ODS 124 - revisited for the 21C

[Stephen1966]

Excellent information on the reverb phase, makes sense. If you want to keep your dry and wet signal in phase, there's another trick to do that job: replacing the mixer triode with a cathode follower. That also means you don't need the voltage divider in the dry signal path anymore.

But still, I've heard some clips of a #60 clone and that reverb sounds simply delicious. I highly recommend to take a look at them (I do not have al link though). If you allow me to pollute this thread one more time with a schematic...

Still playing catch-up here... Here is the implementation schematic for the #060 type reverb.

ODS RVB 5.0.pdf

Pay no attention to my voltages, all that has to be figured out later.

Stephen

[Stephen1966]

Haha yes those descriptions! Honestly I'm not sure what to expect from the next twi ideas. The follower has some advantages, but are they big enough? I expect the follower to stay clean no matter what you throw at it. It should be able to handle the maximum voltage v1b is able to put out. It is not a follower in the way the one in a Bassman or jtm45 is.

The other idea is fun too I think. Ive never seen a cathodyne splitter used like this before (and there's probably a good reason for that). I thought I might share it, maybe you guys have seen this idea before or have experience with it.

So in this three tube design we have a single CF (heading in the direction of the #002). I'm warming up to the idea of using a switch to create a virtual earth before the reverb driver, to switch the reverb on and off, but in this design with the reverb off, the signal from the preamp goes through the CF on the way to the Master (U50?). For the moment, I have bypassed it entirely but I can see the advantages of keeping the CF in the circuit: unity gain (practically speaking), linear frequency response, very low harmonic distortion and the only problem of ensuring the stage provides enough headroom. When we are in a heavily overdriven situation, we might also rely on the soft clipping and compression the CF provides. In real terms, I'm not sure this would be too noticeable... certainly not by that drunken girl barely able to stand in the front row But I wonder if this is going to take us out of the tonal landscape of the 124 to another place... Only one way to find out I guess

This is likely to change but here is the implementation schematic for this design.

ODS RVB 6.0.pdf

Stephen

[rootz]

Will look into it carefully after my studio work today...

[Stephen1966]

Yeah, that is in the direction of what I meant, but not quite correct yet.

Follow my corrected schematic as attached and add the pots and relay like in my drawing. Maybe I’ll have some time left tonight to update my schematics in a more clear way. For now this will have to do.

This is a follow on from https://ampgarage.com/forum/viewtopic.p ... 98#p427198 (see above).

So in this one, I've pieced together the notes you made for the dual CF design using a dual concentric pot for the OD and CL levels. That is perhaps where we hit a stumbling block because I couldn't find a dual 250k + 250k pot. The nearest I could find was this https://www.gear4music.cz/cs/Kytara-bas ... gJsb_D_BwE. Maybe you might know of a supplier who can provide them. All is not lost, however. The CTS pots have a removable phenolic trace so it is perhaps possible to use one of these and replace the 500k trace with a 250k from another CTS pot. They appear to be the same kind and size of phenolic plate. I don't know if you could do this with any other brand of pot.

I've gone all in on the idea of using a SPDT relay as an on/off switch for the reverb and the second G6E relay is (I hope) wired correctly in parallel with the OD relay so that when that fires up, so should the CL/OD relay. The wipers of the concentric pot should probably be connected to the opposite pins (RET CL to nc1 = the off position and RET OD to no1 = the on position). I'm still having trouble seeing how this concentric pot would work to balance the OD and clean levels. If lug 1 on the RET OD pot is connected to no2 on the OD Relay, the RET OD pot could receive a signal from the OD out, creating a second "return" pot in series with the RATIO at the end of the OD signal chain. That would make sense, but it could only attenuate the signal from there. Alongside the RET CL I can see the advantage of balancing the signals, but in that case, why not use a single pot for the OD (or clean) level? Or, use a stereo 250k blend pot like this https://www.tube-town.net/ttstore/alpha ... 16-mm.html. The level still has a degree of control with the SEND pot at the input of the reverb driver. In any case, the way I have it figured out here, may not be correct - the connection with the OD relay would inject a dry overdrive signal in parallel with the wet OD signal after the tank. That might be no big deal, but I could still have this thing entirely back-to-front. It isn't shown in my schematic for that reason.

There was also - if I understood it well - the question of whether the DC voltage hitting the grid of the first CF triode was too high. I'm not sure if this is meant to be some necessary DC coupling and the 22n cap and pair of 1M resistors are meant to tame that, but if it is only the AC component of the voltage we are interested in, could we just use a small, DC blocking cap (.005uF) in place of the net? I would really expect the grid and cathode voltages of the CF to be very high, but not plate load so. Anyway, here is the next schematic...

