Calibrating gain stages

[lightfoot]

I was wondering about the best process for checking to see if each gain stage is beginning to clip more or less at the same time. Would you apply a signal at the phase inverter 1st to see how much the output tubes can handle and then work back towards the preamp?

This question assumes that I’m aiming for a clean-voiced amplifier.

[romberg]

For a lot of amp topologies, things tend to break down at the power tubes first and then work their way backwards towards the input. So, (most of the time), you can put a scope on the output. Then inject a reasonable signal into the input and turn up the gain until you start to see the signal distort on the output.

Most of the time the first thing to distort will be the power tubes. You can confirm this by attaching the scope (or a second channe of the scope onto the power tube grid(s). If the output it distorting but you have clean signal on the grids then you've found the first thing to go. If the signal on the grids is also distorte then look at the PI input.

You can just work your way down the chain until you find a clean undistorte signal. Then increase the gain some more and see if something else distorts.

If the amp has a master volume of some sort then the above applies only if the master is set on 10. Lowering this control can cause an earlier stage to distort first (at a lower volume). A PPIVMV (my favvorite) usually causes the PI to go first. And a MV before the PI tends to cause the last stage of the preamp to be first. The more gain/volume contros an amp has the more flexability there can be in the order stages can distort in.

It is probably pretty rare for multiple stages to all start to distort at exatly the same time. But it can be useful to know what actually is causing the distortion in an amplifier and at what volume/gain settings.

Mike

[lightfoot]

Fortunately I’m working on a build that’s a single channel amp with the following topology:

- V1B (5751) (V1A not used)
- Volume/single tone control
- Paraphase PI (12AY7)
- Output tubes (6L6’s)

I just need to get the grain relationship between the 1st and 2nd gain stages where I want them.

If I use my 5F1 champ as a model the amplitude of the 2nd gain stage (measured at the signal side of the signal capacitor) matches the the amplitude of the 1st gain stage with the volume at “5”. I’m considering using that as a model or a reference point for the push/pull build. Of course my guess is that this should be measured and adjusted with the tone control set at the half-way point.

As it stands right now the 2nd gain stage (which is also the phase inverter) matches the 1st stage with the volume at 3.5 so I need to loose a little gain. Most likely I’ll find a slightly higher cathode resistor to drop the gain just a little or add a little more negative feedback…

Thanks
Devan

[romberg]

If your goal is no distortion/clipping then I'd not overthink it too much. Build the whole amp. Then find the point at which clipping starts on the output by adjusting the volume control set the volume to the point where it is just short of clipping.

Then with the amp off, measure the pot. Say you are totally clean on a 1m volume pot right up to 750k. So, you know that you need to knock off 25% of the signal coming out of your first stage to keep everything clean. You can add a single resistor before the volume pot to change the ratio of the voltage divider and decrease the amount of signal leaving stage one. In made up example you want to loose 25% of the signal. So, some math leads to (.25 * 1m) / (1 - .25) or 330k.

So, you could have the v1 output go ||-----330k----volume pot----- ground. And now when the volume is on 10 you are dropping 25% of the signal you would have had without the extra 330k resistor. This is a common thing done to make a voltage divider and a volume pot at the same time.

Mike

[lightfoot]

I’ll definitely take a look at output stage clipping as soon as my dummy load gets here. It may take some time though because it’s coming from China I think…

Thanks for the suggestions

[Stevem]

You do not need a dummy load to check your preamp stages, but note that a purely resistive load does not act like a speaker which has a inductance factor.

Also note that since your first preamp stage is fully class A that means that one half of its output sine wave will first saturate then start to clip, and the other half will more or less just clip.

You can get some nice fat harmonic overtones with saturation before clipping sets in.

[lightfoot]

You do not need a dummy load to check your preamp stages, but note that a purely resistive load does not act like a speaker which has a inductance factor.

Also note that since your first preamp stage is fully class A that means that one half of its output sine wave will first saturate then start to clip, and the other half will more or less just clip.

You can get some nice fat harmonic overtones with saturation before clipping sets in.

Since I’m only applying a 1khz signal at 100mv I’m not seeing the 1st stage clip. I need to see at what amplitude it starts to clip when I get a chance though.

I’m seeing 6vac off the 1st coupling capacitor with a 100mv signal applied to the input at this point. The tube in V1 is a 5751.

Another thing that I could try is checking the 1st and 2nd stage relationships after installing a 12AX7 in V1 and installing a 12AU7 in V2 for the P.I. Might be worth some tube swapping first….

[SoulFetish]

I would just add that how each gain stage is biased makes a huge difference in how the amp sounds, and responds at the beginning of clipping through heavy overdrive (as well as if a particular stage has a cathode bypassed or not). For instance, if you set up you test circuit with trim pots for cathode biasing, you can hear how changes to one stage might give you more gain, but the overdrive might sound "buzzy" or lack articulation. Yet, this can be mitigated adjusting the bias one or all of the following stages.
I was just testing my most recent amp design in this way because I was experimenting with some different cascode configurations for the input. Turns out I didn't like any of them for this amp. But, since the chassis was drilled out and set up for a compound stage, I installed a DC coupled cathode follower as an addition to the input stage and bootstrapped the input with a split load. Even with modest changes in grid bias voltage at the first stage you could see a drastic change in sine wave symmetry and shape at the output. I think with some adjustments to the biasing of the phase inverter and 4 preamp stages I was able to dial in a pretty decent sound from clean at about "3" on the guitar volume knob, to pretty heavy overdrive with the guitar volume all the way up.
But honestly, no matter how much I think I've got an idea which is sure to improve the amp, every significant change I've been experimenting with since the first iteration of the amp end up being a disappointment .