Actually she's just monitoring me waiting for me to say it's time to eat. Everything else is secondary when it gets around dinner time.
Because of the 4 power transformers and six chokes in each monoblock, I generate some substantial chassis voltage mismatch between the amps and the other components, which had left a stubborn little bit of ground loop hum between the components up to now. Well, I have finally gotten the hum/buzz in the amp down to zero. I mean ZERO, with my head right up to the speaker drivers. An almost inaudible thermal hiss is all that remains. Yay!
The final touch was the addition of a Parallel Earthing Conductor (PEC), from the input RCA jack of each amp to the chassis of the preceding component (in my case the shell of its power umbilical). Picture 1 shows one of the PECs, made from 9.5 gauge copper Litz wire. I used these neat copper spring clips on the end which are PERFECT for clipping onto the shell of RCA and umbilical connectors - they are actually made for connecting geophones for seismic monitors up to their associated electronics, and are available from Digikey for $1.48 each! They have a vice-like grip! Picture 2 shows one installed on the shell of the RCA connector.
I picked up this trick from reading about pro-sound installations, which often use a very heavy gauge PEC running alongside the interconnects to handle the interchassis currents from their massively powered gear. They can have enough current running to actually fry the shield wires of their interconnects without a PEC!
Before the addition of the PEC, there was a low-level hum/buzz that was audible with my ear within 6"-12" of the speaker driver. Now it's dead silent. No negative effects that I can hear, either.
Just thought I'd pass it on, it's a nice tweak to get that last little bit of ground loop hum out of the system, without lifting any ground pins or using any loop breaker circuits.
The final touch was the addition of a Parallel Earthing Conductor (PEC), from the input RCA jack of each amp to the chassis of the preceding component (in my case the shell of its power umbilical). Picture 1 shows one of the PECs, made from 9.5 gauge copper Litz wire. I used these neat copper spring clips on the end which are PERFECT for clipping onto the shell of RCA and umbilical connectors - they are actually made for connecting geophones for seismic monitors up to their associated electronics, and are available from Digikey for $1.48 each! They have a vice-like grip! Picture 2 shows one installed on the shell of the RCA connector.
I picked up this trick from reading about pro-sound installations, which often use a very heavy gauge PEC running alongside the interconnects to handle the interchassis currents from their massively powered gear. They can have enough current running to actually fry the shield wires of their interconnects without a PEC!
Before the addition of the PEC, there was a low-level hum/buzz that was audible with my ear within 6"-12" of the speaker driver. Now it's dead silent. No negative effects that I can hear, either.
Just thought I'd pass it on, it's a nice tweak to get that last little bit of ground loop hum out of the system, without lifting any ground pins or using any loop breaker circuits.
Attachments
Well, I made a video on my phone (Samsung Infuse 4G) but it won't upload to Youtube, using the Youtube app and my home Wifi. Just keeps säying "waiting to retry" over and over and over. I've wasted the last three hours trying to get this to work and now my conclusion is - F this.
It was easier building the damn amp.
OK, I managed to get over the "Youtube Crisis" by downloading the video from my phone to my PC, then uploading from there. Enjoy!
https://www.youtube.com/watch?v=3_cAERyP4_8
PS: Sorry about the grumpiness...I hate it when technology doesn't work!
https://www.youtube.com/watch?v=3_cAERyP4_8
PS: Sorry about the grumpiness...I hate it when technology doesn't work!
Last edited:
Dude... Just build a good EL34 PP amp or even better a SS amp with equal wattage and spend the 10K on objectively good speakers.
Hate to say this but 20 years into the future if/when the cognitive dissonance has worn off you're gonna wish you'd built an insanely good full system, not just an insane amp
Hate to say this but 20 years into the future if/when the cognitive dissonance has worn off you're gonna wish you'd built an insanely good full system, not just an insane amp
I actually tried to make a transistor amp like that, but this is what I ended up with...
An externally hosted image should be here but it was not working when we last tested it.
Been there, Blown that! I am known as Tubelab, and also known for melting more than a few tubes. What is not known is that I have fried far more silicon, GaAs, GaN, and SiC than tubes in my lifetime. Much of that was due to my 41 year career as a transmitter designer at Motorola.
A SET is by definition class A. In theory the average plate current does not change with signal. In reality the Mu of the tube will change with instantaneous plate current and voltage. This will impart some gentle predominantly second harmonic distortion that increases with power output. When you build a 200 watt SET and run it at an average power around 1 watt the effect is very minimal.
Note the current meter behind the tube. It is measuring plate current (about 270 mA). It does not move until the amp is driven into severe clipping by a blonde with an electric guitar cranked to 11. At this point the amp is nowhere near class A!
Attachments
Well look at your characteristic curve. It's a curve isn't it?
If you stick to a smaller region of it the deviation from the tangent is lower. So the deviation the lower the distortion.
As this amp has more power, for a given level will fall into a smaller part of that curve than a lower power one... Common sense really
You don't interpret I-V curve linearity from the tangents to each curve. Below is a set of plate curves for the 6E5P in triode mode (my driver tube). I stole this from Ale Moglia's website, BTW, so that's not me reflected in the CRT! Each curved line represents the I-V at a particular grid voltage (noted at the top of each curve). As the grid swings with the signal, it would go left to right across the plate curves, intersecting them. The more evenly spaced the plate curves are from each other, the more linear the device is. A triode is extremely linear and as such can often be run without negative feedback (except for that inherent in the triode itself).
Unfortunately, I have not found such curves for the 833C output tube, as it's usually biased into A2 territory (drawing grid current) for it's day job as an AM transmitter tube. I'm running it biased into class A1 for the first 40W or so.
Attachments
Last edited:
As a measure of frequency response, the 2kHz square wave is often used. Here's mine (also posted earlier in this thread) with a 4R dummy load in place. Nice flat tops and steep sides for a monster OPT. A little bit of overshoot in the front but ringing is well controlled. Considering the amount of parasitics in this 62lb beast, that's pretty damn good!
Sound-wise, there no annoying or fatiguing sound in the highs to indicate that the overshoot is anything serious, so I'm leaving it as is. Once it's hooked up to a real speaker, that will change pretty drastically, anyway...
Sound-wise, there no annoying or fatiguing sound in the highs to indicate that the overshoot is anything serious, so I'm leaving it as is. Once it's hooked up to a real speaker, that will change pretty drastically, anyway...