ODS RVB 7.0.pdf

Oh! what fun

Stephen

[Stephen1966]

Will look into it carefully after my studio work today...

You lucky lucky devil

[Stephen1966]

Thinking about that last design (RVB 7.0) I understand a little better that it requires a DC coupling with the plate voltage of the earlier gain stage - V3a in my schematic (sorry, the numbers are confused - but consistently). It seems to me that the anode to grid voltage should probably be in the region of 149VDC as simulated.

HT (B+5) = 365VDC
Av (V3a) = 216VDC
HT - Av = Vga = 149VDC

I'm guessing the bias point would be around -1V

So, we could increase the plate load resistor (V3a) and the cathode resistor (V1a) to make the load line less steep. We might also increase or decrease (I'm not sure) the cathode resistor of the preceding stage, though at 1k it seems you may have already thought of that. We would still have a hot bias in the CF so we should use a grid stopper to prevent oscillation and the grid leak to prevent arcing.

Another step might be to use a tube in the textbook sense, with higher anode current capabilities: 12AU7

The maths is all new to me, so if I have misunderstood, please correct me...

Stephen

[rootz]

Hi Stephen, I had kind of a busy week, so not much posting here. I now have kind of an quiet weekend, so here we go again

You are correct I adjusted the cathode resistor of the triode in front of the CF. I'd like 820 Ohm even better. It is just to keep the plate voltage under control and go for more centre bias.
This is important, as Vkf (the allowed voltage between heater and cathode) is 180V for an ECC83.

There are some possible solutions to limit Vfk: 1) using a voltage divider with the top resistor bypassed by a cap after the gain stage (cap is mandatory, but a necessity if you want to preserve full AC signal swing), 2) adjusting plate and cathode resistor values (higher plate value, lower cathode value --> biased a bit hotter but a lower quiescent plate voltage), 3) heater elevation to e.g. +50 V or +75V DC.

Again, there is no free lunch: 1) needs more parts and includes an extra cap in a feedback loop. Probably won't lead to stability issues here, but might when there are more components like caps and transformers in the feedback loop. Not sure if it alters the clipping behaviour in a noticeable way, 2) lowers the headroom of the preceding triode a bit, is the way it is done in e.g. the #002 and many other amps though, 3) requires some extra parts, but an easy way to get to safe Vfk values for all tubes. Possibly lowers noise/hum from coupling between filament and cathode a bit too in an ODS. You do have to keep an eye on the maximum resistance between ground and the heaters. 100k is probably safe for the preamp tubes. Not sure if it is for 6L6 tubes though and can't seem to find info about it in the datasheets.

I would not select a different tube, e.g. a 12AU7. The big downside is you lose a lot of gain, which would seriously limit the reverb capabilities. You do get a lot more headroom on the input side of things though. But no free lunch here either (it just doesn't exist, does it?).

I noticed there is a design flaw in our most recent schematics. The CF's have a different bias point and thus respective different cathode voltages. I don't like that and will correct that in an updated schematic. I also don't like that the CF's are unprotected agains high start up voltage on the grids: full B+ with a ground potential cathode when the tube is cold and you don't use the standby switch properly. Easily corrected with some cheap extra parts, which you can probably place on the tube socket. In the following post I will add some new schematics.

[Stephen1966]

Hey rootz, I figured you were busy, me too! Sure, a lot to digest here, at the moment I am working on the first design (after the 060 - "RVB 5.0") This is the one with a heap of photos and the hand-drawn schematic which I am working through at the moment in order to get the layout. The plan is to build prototype boards for each of the three reverb circuits - the 060/the single CF/the dual CF. 060 just seems like a better place to start since it's relatively well documented. Most of this week has been about sourcing parts - I think I will go blind if I have to read another datasheet

When i get back to the dual CF design, I'm leaning towards using a couple of 12AX7s and a 12AT7 for the driver, for all the designs, in fact. Adding a few parts here and there won't be a problem when I'm making my own boards. If we raise the heater voltage are you saying it has an effect across all the tubes? I read about this but can't recall it right now.

I've also figured out the power supply for first reverb. My transformer should put out 447V. I only have nominal values since who knows where it will go in real life but I have the five standard nodes (not counting the sixth to the FET for the moment)...

• B+1 427V

• B+2 423V

• B+3 412V

• B+4 313V

• B+5 306V

In order to bring it into the ranges in your simulation, I'm adding three dropping resistors on B+3. The supply is pretty well filtered here so I can probably get away without using another filter cap. But if I did, 22uF should do it.

