Somewhere toward the beginning of this thread I mentioned that I had measured the noise of the P3 -- with the stock LM7x15 regulators etc. It came in as Wayne illustrates in the schematic (640pV/√Hz). I don't think you'll do any better with a SR, or Silent Switcher. If you can closely match JFETs you could tweak it a little lower by eliminating the source resistors. This has been mentioned before.
I got back into playing with my amp after neglecting it for a few months. I was having a hum problem and have to echo the need to have very good connectivity between all the chassis panels. I thought I did but then I took someone’s idea of using a Dremel flapper sander and went to town in areas that did not show. That fixed my hum issue. I do have the whoosh / rf static noise at idle I have seen others mention - oddly significantly more so in one channel. I soldered a 1.8k resistor on top of the 2.2k (so I didn’t have to take everything apart) to lower my gain a bit, which helped. It seems to be below the noise floor of my records so I may live with it. Otherwise the amp sounds great!
With my Pearl 2 I also had some noise, and I followed the advice from somewhere earlier in this thread to connect the audio grounds via a 10nF // 10R to the ground post in stead of directly. That helped and I now have only a little bit of hiss left, which is unavoidable as a phono pre is a high gain amp. I will certainly try this with the P3 as well.
Hello,
I measured the Sparkos 15 volt regulators, they were not as clean as the 7815 and 7915 regulators.
Thanks DT
Probably not, but try here:
https://www.diyaudio.com/community/threads/pearl-3-pcb-pair-for-sale.407705/page-2
Hello,
As you say "just for show". Okay just for fun.
What do you expect to gain? How will you document it?
Heat moves very slowly. Voltage and current changes due to signal modulation are very low. There is very little, as in tiny, temperature change due to signal modulation.
Try using 0.1% resistors. A little over $1.00 each. The heat they make will not make any difference either.
https://www.mouser.com/ProductDetail/Vishay-Dale/RN60E10R0BB14?qs=7cfb17uqMF9uzPAWzbQpEQ==
Thanks DT
Well I'm not a scientist, so I will not have to prove anything.
I still believe that coupling those fets together will make their drift more aligned.
The 10r resistors are like all the rest matched together, I have a few hundred of them so found some really close ones, and they are 750mw types, and all parts are checked with magnet, I dislike steel legs.
I still believe that coupling those fets together will make their drift more aligned.
The 10r resistors are like all the rest matched together, I have a few hundred of them so found some really close ones, and they are 750mw types, and all parts are checked with magnet, I dislike steel legs.
Just for fun I asked chatgpt about any advantages in bonding together jfets, here is the answer.
Please state any advantages in bonding together 4 2sk170 jfets they operate in parallel, and in a high gain low noise
Bonding together 4 parallel 2SK170 JFETs in a high-gain, low-noise application, offers several distinct advantages that can enhance performance. Here are the key benefits:
Please state any advantages in bonding together 4 2sk170 jfets they operate in parallel, and in a high gain low noise
Bonding together 4 parallel 2SK170 JFETs in a high-gain, low-noise application, offers several distinct advantages that can enhance performance. Here are the key benefits:
**1. Enhanced Thermal Stability
- Consistent Temperature: Bonding JFETs together ensures they share the same thermal environment, leading to consistent operating conditions and reducing the likelihood of thermal drift. This stability is critical for maintaining low noise levels and consistent gain.
- Reduced Temperature Gradient: A uniform temperature across the JFETs helps maintain balanced characteristics, crucial in high-gain circuits where even minor temperature differences can lead to significant performance variations.
2. Improved Electrical Matching
- Maintained Matching Over Time: Lab-matched JFETs are initially close in parameters like Vgs. By bonding them, you maintain this matching over time, which is crucial for balanced operation and low noise.
- Reduced Parameter Spread: Keeping JFETs at the same temperature minimizes the spread in their parameters, leading to more predictable and stable performance.
**3. Lower Noise Performance
- Thermal Noise Reduction: By maintaining a stable temperature, you reduce thermal noise, which is crucial in a low-noise application. Variations in temperature can increase thermal noise, so a shared thermal environment helps keep noise levels low.
- Minimized Flicker Noise: Consistent operating conditions help reduce flicker noise, which can be a significant contributor to overall noise in low-frequency applications like phono stages.
**4. Balanced Current Sharing
- Equal Current Distribution: Bonding JFETs together helps in achieving balanced current sharing among them. Uneven current distribution can lead to one JFET operating out of its optimal range, increasing noise and reducing gain.
- Stable Biasing: Equal temperature helps in maintaining stable bias points across the JFETs, crucial for consistent high-gain performance.
**5. Reduced Distortion
- Harmonic Distortion Reduction: Thermal coupling reduces mismatch in operating points, minimizing harmonic distortion. In high-gain circuits, small mismatches can result in significant distortion, so thermal bonding helps maintain linearity.
- Stable Operating Conditions: Consistent temperatures prevent drift in operating conditions, which helps in maintaining low distortion.
**6. Improved Reliability and Longevity
- Thermal Stress Reduction: Bonding reduces thermal stress by preventing large temperature swings that could lead to mechanical or electrical degradation over time.
- Enhanced Durability: Consistent thermal conditions can improve the longevity of the JFETs by reducing wear and tear due to thermal cycling.
**7. Simplified Thermal Management
- Easier Heat Dissipation: A shared heatsink provides a larger surface area for heat dissipation, making it easier to manage and dissipate heat generated by the JFETs, which is especially beneficial in high-gain circuits where power dissipation can be significant.
