Tweaking My 102 Amp: A Massive Resistor Makeover

[martin manning]

Yes, Radiotron Designer’s Handbook, 4th Ed. I’ll go read Merlin’s explanation on that; I’d like to understand better. Regarding this:

The 220k load line is not as steep as the 100k one. That means it’ll afford a greater output voltage swing before clipping — in other words, more gain for the stage — which at the same time means less harmonic distortion.

I think you are conflating gain and headroom.

We are going off the subject here, though. ;^)

[ayan]

I think you are conflating gain and headroom.

I really don't believe I am. However, the larger resistor will yield both more gain -- before clipping -- and more headroom to the stage.

We are going off the subject here, though. ;^)

As the OP of this thread, I disagree. I did agree with you when you went "off topic" (your words, not mine) and down the 2.5% difference in resistance rabbit hole. Charge away, Martin, we're totally on topic and I think there is good information to be had from this dialog.

Gil

[martin manning]

Gain is the slope of the transfer curve (dVa/dVg), is it not?

What I mean by off topic is we are now discussing the sound of changes made to the circuit vs. the sound of new resistors vs. old ones.

[ayan]

Gain is the slope of the transfer curve (dVa/dVg), is it not?

What I mean by off topic is we are now discussing the sound of changes made to the circuit vs. the sound of new resistors vs. old ones.

Maybe we can drill that down after you’ve had a chance to read the reference I provided. Deal?

Gil

[martin manning]

Hold on... here is gain (specifically small signal gain like we were discussing above):

-mu Ra/(Ra + ra)

This is defined at the quiescent point, and it has nothing to do with headroom or clipping.

[jcsifu]

Gil- after following this thread for a while I just wanted to verify that I am correct in assuming that you changed out the plate and cathode resistors from Dale metal films to the LCA0414 Draloric Carbon film resistors and prefer the change. Is that correct?

[ayan]

Hold on... here is gain (specifically small signal gain like we were discussing above):

-mu Ra/(Ra + ra)

This is defined at the quiescent point, and it has nothing to do with headroom or clipping.

To recap, I will list again my original claims that a larger plate load resistor results in:

(a) More headroom for the stage. CHECK
(b) Less harmonic distortion at the output (taken off the plate) before clipping occurs. CHECK (well, unless you have some place you can point me to that suggests otherwise)
(c) A greater output voltage swing before clipping (and, quoting myself, "in other words, more gain for the stage"). I'll give myself a CHECK here, the intended use of the word gain was clearly to describe the output compared to the input.
(e) You brought up the definition of gain at the quiescent point, mu Ra/(Ra + ra). I'd like to honest and say that this is not necessarily what I was referring to in (c) above. I was talking about the output signal level at lowest possible grid voltage before clipping occurs, saying the higher plate resistor value would allow for a greater positive excursion at the output. However, obviously mu Ra/(Ra + ra) will also increase with the plate resistor value.

So, I think we can conclude that all four above are causally correlated with the use of the larger plate resistor.

Gil- after following this thread for a while I just wanted to verify that I am correct in assuming that you changed out the plate and cathode resistors from Dale metal films to the LCA0414 Draloric Carbon film resistors and prefer the change. Is that correct?

Hmmm, not quite, let me summarize again. The LCA0414 Dralorics are carbon film resistors and replaced almost everything (not counting the power supply resistors, but counting the bias feed resistors mounted on the preamp board), except: (a) V1 grid stoppers (at least for now I still have old NTE metal films there); (b) Power amp feedback loop 4.7K resistor (at least for now, I still have an old NTE metal film there); and (c) V1 and V2 plate and cathode resistors. The old plate resistors were Dale MFs, the old cathodes KOA Speer MFs; all were replaced with Roderstein MF resistors. Hope this makes it clear now.

Gil

[Mr. dB]

Now, how close do you think you could get to this amp with resistors that are still being made?

[jcsifu]

Oh, ok my bad. I did misunderstand. I have tried a few different "high end" low % CF such as Takmans and the likes before but had always returned the Dales back into duty. I was curious and had some questions for you, but now they are moot.
thanks for clarifying that Gil.

Hmmm, not quite, let me summarize again. The LCA0414 Dralorics are carbon film resistors and replaced almost everything (not counting the power supply resistors, but counting the bias feed resistors mounted on the preamp board), except: (a) V1 grid stoppers (at least for now I still have old NTE metal films there); (b) Power amp feedback loop 4.7K resistor (at least for now, I still have an old NTE metal film there); and (c) V1 and V2 plate and cathode resistors. The old plate resistors were Dale MFs, the old cathodes KOA Speer MFs; all were replaced with Roderstein MF resistors. Hope this makes it clear now.