By my calculations then I have three HT voltages to work with:

• B+3.1 with 2.7k giving 385V for the reverb transformer (Hammond 1750A) and driver tube

• B+3.2 with 39k giving 338V for the recovery tube

• B+3.3 with 11k giving 325V for the send and return tube (first and last triodes in the circuit)

That's with the pair of 12AX7s and 12AT7 plate currents (1.2mA and 10mA).

The voltages in your simulation are somewhat higher, so either my maths are bad or you plugged in a PT with a higher secondary voltage (probably both ) mine - a Hammond 290FEX - puts out 320V-0V-320V across the HT rail.

Do you see any problem if I shoot for the voltages you simulated, given my lower HT voltages elsewhere? I am curious, because after the load resistors the voltages with the 12AX7s all come in safely under 250V

Stephen

[rootz]

Yes, elevation will affect all tubes, unless you would use a separate transformer to power the preamp heaters for example. Elevating 40 to 50 V is perfectly safe, cathode bias for the power tubes would have the same effect in that respect.

My voltages are off and based on a power transformer from a Peavey Deuce. Standard is a Fender Twin transformer, which would have around 440V on the plates. I includes an updated (almost) stock #60 schematic. It might differ a bit from versions on this forum, but mostly on details.

I can't see how you've drawn out the power supply. The #60 PSU is more like a SSS #002: the reverb B+ nodes are in series with all the rest. This is an easy way to work if you use a dual cap can for the clean and OD preamp. What you write suggests you have a string for the reverb parallel to the rest of the amp, am I correct? Also possible, but then you'd still need a cap per node to separate them at an AC level.

There are some design goals for the PSU I think:
- the reverb driver tube should see something like 365V, like in a SSS;
- I'd like the reverb triode voltages, especially the mixer, 10 to 20V above the preamp: a bit more headroom (preamp clips first);
- the other voltages should match the #124 (that's what you were aiming for, right?) voltages within say 5%. I think clean and OD pre voltages are most important. Make them to high and the amp will stay too clean, make them to low and thing will get a bit compressed like in a Wonderland. 190 to 200V for the clean preamp plates is the goal IIRC.

[rootz]

In short: your PSU design needs more work. The 12AX7 tubes in the preamp draw approx 1.3mA, PI a bit less. The reverb drivers draw around 1.5mA.

Also: have a look again at the #60 layout I posted (page 14?). That one looks good, but lacks a bit of the PSU. There is no cap can in the layout and it should be there. It would save you a load of time laying out the board though. I hope and think it can easily be adapted for a more modular design where you can experiment with different reverb designs.

[Stephen1966]

In short: your PSU design needs more work. The 12AX7 tubes in the preamp draw approx 1.3mA, PI a bit less. The reverb drivers draw around 1.5mA.

Also: have a look again at the #60 layout I posted (page 14?). That one looks good, but lacks a bit of the PSU. There is no cap can in the layout and it should be there. It would save you a load of time laying out the board though. I hope and think it can easily be adapted for a more modular design where you can experiment with different reverb designs.

I've sent a PM with the work in progress, as you say it still has a few kinks that need to be worked out but you can see the power design there. In parallel, yes, that was the thinking. However, I'm not averse to redesigning it along the lines of the SSS. My head is still in that horrible drawing of the 060... making some progress though...

Stephen

[rootz]

Can't remember a drawing of the 60 Stephen. Can you post that one here? I do remember a drawing of the 13, but that is a whole other amp.

I replied to your PM's by the way. I prefer to keep most of the work in the thread, so others can look at and learn from it too though. But, whatever you are comfortable with.

[rootz]

Attached an updated design for a dual CF mixer design. I adjusted some resistor values to match safer cathode voltages for the CF's (Dumble got away with it in the SSS, Marshall in many of the amps with a CF driven tone stack). Also added some protection to the wet mixer CF. The double reverb controls are not mandatory, you could replace those with one 100k control, either log or linear.

Let's call this variant: 'full CF mixer 3 triode reverb'. Makes referring to it a bit easier.

[rootz]

An alternative to the 'full CF mixer 3 triode reverb' would be this attached schematic. I replaced the CF with a plate follower. I doubt you'd ever hear much difference to a CF. But who know, I don't exactly have bat ears anymore.

Let's name this one 'PFCF mixer 3 triode reverb'.

[Stephen1966]

Can't remember a drawing of the 60 Stephen. Can you post that one here? I do remember a drawing of the 13, but that is a whole other amp.

I replied to your PM's by the way. I prefer to keep most of the work in the thread, so others can look at and learn from it too though. But, whatever you are comfortable with.

I see your point - I don't embarrass easily but others are following this thread and I wouldn't want anyone thinking this was the finished article. I'll repost with HUGE disclaimers