- Compact Design: Bonding JFETs together allows for a more compact and efficient design, simplifying the layout and thermal management of the circuit.
Summary
Bonding 4 parallel 2SK170 JFETs together offers significant benefits for high-gain, low-noise applications:- Improved thermal stability leads to consistent and reliable performance.
- Better electrical matching and reduced noise.
- Balanced current sharing and reduced distortion.
- Enhanced reliability and simplified thermal management
And I asked about the resistors,
In a high gain low noise circuit, which resistor type Is preferred as emitter resistor, a 0.1% 125mw 10 ohm, or a 1% resistor also 10 ohm, but with 750mw rating, both types metal film
For a high-gain, low-noise circuit, the choice between a 0.1% tolerance, 125 mW, 10-ohm metal film resistor and a 1% tolerance, 750 mW, 10-ohm metal film resistor as an emitter resistor involves several key considerations:
In a high gain low noise circuit, which resistor type Is preferred as emitter resistor, a 0.1% 125mw 10 ohm, or a 1% resistor also 10 ohm, but with 750mw rating, both types metal film
For a high-gain, low-noise circuit, the choice between a 0.1% tolerance, 125 mW, 10-ohm metal film resistor and a 1% tolerance, 750 mW, 10-ohm metal film resistor as an emitter resistor involves several key considerations:
Factors to Consider
- Power Rating and Thermal Noise:
- Self-Heating: A higher power rating (750 mW) means the resistor can handle more power before significant self-heating occurs. Self-heating can increase thermal noise and cause changes in resistance that impact circuit stability and noise performance.
- Heat Dissipation: The 750 mW resistor is less likely to heat up significantly in your application, leading to more stable resistance values and lower thermal noise.
- Tolerance and Precision:
- Resistor Tolerance: While 0.1% tolerance is much more precise than 1%, this level of precision is generally less critical for noise performance in low-value resistors like 10 ohms. It is more about ensuring the resistance value is accurate for circuit design purposes.
- Impact on Noise: Precision in resistance (0.1% vs. 1%) does not significantly affect the noise characteristics of the resistor itself.
- Resistor Type and Noise:
- Metal Film Resistors: Both resistors are metal film, which is ideal for low-noise applications. Metal film resistors inherently have low noise compared to other types like carbon composition.
- Current Handling and Stability:
- Power Dissipation Capacity: The 750 mW resistor can handle higher currents without overheating or changing its resistance value significantly, which is advantageous in high-gain circuits where current stability is crucial.
- Stability Under Load: Higher power resistors tend to maintain their resistance value more reliably under load, which is important for maintaining consistent gain and low noise in high-gain circuits.
Preferred Choice:
The 1% tolerance, 750 mW, 10-ohm metal film resistor is preferred for the following reasons:- **Thermal Stability:
- Higher power rating means less self-heating, leading to more stable resistance and lower thermal noise. This is crucial in high-gain circuits where even small temperature-induced resistance changes can affect performance and noise levels.
- Current Handling:
- The higher power rating allows the resistor to handle higher currents without significant resistance changes, ensuring stable operation and maintaining low noise in high-gain applications.
- Noise Considerations:
- Both resistors are metal film, which are inherently low-noise, but the 750 mW resistor’s greater ability to dissipate heat contributes to a lower overall noise profile.
- Long-Term Stability:
- The 750 mW resistor’s higher power rating provides better long-term stability, which is essential in high-gain circuits where consistent performance over time is required.
Conclusion
In a high-gain, low-noise circuit, the 1% tolerance, 750 mW, 10-ohm metal film resistor is the better choice. Its higher power rating leads to lower self-heating, greater thermal stability, and improved current handling, which are more critical for maintaining low noise and consistent performance than the slightly higher precision of the 0.1% tolerance resistor.It's an interesting question; so here's an equation:
Short of some serious cooling, I don't think this is going to be particularly significant. Likewise, the increase in wattage of the resistor, and lower running temperature is going to have a smaller effect.
My suggestion for noise would be ground plane and a faraday cage around the sensitive parts of the circuit to reduce the effect of EMI.
So if you can lower T, you can reduce noise. 20C is 293K; 50C is 333K and the increase in E is a factor of √(333/293) or 1.066, ie 6.6% for a 30 degree reduction in temperature.
Short of some serious cooling, I don't think this is going to be particularly significant. Likewise, the increase in wattage of the resistor, and lower running temperature is going to have a smaller effect.
My suggestion for noise would be ground plane and a faraday cage around the sensitive parts of the circuit to reduce the effect of EMI.
Hello,
Food for thought.
https://www.ap.com/category/news/from-the-test-bench-resistor-noise-and-non-linearity
https://convexoptimization.com/TOOLS/HoferUltraLowTHD+NPart1.pdf
https://convexoptimization.com/TOOLS/HoferUltraLowTHD+NPart2.pdf
https://slideplayer.com/slide/17832865/
An added thought: always measure / test your assumptions.
By my measurements:
Vishay low PPM TC, PTF resistors, cost many times more, they are the best measuring through hole resistors. Vishay RN mil-spec low PPM 25-50
PPM, measure almost as well. The added thermal mass of RN60 vs. RN55 improves the measurements. RN65 resistors do not fit on the PCB's. Next to carbon composition resistors Vishay foil resistors are the worst measuring resistors.
Thanks DT