi, Al! Thanks! Yes, they do - I did found it.. and it's the circuit lunchie did post here

But it's a bit strange circuitry I must say, even when I know only little of tube theory I would say, that output stage does not seem to give sufficient boost to signal going back from typical guitar stompbox (as it seems to be The AC-Coupled Cathode Follower and cathode follower cannot provide a voltage gain greater than 1) - as I say above, you need like 80V peek-to-peek signals pass to power-amp input for full drive, that is almost 100-times more than stompbox will give (maybe it's for some special FX units?).. and I'd say lunchie practically proved it..

I made a little bit different effects loop into my hi octane, but here is nice schematics: http://www.freewebs.com/valvewizard/accf.html
I also included bypass switch to bypass entire effects loop circuitry.

I actually made a combination of cathode follower from this schematics for fx-send including variable level with 100k pot (entire schematics somehow didn't work, so I improvized a little bit )
{img}https://taweber.powweb.com/store/el_schem.jpg{/img}

and gain stage from schematics on valvevizard for fx-return and instead of 12au7 I am using 12ax7, which I feel works better for the sound.

<czech>Ahoj Vlado, diky! </czech> Man, valvewizard page is great, thanks for this link! Please read my comment about Weber's loop above - would you agree?

Here's a link for a couple schematics - I can't say they are good or bad as I have not implemented any into my amps - soon...

http://www.blueguitar.org/schems.htm  not quite halfway down under misc amp mods - look for Kurt's FX loop.

Hi Patrick! Thanks! Hmm, I'd say that Kurt's FX loop is basically the loop lunchie did end with! So it's definitely interesting candidate!

Cheers, T.

[fpaul]

The Kleinulator layout I plan to use has a charge pump to 33volts.  Still a long way from 80 volts though.

[lunchie]

<czech>Neni zac, doufam, ze ti to pojede </czech>
I agree with the weber effects loop - I created it as described on schematics and it really didn't work - the recovery stage has to be gain stage and not cathode follower as it really doesn't have enough gain to drive power amp. So I immediately improvised and modified the recovery stage from cathode follower into gain stage from valve wizard page. It started to work, but then I realized that it's putting too strong signal into the FX-send - I added 100k pot to have variable fx-send signal strength. Also it seemed that 12au7 tube was not that great in that circuitry (it was kind of changing the sound of the amp - maybe because of the 12au7 I had on hand) and I tried 12ax7 and with and without the loop, the amp sounded same. Also thanks to 100k pot, the loop can be used as volume boost for leads - it feels really helpful and my cheap danelectro delay actually works in loop now .
12au7 takes too much current from the power supply (about 10mA per stage?) and it can drive the B+ for other pre-amp gain stages down - I took B+ from last pre-amp tube (can be taken also from phase inverter tubes). I opted for 12ax7 because of this fact, as 12ax7 takes about 3mA for both gain stages together.

Also I found the effects loop, though I haven't tried it.


EDIT:
I just checked out some FX-processors and they usually have -10dB (home recording) and +4dB (studio recording) level output - so it looks like weber fx loop is for studio levels ...

[aron]

For most people, if you really want to work on tube amps, the O'Connor books are a must. Filled with almost examples of loops, circuits, mods almost everything you can you use.

They are a little pricey, but IMO very worth it.

http://www.londonpower.com/tube-amp-books.htm

[PRR]

> I would not believe 9V circuit to do the job properly - you need like 80V peek-to-peek signals pass to power-amp input (=phase-inverter input) for full drive..

Oh, fah.

The power tubes do need like 80V peak-to-peak.

They are driven by a Driver (including a phase-splitter or inverter) with a gain of 15 to 50. So 2V to 5V is the most you need.

Kuehnel's The Fender Bassman 5F6-A has good discussion and calculation of Bassman internal signal levels. Since most large amps are "based on" the 5F6A, and most small amps are less of the same, this would be another good reference (out of several).

But "effect loops" are usually for connecting GUITAR effects. Stuff you could put in your guitar-amp cord. Guitar levels are 20mV-200V rms or a little more. Historic guitar-amp inputs mangle anything over 1V-1.5V peak. So 3V peak-peak, or well within the ability of 9V-batt systems.

Sometimes the "effects" are line-level. Hi-Fi stuff, or even Pro-Audio interfacing. Hi-Fi stuff never needs more than 2.8V peak. Pro gear may be specified to handle well over 10V peak, but for guitar dynamics it will be fine at 1V-2V peak.

So....

You are usually "looping" at some point much higher than guitar level, several stages along an amplification chain. Guitar pedals will overload. You must attenuate. You can do this passively. But guitar-amp stages are often quite high impedance, resulting in fairly large attenuator impedance. That should be fine for gitar-pedals, but some hi-fi gear has impedance of 22K or 10K. It's a tight budget. You usually want a buffer.

The amount of attenuation needed depends on where you tap the amp and the amp's gain structure, and its level structure for the specific player. You usually want some way to trim the attenuation.

Then however much you attenuate to get out to the loop devices, you want to gain-up to get back to where you inserted. This stage must be able to drive the driver, which is not 80V, but may be 2V-5V p-p, and the upper end of this is near the limit of simple 9V power (especially battery).

And of course you don't build a battery-powered buffer/booster into a wall-powered amp. But several mA at 10V-30V is an awkward thing to get in a 300V-400V amp that only has a few mA of clean DC. You must waste 90% of your voltage, and you may end up tapping the "dirty" DC at power stage to avoid sagging your preamp stages.

