I recall that C1 and C2 could not be too high with a practical value of 47uF, as the Idss was relatively low and it would take a long time to charge up the capacitors. Is this now changed?
Is there a reason why R9 4k2 is larger than R11 4k?
If irfp240 is better than the irfbc40, then I am happy to buy the new ones.
Is there a reason why R9 4k2 is larger than R11 4k?
If irfp240 is better than the irfbc40, then I am happy to buy the new ones.
Sorry for so many posts and questions. I am getting excited. I have been planning to work on my regulator this weekend and you posted the good news this morning. I now have to stop my work and am waiting for some clarifications.
Comparing to the 5k, on top of the stability enhancement of adding D3, D4 and D2, the following have been changed:
1. The 2n3819 is replaced by 3k3.
2. The JFET underneath mpsh81, 2SK170 is replaced by J201 in parallel with a resistor.
3. The JFET that drives the voltage ref, J201 is replaced by 2SK170.
4. Collector of Q1 and 2n3819/3k3 used to be connected to the senses, now connected to Rail or Ground respectively.
Could you give a brief explanation on why you did the above 4 changes?
Comparing to the 5k, on top of the stability enhancement of adding D3, D4 and D2, the following have been changed:
1. The 2n3819 is replaced by 3k3.
2. The JFET underneath mpsh81, 2SK170 is replaced by J201 in parallel with a resistor.
3. The JFET that drives the voltage ref, J201 is replaced by 2SK170.
4. Collector of Q1 and 2n3819/3k3 used to be connected to the senses, now connected to Rail or Ground respectively.
Could you give a brief explanation on why you did the above 4 changes?
This is a new circuit. I built one version of this and was trying to get rid of the electrolytic or large caps. In my build I used a 124nF MKP with a 2R. Any lower value made it oscillate. However, by now you should know that you will have a different layout and you'll need to experiment with these values. Take them only as suggestions, points to start from.
You may not experience any oscillation with this one either, but I wanted to have R15 there as an option to control the output impedance.
You're right about C1 and C2. The capacitance there lets you compromise between the resistor noise and a long start up time. Larger capacitance means lower noise, but longer start up time.
The only requirement is that R9 and R11 should be close in value, so that Q2 and Q3 have roughly the same voltage across.
Argh, number 4 is just an oversight on my part. I went back and changed the schematic. Also, I removed one of the capacitors across R9 and R11, even though one should still choose for himself what value to use there. As for 1, 2, and 3, it just made more sense in the overall circuit to make these changes. I don't claim any audible improvement. Just hoping this version to be less hassle to build and be more stable temperature wise.
Bill, I'd suggest that you first build a prototype and test your layout. I plan to replace the CCS with a simple current source based on LT3083. 
At least for the high current versions (> 800 milliamps).
LT3083 - Adjustable 3A Single Resistor Low Dropout Regulator - Linear Technology
May or may not be an improvement. Remains to be seen.
At least for the high current versions (> 800 milliamps).LT3083 - Adjustable 3A Single Resistor Low Dropout Regulator - Linear Technology
May or may not be an improvement. Remains to be seen.
Ikoflexer,
Thanks.
I am in production mode after being idle for a long time. I have 2 weeks compulsory leave from work during Christmas and I want to make most of it doing HiFi building. I am pretty happy with what I have and am rebuilding the regs only to alow running higher current on a power amp chassis. I guess I will not wait for the new CCS to be tested and proven, because I want to see the result now and I can see nothing wrong with the 5k version.
I am struggling with the choice to build a hybrid of 5k-5m or the new 5m. I have no stability issue whatsoever with 5k and have got all the parts. Voltage changes caused by temporature drift are at most a few volts which do not worry me at all.
Changing IRFBC40 to IRFP240 and 2n5088 to 2n5089 (slightly increased hFE) and reduce the Idss dramatically from using 2SK170 to J201 for the mpsh81 are risks for me. But as long as they can only make it more stable and less prone to oscillation then I am happy to build the new one.
Thanks.
I am in production mode after being idle for a long time. I have 2 weeks compulsory leave from work during Christmas and I want to make most of it doing HiFi building. I am pretty happy with what I have and am rebuilding the regs only to alow running higher current on a power amp chassis. I guess I will not wait for the new CCS to be tested and proven, because I want to see the result now and I can see nothing wrong with the 5k version.
