Both ValveWizard (Merlin B.) and Bruno P. have it right, IMO. Merlin's advice might be more relevant for single-ended designs.
Don't worry about it too much. Just wire it as usual and you should have no problems. One thing to watch out for is to isolate the RCA inputs from the chassis.
Don't worry about it too much. Just wire it as usual and you should have no problems. One thing to watch out for is to isolate the RCA inputs from the chassis.
I think so too, but with an aluminum chassis and keeping signal wires away from the power bits, I am not even sure using 2-conductor shielded cables would be necessary. And grounding the shield of those 2-conductor cables at both ends could be a pain.
It’s interesting to take it all in and get a sense of why an approach is best.
Has anyone else had any trouble soldering in the trimmers?
The ground plane pad is very thin and with the thermal reliefs, I struggled to get any solder to stick or flow through.
Picture through the magnifying lamp.
Any tips?
I thoroughly cleaned the leads and pads, tried extra flux, increased the temp of the iron to 385 momentarily, but very thin results still.
The ground plane pad is very thin and with the thermal reliefs, I struggled to get any solder to stick or flow through.
Picture through the magnifying lamp.
Any tips?
I thoroughly cleaned the leads and pads, tried extra flux, increased the temp of the iron to 385 momentarily, but very thin results still.
The solder must flow, as the guild navigators would say. (10 points if you get that movie reference)
Hello,
I am going to try lowering the gain of my SCG. I had set it up originally with the 10k / 1k RCSH / RG combo as outlined in the BOM, and now I am going to change RG to 4.75k. I am unclear on what I voltage I should set across R107/207 because I don't know how to figure out what current the Schade network will drain for a 10k / 4.75k combo. I am using 10ohms for R107. I'd appreciate any help.
Also, when changing the gain, am I still shooting for 20 to 30 volts for Q102, and 65 volts for Q101?
Thanks,
Alan
I am going to try lowering the gain of my SCG. I had set it up originally with the 10k / 1k RCSH / RG combo as outlined in the BOM, and now I am going to change RG to 4.75k. I am unclear on what I voltage I should set across R107/207 because I don't know how to figure out what current the Schade network will drain for a 10k / 4.75k combo. I am using 10ohms for R107. I'd appreciate any help.
Also, when changing the gain, am I still shooting for 20 to 30 volts for Q102, and 65 volts for Q101?
Thanks,
Alan
For Q102, yes, 20 volts, and yes, 65 V for Q101.
The 10k/4.75k combo will result in less current in the feedback network. You can leave the CCS current unchanged---the feedback network will just draw less. Anything between 27 mA and 30 mA should work, which would mean 0.27 to 0.30 across R107.
Curious to hear how it sounds and measures compared to the stock version. Do let us know your impressions 🙂
The 10k/4.75k combo will result in less current in the feedback network. You can leave the CCS current unchanged---the feedback network will just draw less. Anything between 27 mA and 30 mA should work, which would mean 0.27 to 0.30 across R107.
Curious to hear how it sounds and measures compared to the stock version. Do let us know your impressions 🙂
I made the change and adjusted the voltages. Using a 4.75k resistor, the gain moved from 10X to a little over 3X. I did find more useful range out of my 24 step attenuator, but subjectively the preamp loses its rich and powerful feeling bass, sounds a bit less warm overall, and doesn't sound as punchy. For my tastes, I consider it a failed experiment, so next I'll try maybe 6X gain, or go back to full gain and buy an attenuator with more steps.
Objectively, the noise floor is lower as expected, although I had some junk being picked up when I took the measurement. At the same volume of about 1dB over unity gain, the level of the second harmonic is about 9dB lower than that at 10X gain. The third harmonic level is about 2dB less than at 10X gain.
Below is a picture of the full gain vs low gain harmonic measurement. Full gain is in white and low gain in green.
Alan
Objectively, the noise floor is lower as expected, although I had some junk being picked up when I took the measurement. At the same volume of about 1dB over unity gain, the level of the second harmonic is about 9dB lower than that at 10X gain. The third harmonic level is about 2dB less than at 10X gain.
Below is a picture of the full gain vs low gain harmonic measurement. Full gain is in white and low gain in green.
Alan
This is absolutely wonderful feedback! I think I see what is happening. Can I ask what you are driving with the pre and the input impedance of the following stage?
I am not sure if you are asking about listening or measuring. When listening I am driving a pair of Pass Aleph 2 monoblocks with 10k input impedance. When measuring I am driving the line inputs of a Prism Sound Orpheus studio interface with 14.5k input impedance. I could also measure using a friend's QuantAsylum measurement interface, but the noise floor of the Orpheus is lower, so it is easier to see the preamp's noise and harmonics.
Edit, I have some follow-up questions:
1) When I switched to lower gain, I lowered Q102 to 20 volts. It had been at 25 volts when running 10X gain. Would that affect the sound flavor?
2) I am using a 25k stepped attenuator. I also have a 100k stepped attenuator which would give me a little more volume adjustment range. Is there any advantage of using the lower impedance attenuator?
Thanks
Edit, I have some follow-up questions:
1) When I switched to lower gain, I lowered Q102 to 20 volts. It had been at 25 volts when running 10X gain. Would that affect the sound flavor?
2) I am using a 25k stepped attenuator. I also have a 100k stepped attenuator which would give me a little more volume adjustment range. Is there any advantage of using the lower impedance attenuator?
Thanks
Last edited:
1) 20V is fine.
2) As adason says, less noise is the reason to use lower impedance, but I wouldn't hesitate to use the 100k.
You could try keeping the original 1k resistor at 1k and try lowering the 10k to 6k. This will mean higher current in the feedback network--about 6.5 mA---but it will keep the output impedance low and should result in better bass. The 6k resistor should be rated for at least 1W and probably 2W. The pot will be at about 3k, dissipating about .12W, should be okay. These values are based on an Spice model. Might be worth checking the current through the 1k resistor and the trimpot resistance for 60-65V.
2) As adason says, less noise is the reason to use lower impedance, but I wouldn't hesitate to use the 100k.
You could try keeping the original 1k resistor at 1k and try lowering the 10k to 6k. This will mean higher current in the feedback network--about 6.5 mA---but it will keep the output impedance low and should result in better bass. The 6k resistor should be rated for at least 1W and probably 2W. The pot will be at about 3k, dissipating about .12W, should be okay. These values are based on an Spice model. Might be worth checking the current through the 1k resistor and the trimpot resistance for 60-65V.
Last edited: