I recently started poking around and modding an EL34 PP amp that I've been using for the past 7 years or so. I measured B+ from the supply at 500V which was surprising given that the last stage PSU cap that the B+ is taken from is only rated at 450V.
After poking around some more it looks like someone (previous owner perhaps) modded the PSU to draw from the 400V xfmer output tap instead of the 360V tap, thereby raising the B+ and pretty much over-volting all of the caps in the amp by 20-60 volts.
My question is, has anyone experimented with increasing or decreasing supply voltage and its affect on the sound quality? This amp sounds pretty good so I'm not sure I want to mess with it. Is it still safe to continue using the amp in this configuration, considering all the caps are being run at or slightly above their ratings? On the other hand it's been run like this for nearly a decade and nothing has blown up so...
After poking around some more it looks like someone (previous owner perhaps) modded the PSU to draw from the 400V xfmer output tap instead of the 360V tap, thereby raising the B+ and pretty much over-volting all of the caps in the amp by 20-60 volts.
My question is, has anyone experimented with increasing or decreasing supply voltage and its affect on the sound quality? This amp sounds pretty good so I'm not sure I want to mess with it. Is it still safe to continue using the amp in this configuration, considering all the caps are being run at or slightly above their ratings? On the other hand it's been run like this for nearly a decade and nothing has blown up so...
i think it would cause distortion because too much or too little plate voltage changes the design parameters of the amp by creating more or less current flow causing the load line to shift away from the most linear portion of the tube characteristic curve. you would have to change the bias point and the total grid voltage swing to correct this problem but then you would end up with less amplification.
increasing plate voltage would increase volume, i think, but also cause a decrease in fidelity.
im just a noob.
Last edited:
General speaking here. With my guitar amps on the pre-amp tubes; reducing the plate voltage & helps to generate more harmonics & raising the plate voltages typically cleans up the amplified signal. All the time with the same cathode resistor/capacitor combo.
With the final amplifier section; I do the same thing to vary the sound except that it's important to know where you place the tube in the positive & negative sine & what the compression level will be. I raise the cathode resistor when I raise the voltage (well over Max spec) to move into a cleaner tone placing the tube in a more positive sine. The more the tube is away from the negative sine, typically the more clean sound it will pass. The higher the cathode resistor, the less current the tube will conduct, however then lowering the plate resistor & then you thus raise the plate voltage (as you question) will cause more gain as already stated. Exceeding the optimum gain will again enter into the various types of distortion, however slightly different then operating in the negative sine as mentioned above.
Now this is with my specific design operating down in the "brown sound" region (low pre-amp B+) with class A / SE configuration. With the class AB (Push-Pull) one can work with current caused B+ "Sag" and go from low drive (high B+) clean to heavy driven brown (low B+) sound with a high enough plate/bias resistor.
The best way is to do all the reading you can get on the subject as you are doing now, and then acquire a decade box full of resistors, variac (of which you may already posses) and start switching up resistors. I have 5 decade boxes, 3 resistor & 2 capacitor & still end up soldering in other swappable parts dialing in tone.
So you see in my world no component is safe! Double up those 450 Vdc caps in series for a 900 volt rating. Then measure the AC component on the DC line to watch how bad it gets. Add 2 more in parallel if you need to. Caps can be bombs, ask my plastic army men (when I was 10) what a 5Vdc, 2000mfd cap does when over' volted. I refuse to pay x 4 for any cap over 450Vdc rating.
Enjoy,
Doug
With the final amplifier section; I do the same thing to vary the sound except that it's important to know where you place the tube in the positive & negative sine & what the compression level will be. I raise the cathode resistor when I raise the voltage (well over Max spec) to move into a cleaner tone placing the tube in a more positive sine. The more the tube is away from the negative sine, typically the more clean sound it will pass. The higher the cathode resistor, the less current the tube will conduct, however then lowering the plate resistor & then you thus raise the plate voltage (as you question) will cause more gain as already stated. Exceeding the optimum gain will again enter into the various types of distortion, however slightly different then operating in the negative sine as mentioned above.
Now this is with my specific design operating down in the "brown sound" region (low pre-amp B+) with class A / SE configuration. With the class AB (Push-Pull) one can work with current caused B+ "Sag" and go from low drive (high B+) clean to heavy driven brown (low B+) sound with a high enough plate/bias resistor.
The best way is to do all the reading you can get on the subject as you are doing now, and then acquire a decade box full of resistors, variac (of which you may already posses) and start switching up resistors. I have 5 decade boxes, 3 resistor & 2 capacitor & still end up soldering in other swappable parts dialing in tone.
So you see in my world no component is safe! Double up those 450 Vdc caps in series for a 900 volt rating. Then measure the AC component on the DC line to watch how bad it gets. Add 2 more in parallel if you need to. Caps can be bombs, ask my plastic army men (when I was 10) what a 5Vdc, 2000mfd cap does when over' volted. I refuse to pay x 4 for any cap over 450Vdc rating.