Once you define the basic goal, then it is "just implementation". One 12AU7, half to buffer the attenuator and half to gain-up the return. Or a dual op-amp chip, with some way to power it, and protection against the >36V signals and thumps which may get into it. Or the "12AU7" plan with a couple $1 500V MOSFETs, which can drink near-raw B+ and throw-off the excess as heat. Or specific cases: a cathode-bias output stage has 20V-30V clean DC handy which can be tapped to feed a TL072. A channel may have a loss-network already, which avoids some of the down/up issues. Some amps like Ampeg VT40 take a lot of gain in the preamp, run downhill through tone controls until they get to near guitar level again, and a high-gain output stage: this may be very easy to tap.

And -then- you have to think about bypassing it for straight sound. Are the buffer/booster stages perfectly transparent in your specific amplifier? Or do you cut them out? If the buffer loading is absolutely negligible, you may only have to switch the Return; else you'll want to disconnect the Send buffer also. Switches can't be remote controlled. Relays clack and eat power. CMOS may not handle the higher signal voltages post-booster. Opto-cells do not have infinite off/on ratios, and do need several mA at quite low voltage, and are slow (which can be good).

[Lurco]

http://www.diystompboxes.com/smfforum/index.php?topic=80821.0 

[Thomeeque]

> I would not believe 9V circuit to do the job properly - you need like 80V peek-to-peek signals pass to power-amp input (=phase-inverter input) for full drive..

Oh, fah.

The power tubes do need like 80V peak-to-peak.

They are driven by a Driver (including a phase-splitter or inverter) with a gain of 15 to 50. So 2V to 5V is the most you need.
...

Hi Paul!

Thanks a lot for this post, it brings lot of light here!

I have only one counterargument at this point and it would be that you may want to overdrive PI and for this you need proper headroom.. Please, see these two posts:

http://forum.metroamp.com/viewtopic.php?p=257587#p257587
http://forum.metroamp.com/viewtopic.php?p=259874#p259874

I admit that this is not limitation for everybody but on the other hand it is something to keep in mind IMO.

http://www.diystompboxes.com/smfforum/index.php?topic=80821.0 

Hi Lurco! What's the question?

T.

[PRR]

> you may want to overdrive PI

So why must it be next to the driver? Do your gain/level diagram and put it where you think is best.

Here is a famous-brand amp with an effects-loop which seems to incorporate "good levels". Note that there are two gain-stages between the effects-return and the power section driver. And yes, I grant that this example is WAY over-complicated. Altho if you ignore all those relays and the V3 loop-tube, it is really a pretty ordinary amp.

[lunchie]

> you may want to overdrive PI

So why must it be next to the driver? Do your gain/level diagram and put it where you think is best.

Here is a famous-brand amp with an effects-loop which seems to incorporate "good levels". Note that there are two gain-stages between the effects-return and the power section driver. And yes, I grant that this example is WAY over-complicated. Altho if you ignore all those relays and the V3 loop-tube, it is really a pretty ordinary amp.

If I am reading the schematics correctly, the effects loop precedes the reverb tank driver, which is done from 2 paralel triodes and reverb recovery and than there is 1 gain stage most likely to drive phase inverter. IMHO it's standard effects loop design positioned right after pre-amp and before reverb.
I can't see any point in having effects loop before any pre-amp tubes participating in distorting the signal.

[sean k]

The valve wizard is my new favourite guy for shooting from the hip ways to do tubes/valves. I think his explanation of why the Soldano dual triode with a volt amp with plate to grid direct connected cathode follower is the bees knees  and the kinda pure poetry you just gotta believe is gonna do what you really couldn't figure out you actually wanted!

[Thomeeque]

 Hello, I'm back!

After almost endless searching, thinking, consulting, deciding and reverse-deciding I've come to state, where I'm trying to design my own loop circuitry to suit my needs best..

My amp will be basically 50W Marshall JCM800 Master Volume (2204) clone with few mods (two "virtual" channels, boost switch, post-PI Master Volume, FX loop and build-in reverb).

Here's a block diagram of future amp:



My goals for what I call FX BLOCK (FX loop + build-in reverb) are:

• topologically placed directly in front of PI (after CHANNEL VOLUME controls)
• no modifications of original circuitry of amp besides true-bypass switch
• reverb after FX loop return
• IN/OUT signal level range up-to the range mentioned by MetroAmp's guru SDM = around 80V peak to peak.
• standard guitar stompbox levels for the loop
• FX loop bypass footswitch (for reverb I don't need footswitch - I use one constant level of reverb almost all the time)
• only as much active stages and only as much gain as needed
• only as much controls as needed (but goals above are more important)
• SEND functional even "standalone" as a preamp out

I've came with this:



(REVERB block will be something "standard", no issue here.. hopefully )

Since probably all FX units I'll be connecting to this amp will be solid state and since I didn't want to use valve-based reverb driver, I've decided to make whole FX BLOCK solid state - even I really love tubes I don't want to add tube where solid state technology can do same or better job. For puristic moods there will be PURE MODE switch.

Please, check it and let me know what you think, I'd really appreciate it!

Stage F is my problem #1 now, what I have drawn does not "work" correctly when I try to emulate it in LTC, but it's not big surprise as I'm absolute MOSFET beginner It's more like "placeholder" for now..

Thanks!

Cheers, T.