Negative F/B resistors effect on Presence control in D'LIte?

[Mark]

I have a question for D'Lite builders.

I was looking at a D'Lite layout diagram and I was thinking that it is odd that the D'Lite amp has a 6.2K negative feedback resistor. The 100watt (82 watt) Dumble would have more output voltage at the secondary of the output transformer and uses a 4.7K resistor. Initially I did think the lower wattage D'Lite wouldn't drive the speaker as hard and damping might not be a great issue. I then thought, the lower amount of damping voltage across the 390 ohm resistor would also effect how effectively the presence control worked and thus the sonic performance of the amp.

If it was attached to the 8 ohm tapping I think it would be more "correct" according to the numbers. Though it would still only have two thirds of the voltage a Dumble would see at this spot.

My thoughts at the moment are that a 2.7K resistor would deliver the correct feedback voltage, though the amp might sound a little too stiff (depending on speaker choice of course.) A good compromise might be to simply change the feedback from the 4 ohm tap to the 8 ohm tap?

This is not to say Brownnote have missed something or gotten something wrong. The D'Lite amp does use a smaller treble cap than a Dumble, 270pF versus 390pF, and the bright cap is larger, this might perfectly offset the presence control for all I know. Still there is no harm in asking the question.

Are there any opinions on this one?

[Mark]

Still no takers?

[stelligan]

I like a 4k7 there in my D'Lite. Opens them up big time IMHO. Some discussion of that and other D'Lite tweak topics are contained in this thread:
https://tubeamparchive.com/viewtopic.php?t=4153

[Mark]

Dear Stelligan

Thanks for the reply, much appreciated particularly the power supply mods.

This how I saw it (and I could well be wrong.)

A 100watt Dumble using Twin reverb transformers should put out 82 watts as a Twin reverb does.

This would mean 43vpp out of the secondary of the output transformer. With the standard 4.7K negative feedback resistor (-F/B) one should expect to see roughly 3.3vpp at the node of the 4.7K resistor and 390 ohm resistor.

In the case of the (30 watt [4 x 6V6's]) D'Lite with a 6.2K -F/B resistor I calculated that 1.53vpp would be present at this point.

Moving the -F/B resistor from the 4 ohm tap to the 8 ohm tap should yield 2.18vpp.

Stelligan, you could try moving your - F/B resistor from the 4 ohm tap to the 8 ohm tap. It would be closer to the voltage a Dumble would see at this point. I calculated that a 3.9K resistor on the 8 ohm tapping would be equivalent to Dumble figures mentioned above.

Thus in the case of a 30 watt D'Lite 2.7K should deliver the required 3.3vpp at this point. As far as a regular 2 X 6V6 D'Lite the N/FB resistor value comes in at around 2.5K, I'd stick with 2.7K, it's close enough.

Of course I should point these are calculated values, while Moss has actually built and experimented with various values. So perhaps my comment may have merit, and maybe they don't?

[heisthl]

If matching the voltage was the only factor you would be right. But what you really need to match is the tonal effect that the presence imparts on the output section. Which is dependant on power, frequency response and damping effect. I believe the original D'Lite picked the higher value because sonically it worked out best for the response of the 6V6 output section and the OT they were using. There are many ways to go here - look at a variety of commercial schematics and you see a variety of feedback values. The great thing about it is with only 3 values (2 resistors and 1 cap) and a limited number of OT taps it is very easy to experiment.
That said - I always use the 4 ohm tap,4k7,390 and 1uf values and make changes only if the control doesn't work right.

[Normster]

Mark, heisthl pretty much hit the nail on the head. When Moss and I were prototyping the D'Lite, we did a lot of experimenting with different output transformers and feedback values. Our conclusion was that you can't really "down-size" a 100 watt amp. Instead you have to optimize everything for the iron and output section. The 6.2k value really gave the amp a much livelier feel than the more traditional 4.7k/390R or 8k/1k values, both of which result in about an 8.1 dB gain reduction. With lower values, the amp sounded congested using the vibrolux iron. At 6.2k/390R, the amp will really sing with lots of harmonics at a relatively low volume.

FWIW, I had my 100 watt Bluesmaster clone at Moss's place yesterday comparing it to a 44 watt Bluesmaster reverb. While the amps sounded very similar, the 100 watt had to be brutally loud to achieve the same harmonic content.

[Mark]

The 6.2k value really gave the amp a much livelier feel than the more traditional 4.7k/390R or 8k/1k values

Just thought I'd report back to everyone. I tried subbing in the 2.7K value today, as Norm said the amp wasn't as lively, though I could tell the presence control worked better with the 2.7K.

I should firstly point out that the volume level was a bit above normal home listening levels, so I can't say how it would have sounded at full working volume, but I think it is safe to say much of the same.

Back to the amp, with the 2.7K value in the amp, it sounded bland and continued to sound bland regardless of where the presence control was set.

It is a matter of effective presence control operation versus speaker dampening.

Personally, I couldn't justify getting the presence control to have more range and adversely effect the tone of the amp.

My thoughts on this resistor are it should be one of the last resistors you tweak. I would be taking into consideration the speaker I am going to use with the amp (I doubt high wattage speakers need the dampening lower wattage speaker would) the volume I am going to play at, and possibly any effects (such as delay) that I might use. I would adjust the resistor accordingly.

In all I'd say Norm and Moss got this value correct, I have seen posts where guys have changed the negative feedback resistor to a more Dumble-like value, but I can't say I heard enough benefit to cause me to change to a lower value.

Should I have a change of heart, I'll post it up.

[jelle]

Mark,

I worked on Nitroliq's Dlite today and I tried a 3.3K.....same opinion here, I put the 6.2K back in.

I Skylined the EQ and we liked it a lot.

Have fun,

Jelle

[Mark]

My previous post has been altered.