So,
I've been mainly happy with my OTS 100w and had about 6 months to get used to gigging it and to find the right speakers etc. I've found a startling discovery which has really made the amp work for me. I wanted to share it with everyone here in case they may like the improvement.

The issues I had were:
1 - Although the tone controls had a noticeable effect on the tone, they weren't as responsive as other amps I have had: the tone always seemed to have a slightly lo-fi quality for lack of a better word...
2 - The PAB was useless live. Too much bass was introduced unless I cranked the treble control, but this caused problems when PAB was not engaged.... This was especially bad on the clean channel, where the tone sounded flabby at all times with PAB on.

So I played with cathode and plate resistors. Even modded the amp to S&M specs. Also reduced snubbers to 270p. All these made a difference to the overal tone, but did not seem to impact the two issues above.

Today I was looking at a schematic of Dumble 124 and i got to thinking something. The cathode and plate resistors on v1a have been inreased in the ceriatone when compared to this dumble. I have read this helps warm up the sound somewhat. I thought, well I could change them to 124 specs, but would this really be the cause of all my problems? It turns out not.... the solution is more complex to explain, but simpler to implement:

The tone stack in 124 is the same as the ceriatone, so I realised that changing the plate resistor on v1a would have a likely impact on the impedance seen by the tone-stack, but this hasn't changed, so perhaps there is a mismatch, when compared to the original design. Turns out (assuming a 12ax7 output impedance of around 60k) that the change to the plate resistors cause the output impedance of v1a gain stage to increase from around 38k to around 47k... but the tonestack has not been modified, and still has the same input impedance. Guess what this does?  

So, I thought, can I re-create the same impedance match between v1a output and the tonestack input? Yes, I can! The slope resistor is surely the key.... So, what if we scale the slope resistor up by the ratio by which the v1a output impedance changed (47/38). Well, we get around 190k. Didn't have resistors to make up this value, but I had a 180k lying around. Replacing this one resistor made the amp behave beautifully in all scenarios! I couldn't believe what I was missing before I did this!

It seems that someone has introduced a well meaning mod (to change the v1a cathode and plate config) but forgotten to take account of the impedance impact on the next stage. When this is factored in, I think the amp behaves "as originally designed". I must stress that the plate and cathode have not been changed to 124 values, so the well meaning mod has been preserved, its just been taken to its full conclusion and implemented "correctly" (in my opinion obviously).

I urge all OTS owners to try this and welcome and views or comments on the above!

Cheers,
K

Hi exocet,
So it seems I have what you call a "high plate" amp as per the stock OTS config. I am deffinitely not trying to suggest that there is anything incorrect about the "high plate" config, nor the "low plate" config. Neither am I saying that the tonestack is incorrect in and of itself. When you say that the tone stack is based on actual examples do you mean that the combination of a "high plate" amp with a 150k slope resistor is based on an actual example?

There is always an impedance mismatch between v1a and the tonestack (assuming the guideline that "input impedance > 10 x output impedance" can be applied). In any case, this is never as high as 10 times, however, I feel its fair to say that the exact ratio should be preserved in all designs. After all, what reason would anyone have for deliberately changing this mismatch?

I know its a strong point of view, but the effect is so dramatic, and the theory seemingly sound, that I feel happy to assert that a "high plate" config requires a different slope resistor from a "low plate" config. I think the effect of taking this on board is so positive that I feel this goes beyond personal preference, and is in fact correct. I appreciate this is likely to be controversial in its lack of relativism.

Either way, whether this is a design flaw, or just my personal prefference I reccomend everyone first using this amp to experiment with this mod. I would personally love to see it incorporated into the OTS & S&M designs (maybe others require an adjustment too, I havn't checked all the configs...) I have tried this with a variety of loading scenarios (i.e. loads before the amp), guitars, effects pedals. There is no scenario which is not improved by the mod.

The effects of the mod are quite subtle in some scenarios (those which were not previously problematic) but dramatically better in other scenarios (those which were previously problematic). This further makes me feel that this is a design issue, rather than my personal preference. The change is not as simple making everything brighter or everything darker as might be attributed to preference, but seems to "correct" a number of issues

'scuse the intrusion but...

Wow!  I learned some really nice stuff here.  I ordered my OTS and HRM 50 watters prebuilt.  However, I'm not averse to picking up my pencil, braid, and solder to try something new.   I'm not as familiar as the forum's builders, I am not famliar with where such obvious tagnames are located.  I'd like to try it.  Man, if someone could shoot me a PM with maybe a chunk of the layout with notes on this that would be so appreciated... 

Thank you very much for your prompt and helpful reply.  Safety precaution needs noted.  I have a way to drain the caps to below dangerous levels before taking the chassis out of the headshell and will follow up with additional precaustions.

Once I got a finger across the OT primary terminals by mistake in my Frank-en-Champ very early on in my tinkering while the amp was running, burned a tiny hole all the way througgh the skin right into the meat, didn't bleed because wound was cauterized as it was being bored by the arc of rampant b+... and my stupidity.

Thanks again, will look at both layouts.  Then get to most carefully researching and then source the part(s). 

Hi,
My normal method is to use a 100w 16 ohm resistor. I find it a very convenient component because it can also be used as a dummy load to replace the speaker when doing osciloscope work. Its plenty adequate to absorb any power left in the filter caps. I just touch it in parallel with one of the caps and wait a couple of seconds. I then measure any remaining voltage accross each cap using a multimeter set to a high DC voltage range.

Having said all that, I have found that on my OTS simply powering the amp of in the normal fashion (turn to standby, wait 2 secs, turn off, wait 2 secs) will drain the caps. I always measure across the caps just in case even though it has never shown more than 10v when doing this, which is a perfectly safe level. Bear in mind that your OTS may behave differently and you should always, always, always check. "Measure twice, die never", thats my philosophy.

Same waiver as always, I cannot be held responsible for anyone doing this, your life is in your hands, not mine....

P.S. - Love the avatar, Roy is an inspiration!

Hi Bluesking
I tried your 180Kohm resistor mod to the tonestack input and I also replaced the  bias bypass cap from 5uF to 2.2uF, as you described in you pm to me. It’s really a very big difference. Now my amp sounds as I want it to sound. I can get the John Mayer sound and also other smooth lead sounds, which I couldn’t get before. Your mods definitely worked out for me. Thanks a lot for your help and advising.
 

i have performed a detailed circuit analysis on the skyline eq, using different resistors.

results are as follows:

for a given eq setting (ie bass middle treble), and varying the slope resistor only, the effect is to slightly reduce the bass, and make a larger skoop in the mids, dropping approx 1db at 1.5 kHz,

... hence, the more fender like tone!!

i may try increasing it to 200k and see what happens.................

i am happy with the 180k mod and am not keen on trying the 68k

however i have simulated it... you get more signal to the rest of the amp and more bass and mids making the sound less scooped

overall i would say it is not the direction i want to take it; and if anything try slightly increasing the 180 k to 190 or even 200k

however being cautious, i want to try it in a live situation as it is now

the overdrive is much smoother and refined, especially with a les paul and the PAB engaged