Enjoy,
Doug
Last edited:
I've tried running a circuit well below B+ volt spec'd (400-0-400) once due to a problem with a transformer, one leg of the high voltage pair I wanted to use was more than 25% over spec'd (like 460+) and didn't match other so, I wired it to run the lower voltage of the 2 HT pairs, which was balanced, but a bit more than 15 or 20% spec'd for the circuit/tubes - like 350-0-350, maybe less, have to look it up... anyhoo -
It ran/made music, but had something missing (headroom) and it ran hot, meaning the iron/transformers/chokes. The Transformer provider (ed-something ;^) fixed it and now this thing is running at spec'd voltages and the iron runs much cooler. This was a 300b (and still is, although being experimented still). I bet your caps are good quality, were under rated but still, running anything way out of spec, higher or lower, is a bad idea, esp. caps and NOS tubes. Have fun and be safe.
It ran/made music, but had something missing (headroom) and it ran hot, meaning the iron/transformers/chokes. The Transformer provider (ed-something ;^) fixed it and now this thing is running at spec'd voltages and the iron runs much cooler. This was a 300b (and still is, although being experimented still). I bet your caps are good quality, were under rated but still, running anything way out of spec, higher or lower, is a bad idea, esp. caps and NOS tubes. Have fun and be safe.
I once built a beast of a push-pull 300B amp, with a gargantuan, variable regulated B+ supply and a variable negative fixed bias supply. The driver stage had its own separate PSU, set to +440V DC. This allowed me to twiddle knobs and raise/lower the B+ of the output stage in isolation, while adjusting the fixed bias to keep the 300B's biased correctly. That way I was able to compare higher voltage with less current vs. lower voltage with higher current in the outputs stage only.
The OPTs were a pair of UTC LS-63, using the 6k ohm primaries. I never used a NFB loop in this amp. The driver was a 6DJ8 LTP DC-coupled to a 6SN7 diff-pair, RC-coupled to the PP 300B's.
What I found was:
- The high V/low I state gives you more power, but 'sounds' more 'lean,' or maybe 'brighter.' I attribute this to less damping due to higher rp of the output tubes in that state. The loadline would be steeper, so maybe there are more odd order harmonics in the distortion spectra.
- The low V/high I state gave less power, but 'sounds' more 'robust' or 'warm.' I attribute this to better damping due to lower rp of the output tubes in that state. There are probably less odd order harmonics present in the distortion spectra this way.
If you increase the B+ past the voltage ratings of the capacitors, you are likely to blow those caps. That makes a horrible mess and could even be dangerous. There may be other parts that will be over-voltaged too. A resistor might actually burn (smoke and fire). Avoid that.
If you look at the triode curves for EL34, you'll see that it looks really nice and linear with very low plate volts and high plate current. But it won't make much power with 270V on the plate and 90mA plate current (PP class A).
Now look at the triode curves where everybody uses EL34, at about 425V on the plate and 55mA plate current. That's way over to the right, where the curves start bunching together, indicating poorer linearity. But you get much more power that way (PP class AB).
_________________________________________
PS - That gargantuan PP 300B amp had reliability problems due to many mistakes, including a very poor layout. It was eventually stripped and made into other things.
--
The OPTs were a pair of UTC LS-63, using the 6k ohm primaries. I never used a NFB loop in this amp. The driver was a 6DJ8 LTP DC-coupled to a 6SN7 diff-pair, RC-coupled to the PP 300B's.
What I found was:
- The high V/low I state gives you more power, but 'sounds' more 'lean,' or maybe 'brighter.' I attribute this to less damping due to higher rp of the output tubes in that state. The loadline would be steeper, so maybe there are more odd order harmonics in the distortion spectra.
- The low V/high I state gave less power, but 'sounds' more 'robust' or 'warm.' I attribute this to better damping due to lower rp of the output tubes in that state. There are probably less odd order harmonics present in the distortion spectra this way.
If you increase the B+ past the voltage ratings of the capacitors, you are likely to blow those caps. That makes a horrible mess and could even be dangerous. There may be other parts that will be over-voltaged too. A resistor might actually burn (smoke and fire). Avoid that.
If you look at the triode curves for EL34, you'll see that it looks really nice and linear with very low plate volts and high plate current. But it won't make much power with 270V on the plate and 90mA plate current (PP class A).
Now look at the triode curves where everybody uses EL34, at about 425V on the plate and 55mA plate current. That's way over to the right, where the curves start bunching together, indicating poorer linearity. But you get much more power that way (PP class AB).
_________________________________________
PS - That gargantuan PP 300B amp had reliability problems due to many mistakes, including a very poor layout. It was eventually stripped and made into other things.
--
Attachments
Something I just realized is that this amp is currently connected in ultra-linear configuration which means that the B+ to the output tranny primary (550V) will pretty much also appear on the screen and plate of the output tubes. Probably ok for KT88, but the screen voltage is over spec for EL34 so it will limit the types of tubes that can be used unless the B+ is brought down closer to 400V.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.