Gil

[ayan]

Now, how close do you think you could get to this amp with resistors that are still being made?

I sincerely don’t have the knowledge to answer with authority. Others that have built and tweaked to death more amps than I will probably be able to chime in and weigh in on this. As I said in the beginning, building to a given schematic will most likely result in a good, perhaps great sounding amp. It’s the extra few % points that are a very personal decision. Different iron goes a long ways, for example. And I have built identical amps that used the exact same everything and one always sounded better than the other. Furthermore, some amps will sound good using a given circuit, and still sound good if you try something else in them. Others don’t behave in the same way, in my limited experience. My low plate amp always sounded good, no matter what. Not perfect though. I’m looking forward to giving it a makeover and see where it ends up. The changes in my 102 were well worth the effort. I hope to be able to say the same after I’m done with the low plate amp. Wish me luck!

[erwin_ve]

Gil, I can follow your post on gain and headroom, but this is confusing;

" To recap, I will list again my original claims that a larger plate load resistor results in:

(b) Less harmonic distortion at the output (taken off the plate) before clipping occurs. CHECK (well, unless you have some place you can point me to that suggests otherwise)

When using a larger plate resistor( like 220k or 150k) you automatically are in curved part of the load line of a 12ax7, so the upper half is different from the bottom half of the audiowave, which is harmonic distortion. Or do you draw your loadline different?

[martin manning]

Gil, I read through the bit on load resistance vs. second harmonic distortion in Merlin’s book and that’s clear. Now I need to see why my simple THD calc doesn’t agree. But, notice he says at the end of that section that his analysis does not account for the earlier onset of grid clipping with higher load resistances, and that that will produce a “harder or crisper overdrive sound.” That is of course due to distortion. I ran some sims with 100k/1k5 and 220k/3k3 and looked at FFT’s, but I’m not confident that the grid current is modeled correctly.

Recall we are discussing this because Tony asked why HAD used several different plate loads, and I said harmonic complexity. I still believe that is the reason behind it. This is about the effect on the sound due to distortions caused by changing the values of resistors and not the distortion (or other?) effects from the resistors themselves. I still have not seen any evidence indicating there is any significant contribution there. Merlin’s Hi-Fi preamp book shows that resistor-induced distortion, especially for MF types, is very very low. We agree (I think!) that there is a significant effect on the sound when changing the value of the plate load, so in any comparison test which looks for evidence of resistor-induced distortion, I would insist that the value of the resistors be as close to the same as possible.

[didit]

Hello -

Wondering if anyone with depth in ODS experience has tried the current production PRP 1W MF for V1/V2 plate resistors?

https://prpinc.com/products/leaded-meta ... film-audio

Best .. Ian

[jelle]

RDH4 describes a shift in the harmonic content when changing plate resistor from 2x internal plate resistance from mostly 2nd harmonics to 3rd and 5th harmonics mostly when going to a plate resistor equal of 4-5x internal plate resistance of a tube. So the harmonic content shifts when doing that, and the result is very audible in these overdrive amps.

PS: yes that is the Radiotron Designers Handbook, 4th ed.

jelle

The answer is right here.

[jelle]

I have gotten a few messages from people about these resistors. They can be found here: https://www.ebay.com/sch/lampje*2007/m. ... ksid=p3692

When time allows, I will make a list of what I have available and post it here, but for now all is listen on ebay as well, so people can see it there. I apologize for the pricing, these were pulled from the NATO repositories for me and imported.

Also, please do not PM me here, but email through my website. Happy soldering!

Jelle

Ok, here is the list of what I have in stock and can provide you all with. Please email me through www.welagen.com

http://pdf.datasheetcatalog.com/datashe ... 8260_1.pdf

In stock
LCA 0309:
12K
100K

LCA0411
820r
2.7K
68K
100K
220K
1M

LCA0414
150r
2.2K
3.3K
5.1K
10K
11.5K
21.5K
25.5K
53.6K
56K
82K
91K
110K
115K
120K
120K
150K
191K
232k
270K
470K
1.8M
4.7M

LCA0617
2.2K
3.3K
4.7K
47K
100K
150K
180K
240K
270K
1M

LCA0922
2.2K
27K
120K

LCA0933
2.7K
33K