I am struggling with the choice to build a hybrid of 5k-5m or the new 5m. I have no stability issue whatsoever with 5k and have got all the parts. Voltage changes caused by temporature drift are at most a few volts which do not worry me at all.
Changing IRFBC40 to IRFP240 and 2n5088 to 2n5089 (slightly increased hFE) and reduce the Idss dramatically from using 2SK170 to J201 for the mpsh81 are risks for me. But as long as they can only make it more stable and less prone to oscillation then I am happy to build the new one.
Bill, just change this in your 5k: replace the J201 in the reference with the 2sk170 and use a zener across the 2sk170 to protect it. The rest can stay the same. Playing with the capacitors is something that you might want, because if it works well, it's better to have fewer and smaller value capacitors on the output.
Do not replace the 2n5088 with 2n5089. They are both fine, I just left the 2n5089 in because that's the model I use by default.
Do not replace the 2n5088 with 2n5089. They are both fine, I just left the 2n5089 in because that's the model I use by default.
That sounds good.
I have played with the capacitors a lot. I have no problem with 22uF Rubycon ZL || 6 x 1.1uF MKP at the output and 4 x 1.1uF MKP at the load of 2 to 4 opamps.
Since I have only 2SK170 of 9mA+ I guess I may add a resistor to the one underneath the mpsh81s.
P.S. Tip: place the 6 x 1.1uF MKP closer to the MOSFET output, while the 22uF at the end of the string of the 6 x 1.1uF furthest away from the MOSFET output. This sounds fine. If the 22uF is close to the MOSFET output then I heard the "capacitor" sound. I have done all sorts of experiments with 5k so for me 5k is done.
I have played with the capacitors a lot. I have no problem with 22uF Rubycon ZL || 6 x 1.1uF MKP at the output and 4 x 1.1uF MKP at the load of 2 to 4 opamps.
Since I have only 2SK170 of 9mA+ I guess I may add a resistor to the one underneath the mpsh81s.
P.S. Tip: place the 6 x 1.1uF MKP closer to the MOSFET output, while the 22uF at the end of the string of the 6 x 1.1uF furthest away from the MOSFET output. This sounds fine. If the 22uF is close to the MOSFET output then I heard the "capacitor" sound. I have done all sorts of experiments with 5k so for me 5k is done.
Last edited:
Yes, use the new values. The two diodes help keep the voltage more stable if the environment temperature varies. They are optional.
Thanks.
Last I would like to check with you these.
I have ordered some 2sk170GR that have 3mA - 6.5mA Idss from Mouser. These will be used in places for the Vref as well as to drive the mpsh81.
In the Vref I will run around 1mA current after some ??? ohms on the 2sk170GR.
In the place underneath the mpsh81, no resistor is used so the current will be between 3mA to 6.5mA. Is that alright? What is the ideal current here? I noticed that with 5k the current was 6.5mA to 8mA, with 5m J201 it would be 1mA but shunt with a 470k resistor.
Last I would like to check with you these.
I have ordered some 2sk170GR that have 3mA - 6.5mA Idss from Mouser. These will be used in places for the Vref as well as to drive the mpsh81.
In the Vref I will run around 1mA current after some ??? ohms on the 2sk170GR.
In the place underneath the mpsh81, no resistor is used so the current will be between 3mA to 6.5mA. Is that alright? What is the ideal current here? I noticed that with 5k the current was 6.5mA to 8mA, with 5m J201 it would be 1mA but shunt with a 470k resistor.
With the current 5k regulators I built I have the 2sk170 (Idss = 8mA) here underneath the mpsh81. There have been no problems. That was also according to the official schematic of 5k.
Now I have the 2sk170GR (Idss = 3 - 6.5mA) for the new build of hybrid 5k/5m. Should I add a resistor to reduce the current to 1mA without the external resitor shunt?
Now I have the 2sk170GR (Idss = 3 - 6.5mA) for the new build of hybrid 5k/5m. Should I add a resistor to reduce the current to 1mA without the external resitor shunt?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.