I see the Caddock 930. Thanks. Also gave me the chance to read up on R9 experimentation.
In my previous builds of the 1.3 for preamps I never used any snubber parts, but I am wondering whether this would be useful for the Soekris dam 2941? For the +/_ analog and for the + digital. Could you please advice.
Thanks. nash
Its basically about how a certain transformer rings with a certain diodes bridge and how to critically damp that. It definitely can have an impact. To fine tune the three snubber parts (Cx,Cs,Rs) there is Quasimodo. Then some less precise or even intuitive approaches. There has been much written in the relative threads.
Was that a question originally addressing Selfy?
In any case, Y class will have smaller Vgs (OFF) than GR. Zero Volt Vgs (G+S shorted) will be sooner into saturation with Y than with GR.
Its true that a resistor in between flattens transition to the saturation region by introducing Vgs drop, so GR gets to work at 1-2mA fine too.
When you got Y they aren't behaving far off, so for simplicity i.e. no source resistor between thus less components, you could prefer those.
@nashbap there is this https://www.diyaudio.com/community/threads/quasimodo-results-only.313202/ thread where ppl are posting results of quasimodo tests. If you dont care to build a rig you can always look if you find yours there
Thanks Stajo. I could not find the Antek AS0509 but I figured from the discussion of results I should be OK with Cx 10nf, Cs 150nf and Rs 7r for 120V parallel primaries and paralleled secondaries.
Why did you choose the Caddock 930? Is that after you had tried other types? Thanks/
Why did you choose the Caddock 930? Is that after you had tried other types? Thanks/
The Caddock just happened to be in my stash with the approximate right value, and I know they are decent for the purpose. Bourns are ok to. The thickfilms I like best in speaker xover applications are Isabellenhütte PBH series. Here I just want as much heat out of the box as possible. You should not worry to much about wirewound induction though. At those small values its not much to write home about.
Last edited:
Yes, it was because he had already tried it. But everyone's entry is welcome 🙂
Indeed, I intend to use less components if there is no drawback
On the other hand an 0805 resistor will add no significant space with just a 2mm length track if needed
You have many times mentioned for the J3 lower Idss due to their lower Vgs(off) in case of 5V or less output
Both 2SK880 Y and GR are still in production
The difference is that Y will be in saturation while GR in Ohmic region for the same 2mA Idss.
Max transconductance occurs when Vgs=0V, right? Do we need that and prefer Ys?
Or is it better to have a small margin below saturation for DC fluctuations and prefer GRs with a resistor?
Better saturated ccs vs small Vds behavior according to post #4,024 curves should happen when using it at 0.5-1mA with a proper value source resistor.
The voltage drop on the source resistor elevates the source so the gate stands negative in comparison. Effectively a minus Vgs gate bias value is created.
Still fair enough mA to bias the Ubib's BJT transistors I believe. But you can simulate with 0.5mA to 2mA span and check impact on OLG and phase margin.
Possibly 1mA should be a midpoint good choice. I opted for 2mA in the original for nicest reg Zo curve shape. That was with using PF5102 JFETs though.
Max Yfs occurs at Idss (Vgs=0) yes. Higher Idss goes along with higher Yfs, but that's for when running full. As we level down same type Jfets by source resistor we get almost same Yfs between Idss classes. See how different Idss samples Yfs curves bunch together as they travel lower for drain current. This chart comes from the 2SK209 datasheet. Same for 2SK117 same for 2SK880.
The voltage drop on the source resistor elevates the source so the gate stands negative in comparison. Effectively a minus Vgs gate bias value is created.
Still fair enough mA to bias the Ubib's BJT transistors I believe. But you can simulate with 0.5mA to 2mA span and check impact on OLG and phase margin.
Possibly 1mA should be a midpoint good choice. I opted for 2mA in the original for nicest reg Zo curve shape. That was with using PF5102 JFETs though.
Max Yfs occurs at Idss (Vgs=0) yes. Higher Idss goes along with higher Yfs, but that's for when running full. As we level down same type Jfets by source resistor we get almost same Yfs between Idss classes. See how different Idss samples Yfs curves bunch together as they travel lower for drain current. This chart comes from the 2SK209 datasheet. Same for 2SK117 same for 2SK880.
Good day! I need your help. I assembled a board for the positive terminal... but the regulator isn't working; the bulbs are burning out (the output constantly shows around 11 volts). I removed the bulbs and shorted the bulb contacts... now the output is around 6V and still not adjustable. Meanwhile, transistors Q2 and Q3 are getting very hot. What could be the issue? Has anyone encountered this before?
Official board? If not, check its connections are 100% correct to the circuit's schematic.
Measure voltage across R1. It must not be very different than 0.55-0.6V. Else replace Q1.
LEDS burning is wrong try replace Q2 Q3. Mosfets between legs 1,3 must be showing 3-5V.
Measure voltage across R1. It must not be very different than 0.55-0.6V. Else replace Q1.
LEDS burning is wrong try replace Q2 Q3. Mosfets between legs 1,3 must be showing 3-5V.
no, it's not official. But I have other boards on this board. They work fine. But something didn't work with this one (
Replaced Q1, Q2, Q3 and the LED. The LED blinked and slowly went out. While the bulb was on, the output was about 6V. When it burned out, it became 11V+
Is it sure it burns out or it simply extinguishes after the voltage charges up C2 beyond what's necessary? Test the LED directly with diode check mode of a DMM that can light up LEDs when the Ubib's power is OFF.
First of all make sure you have turned down the VR1 trimmer low so its not dictating a high Vout setting beyond what is available for Vin. Then check the R1 and M1 M2 voltages I mentioned. Look closely on the Mosfets body print and make sure you did not mix up the PMOS-NMOS types and their proper positions.
First of all make sure you have turned down the VR1 trimmer low so its not dictating a high Vout setting beyond what is available for Vin. Then check the R1 and M1 M2 voltages I mentioned. Look closely on the Mosfets body print and make sure you did not mix up the PMOS-NMOS types and their proper positions.
R1, M1, M2 corrected.
It turns out I forgot to solder resistor R9. But the stabilizer still does not regulate the output power. I wonder what could have burned out. There is only one left for Q2 and Q3.
It turns out I forgot to solder resistor R9. But the stabilizer still does not regulate the output power. I wonder what could have burned out. There is only one left for Q2 and Q3.
Without R9 the output MOSFET was not conducting and the constant current was forced through other smaller parts.
Now after you restored R9 check all characteristic DC voltages of parts and places vs a working Ubib to narrow down on what might be broken or wrongly put.
Now after you restored R9 check all characteristic DC voltages of parts and places vs a working Ubib to narrow down on what might be broken or wrongly put.
So I tried with another 470R over R9,it's much better, it wakes up the high bass and it's more dynamic,
I think I found a good compromise with a KZ 100µf and UES 220µf,
I also tried several transformers because I realized that the latter had a great influence (Ocore 100VA James transformer,Rcore 30 Va Selectronic,toroidal 25Va Triad Magnetic,and the best one I kept 50Va Triad Magnetic,the latter gives me a dynamic high bass, the right amount of extreme bass and an open midrange.
