Neil Young's "Magic" 5E3

[Mike K]

His typical live electric sound stars about 1m30s in. There is quite a bit more to it than the Deluxe. Of course the other guitar gives him lots of leeway to go off on tangents.

Sounds pretty killer for a flubby ole 5E3.

[Helmholtz]

I get both below the knee, actually. It seems like you put in the full Ra-a for PP calculations, not half.

The calculator requires inserting the full Raa.
From the chart you will find that with an Raa of 8k the loadline for the single tube corresponds to 4k in class A range and to 2k in class B range above.

...the DeArmond I estimate would be about 320/290.

Full load voltages will be lower than that (cathode voltage and screen current increase). I used 310/260

[Mike K]

The calculator requires inserting the full Raa.
From the chart you will find that with an Raa of 8k the loadline for the single tube corresponds to 4k in class A range and to 2k in class B range above.



Full load voltages will be lower than that (cathode voltage and screen current increase). I used 310/260

Yes, that's what I did - I get both knees lower.

Yeah, that's true. I didn't estimate any PSU sag. I have models for the full amp and PSU. I know the values for the 5E3 based on actual measurements. The modelled 6V6s don't correlate exactly with real measurements, but it's pretty close. I don't know the DCR on the DeArmond PT, but I can estimate it based on known 5E3 measurements assuming the lower voltage.

For some reason, and maybe I should ask this elsewhere but I've never been able to model AB in LTspice. It doesn't recognize the cutoff and adjust the reflected impedance. I typically run in "A" and in "B" by manually adjusting the impedance ratio and interpolate between the two.

[Helmholtz]

Yes, that's what I did - I get both knees lower.

.

You might be fooled by the upper plate curve which corresponds to Vgk= +20V! (Don't ask me, why.)
The Vgk= 0 curve is somewhere in the middle. Use your mouse pointer to identify.

[Mike K]

You might be fooled by the upper plate curve which corresponds to Vgk= +20V! (Don't ask me, why.)
The Vgk= 0 curve is somewhere in the middle. Use your mouse pointer to identify.

Yeah! That would fool me!

[Helmholtz]

For some reason, and maybe I should ask this elsewhere but I've never been able to model AB in LTspice. It doesn't recognize the cutoff and adjust the reflected impedance. I typically run in "A" and in "B" by manually adjusting the impedance ratio and interpolate between the two.

I'm not an LTSpice expert but might be due to transformer model.

[Mike K]

I'm not an LTSpice expert but might be due to transformer model.

Yeah, it is. There's probably a command you can put in to couple them in that way, but they don't recognize it automatically.

I see your result now. That's a good tool, but the positive grid curves definitely threw me off (not all tubes have that). I see the correct power calculation now as well. Thanks!

[Mick Bailey]

The last 5E3 I built used a Hammond 291BEX PT and a 1760H OT - both intended for a Deluxe Reverb. B+ was 480v on my mains. I used 470R flameproof screen resistors and fitted JJ6V6S tubes, which have some differences to a typical 6V6. The speaker is a 102dB Eminence Cannabis Rex. This all works really well for live playing and the customer wanted an amp that had a little more clean headroom than a stock 5E3. The downside is that for low volume playing it needs a push with an overdrive pedal to get it to break up.

[Chuck H]

The last 5E3 I built used a Hammond 291BEX PT and a 1760H OT - both intended for a Deluxe Reverb. B+ was 480v on my mains. I used 470R flameproof screen resistors and fitted JJ6V6S tubes, which have some differences to a typical 6V6. The speaker is a 102dB Eminence Cannabis Rex. This all works really well for live playing and the customer wanted an amp that had a little more clean headroom than a stock 5E3. The downside is that for low volume playing it needs a push with an overdrive pedal to get it to break up.

Do you remember the bias current? Cathode resistor value?

No doubting the amp sounds great but I wonder if it's still a "5E3" at that higher voltage. Since the 5E3 vibe overall is somewhat about the lower voltage and running the power tubes at higher current. I had a Cannabis Rex loaded into a similar sized combo for a long time. I liked it a lot but it was at it's best right out of the box. It lost a little too much high end for that particular amp when well broken in. Great speaker. I'd use it again in a brighter amp.

[Mick Bailey]

I have some pictures somewhere - I'll check the resistor value when I find them, though I recall the dissipation being too much for a regular 6V6. Those JJ 6V6S will stand a lot of abuse - more like a tube that's halfway between a 6V6 and 6L6.

[Mike K]

I just did some calcs on power outputs and such. But I want to comment on the voltages first...

Fair disclaimer: I've never built a 5E3 with a "stock" PT. I come up with a B+ around 420V. I've typically run them between 370 and 350V. IME 370 puts them right at 14W dissipation at idle.

