Favorite Ways to Control Compression

[dorrisant]

I've see mention of how screen grid resistors can affect compression.... I've also read where the dropping string resistors (as well as cap value) will affect compression as well. I'd like to understand more about the manipulation of compression within the amp build. Do any of you have evidence (reading material) about either of the two above statements? I'm not solid on this subject and would just like to possibly kick start a conversation about it. Are there any tricks you'd mind sharing on this subject... what worked for you and why?

I feel like I need a test mule that I could experiment with... I know I could pull something out for a mule.

[Littlewyan]

I've heard of screen grid resistors affecting compression by causing the voltage to drop. In an Express I think this is mainly due to the 1K Choke resistor. If you measure the voltage across that resistor when playing the amp full blast, the voltage drops a lot! 100V comes to mind, but don't quote me on that.

Merlin mentions it a few times on his Single Ended page under Screen Voltage http://valvewizard.co.uk/se.html.

I've also heard of anode resistors in the preamp affecting compression. I haven't yet experimented with this myself, but I've heard of 220K providing more compression. I'm building another amp in a few months time and plan on experimenting with this to see just what people mean by 'compression'.

[chief mushroom cloud]

also on valvewizard
http://www.valvewizard.co.uk/dccf.html
I've tested this and don't feel it's the best place to 'add' a compression characteristic....totally screws up the 'response/recovery' or 'feel'...at least to me
Same with screen grid resistors....too funky
Dropping in a slightly underrated HT current spec PT.....that's musical compression to me

[dorrisant]

I found this: https://www.tdpri.com/threads/plate-scr ... on.277603/

Which led to this: http://ampgarage.com/forum/viewtopic.php?t=15156

[Stevem]

I think any discussion on compression needs to start with if the amp in question has a feedback circuit.

[xtian]

Which led to this: http://ampgarage.com/forum/viewtopic.php?t=15156

Well, THAT was an interesting read. Thank you.

[dorrisant]

I think any discussion on compression needs to start with if the amp in question has a feedback circuit.

Any info on that subject?

[xtian]

Sure, let's talk about #183!

[tubeswell]

g2 voltage affects gm. Higher Vg2 = higher gm and vice versa. So if your Vg2 supply is lowered, it will take a bigger g1 input signal for the given output signal at the plate. This isn’t ‘compression’ as such (although the difference is moot to some). Rather it is decreased input sensitivity.

But an unbypassed screen has Ig2 feedback, so as you crank the signal more, you get more Ig2 feedback, which is more like a compression effect.

Merlin Blencowe covers this in his Designing Tube Preamps books

[Stevem]

Sure, and here it is in a nut She'll, though others might want to detail it more.

Global feedback in a amp brings 3 things to the table and takes away 2 others!

It brings a smother frequency response while taking away gain.

It tightens up the low end responce while taking away gain.

It widens the frequency responce while taking away gain .

And it takes away dynamic responce that could be had as the amp gets towards the area of clipping.

And worst of all once the amp is being played loud enough so that the amp has in effect given up that extra gain it'd been using the preamp and output stagtes that are within that golbal feedback loop jump right into clipping and in turn bypass any compression dynamic's you could have used !

For me if a amp is to have global feedback and will be called upon to cover the usage of providing clean sounds and compressed
distorted lead sounds then the feedback circuit needs to be very finely tuned to how cranked up the amp will be played, but also factors of if the amp is tube or SS rectified and how much filtering it has and if and where the amp uses a Choke come into play of the amps overall feel.

[roberto]

I would suggest you to loan a Mesa Boogie 20+20 W DynaWatt power amp and try it.
It has a huge series of resistors on the power supply between plates and screen, with a screen supply capacitor designed to have a time constant of 200 ms. it has a good compression that makes it feel louder.
EL84 on it, so you may like it or not, but it’s not the point of this discussion.

[JD0x0]

IME, relatively high B+ with stiff filtering gives an amp wide dynamics. Anywhere you generate 'sag' is essentially compression, as that's basically a limiter on your peaks. Lots of voltage drop over a resistor, such as a screen resistor or dropping resistor can increase 'sag'.

Don't forget about cathode biasing 'squish' either.

[Gaz]

Just to play devils advocate, I think my approach would be a non expanding compressor pedal with a blend control before the amp or even in the effects loop. All these other methods are valid, but will only work when the amp is cooking at a certain level. I've had the best results control screen voltage or preamp voltage in a high gain preamp. But for me if I really wanna dial in exact amounts of compression I use something like a Cali76 and blend to taste.

[roberto]

Talking about adding compressors, I really like the sound of a good compressor in the loop of an amp.
It adds "that" feel that you can adapt to the volume you need. Then for added sustain during low gain solos, the classic pedal in front of the amp.

[Tobyk]

Thought I’d add some to this subject, which hasn’t been discussed much I believe. I prefer compression being built in to the amp, no pedals needed. Here are some ways that I know of to create some nice compression:

1. Undersized output transformer.
This often gives very good results. The downside being the OT might get hot. Way too small OT won’t work for long, though.

2. Big output tube screen resistors. Something like 10K–47K will add compression, mostly noticably when amp is cranked.

3. Preamp pentode screen bypass cap resistor or pot. This has greater effect than no.2, especially if the pentode is placed in a late gainstage. Try something like 220K–470K between screen bypass cap and ground. Or use a pot.

4. Have the signal from the OT go via a lightbulb before the speaker. A normal jewel light bulb will do. Swart Night Light attenuator uses this approach.

5. Put a 470 ohm resistor between rectifier tube and first filter cap. This will add sag and compression.

These are the five ways I know of. Feel free to share your insight!