[Mike K]

Just FYI I come up with 21W RMS for a "stock" 5E3 with a 250R cathode resistor, 8k primary OT, and 420V Plate/364V Screen idle voltage @1kHz. I come up with 16W RMS for a "stock" DeArmond with 250R cathode resistor, 6.7k primary OT, and 360V Plate/328V Screen idle voltage @ 1kHz.

I did not use voltages from the schematics, but rather those calculated from PT specs. The power is computed using the previously linked load-line calculator (I verified it calculates the correct RMS for PP @g1=0) using Va-k, and Vg2-k under load. Actually powers might be a little high due to a lack of a good tube recto model. The model itself (in class B) seems to overestimate the power a bit from the load-line. I'm not 100% sure why but I think the 6V6 model is a little optimistic.

All this tells me nothing about the sound... I'm more interested in what will have the best driven tone. All-in-all I'm actually leaning toward using an 8K primary with a larger than stock 5E3 core and a PT that puts me in the 370V B+ range.

[Chuck H]

All this tells me nothing about the sound... I'm more interested in what will have the best driven tone.

Well define "best".?. Since this is a very subjective matter. And then you have to choose best for most people or best for you. The lower voltage and to a degree the lower primary OT would be smoother with less headroom. Some would say mushier. The higher voltage and higher primary OT would be tighter and brighter with more headroom. Some would say harsh. In other words...

One players smooth is another's mushy. And one players tighter and brighter is another players harsh.

So what do you think you'd like from your Deluxe? We already know (for the most part) what the stock circuit delivers. Higher voltage with the higher OT primary could be great with that design. Bias adjusted accordingly.

[Mike K]

Haha - you got me there.

I don't know. Best for me is generally tighter and brighter (to a degree) but I also prefer a modest volume and don't care much about headroom.

I was more or less philosophically comparing the DeArmond which has claimed to be higher wattage (doesn't appear to be the case by my calculation and the info I've gathered). It might be a little less mushy due to a bit stiffer PSU and comparing the OT loads, they hit the knee at a very similar point when compensated for the different voltages each design runs.

I was intrigued by DeArmond's larger OT - but it seems I could probably have a preferable (choosing my words here) result using an 8K with a larger core and adjusting the coupling caps in the 5E3 design. I'm most intrigued by trying a 50W ceramic driver as I think that would change the character quite a bit.

[Chuck H]

I'm really remiss in my avocation as an amp builder and designer in that I've never built a 5E3 circuit or even played through one. So take any thing I say about this model with a grain of salt (or a bucket full). But...

It seems to me that the 5e3 circuit has a lot more lows and low mids in the preamp circuit than many other amp models. It also doesn't use any global NFB. Using a guitar like a Telecaster this could work. But plug in a humbucker in a darker guitar and you get that crazy Neil Young overdrive sound. People like it when Neil does it because, well, it's frickin Neil and he uses it artistically. But if my own amp sounded like that I'd think it was broken.

Now... I see current specs for 5E3 PT's ranging from 70mA to 200mA. What the hell? Most products that claim to be modeled from a reverse engineered original spec at 120mA. That's not too low for a 5E3 so I don't think bumping higher than that gains you much in relative stiffness. But then there's the DCR. Which WILL have an affect on sag. This is a harder spec to find and all the manufacturers I checked seem to leave it out. All I'm saying is that if you want a stiffer power supply for a 5E3 you'll need to dig through the available info and find out what the DCR of an original model is and then find a PT with a lower DCR. Higher or lower voltage not considered at this time.

As far as I can tell the stock spec for a 5E3 OT is 8k primary impedance. So that's not actually a change. What the bigger stack buys you might be more bass response. But only if the stock OT is actually saturating under overdrive. A lot of transformer manufacturers tout this saturation effect but in my reading and experience it's actually very rare to saturate most stock guitar amp OT's to the point of LF deficiency in the real world. But like I said before, I'm not intimately familiar with the amp. So is the stock Deluxe OT too small? I've read that it probably was and that blowing them under heavy clipping is a little more common than average. So sure, bump the current spec on the OT. As to the bigger stack and greater LF, Why? With extra lows and low mids in the circuit anyway it could well be that a clipping 5E3, with a little LF loss is exactly what the design needs in order to balance the clean and dirty tonality. You might even find that increasing the current spec for the OT increases LF under clipping too much for this design. I mentioned before that sometimes it's the deficiencies in some of these older amps that make them sound like they do. I'd say that if you have an over spec OT on hand just try that first. If any extra LF in the clipped tone of the 5E3 circuit is noticed, is it for better or worse? There was a little tiff about measurements vs actual sound not long ago. My position was that we sometimes measure things to see why they sound like they do. And I'll stick with that now. Try a bigger OT (which will already have a bigger stack) and see how it sounds. If you do or don't like what you hear then compare measurements for a bead on what to do next, if anything.