Remove Q102. Short former drain and source pads with a small jumper. Replace R108 with 15K 0.5W-1W. Keep it 2mm off board, not contacting. Use IRF9620 or IRF9520 for Q101.
Do the same for the symmetrical unit using NMOS 620/520 for Q201 of course.
All caps 63V, don't rectify a transformer to more than 60V raw DC.
Set CCS at 1 to 1.1A. Watch output MOSFET's sink size. Spare current times output Volts is the dissipation there.
Dear Salas,
Can we put BHC ALC10S Slit Foil Electrolytic 80v cap on the SSLV1.1?
BHC - AudioCap Limited
I mean whether the physical size fit?
Thanks!
Can we put BHC ALC10S Slit Foil Electrolytic 80v cap on the SSLV1.1?
BHC - AudioCap Limited
I mean whether the physical size fit?
Thanks!
If you say spare 200mA on top of load, that will create 10W dissipation @ 50V out. If 100mA then 5W. 1C/W or 2C/W sink.
Sorry Salas, I made a mistake.
Can you show me how to convert the SSLV1.1 to regulate "POSITIVE" voltage 50v 900ma?
NO NEGATIVE VOLTAGE AT ALL. I'm sorry.
Since this is a circlotron, the output stage has 2 x positive voltage, no negative at all.
For stereo, I guess we need 4 SSLV1.1 and one for each positive rail regulation.
Thank you!
Can you show me how to convert the SSLV1.1 to regulate "POSITIVE" voltage 50v 900ma?
NO NEGATIVE VOLTAGE AT ALL. I'm sorry.
Since this is a circlotron, the output stage has 2 x positive voltage, no negative at all.
For stereo, I guess we need 4 SSLV1.1 and one for each positive rail regulation.
Thank you!
I explained for positive and negative, you do only the positive sections, you keep the negatives unused. Break them away and store them for future use.
As much the Mosfets spec. You check the datasheets they give a max amperage for 100C internal. To keep that temp in check, the sinks got to be correct. Means you factor in everything for the current that goes through each and the voltage across each. Watch this video about sinking. EEVblog #105 - Electronics Thermal Heatsink Design Tutorial - YouTube
Some good news.
I also replaced the Q306 and connected the BiB this afternoon.
This time, I measured the voltage across R301 instead of measuring its current by parallel it.
1. The voltage across R301 is about 1.40V. 1.4V/4.7Ohm ~ 300mA not far from my target of 340mA. I input these values into the SSLV calculator. It gives the voltage drop across each green led is appr. 2.06V which is closer to 2.1V than my measured 1.84V.
2. However, I only get slightly more than 22V output. The trimmer was turned to the max. In my built, I am using 2 x 3k + 5k trimmer with 2 green leds. The Idss of the 2SK30A GR were 3~4mA. Maybe, the real Idss is a bit lower than the measured value.
I will replace one of the 3k resistor with a 10k and see if it raise the voltage output.
Salas, any comments on the proposed move?
measure the voltage across each individual component in the voltage reference string.
The total of the individuals should equal the gross measurement across the whole string.
You can then calculate the current through the fixed resistor. Then decide which component would need to be changed.
The total of the individuals should equal the gross measurement across the whole string.
You can then calculate the current through the fixed resistor. Then decide which component would need to be changed.
You never use a DMM in current mode in parallel to a component, you effectively short it.
The IDSS in the Vref circuit is significantly less than when using a battery because the JFET sees only 0.6V across it there (one Vbe). The 2SK30ATM especially has -0.4V VGS minimum at its lowest IDSS samples when you need -0.6V VGS maximum. It shows much reduced IDSS then and it will not satisfy good CCS criteria in its medium range samples in this application. Get 2SK117GR at least for that position. From your resistor values and Leds in the Vref there should be only about 1.5mA running right now. Insufficient. You can verify that by measuring voltage drop across a 3K you got there and derive the running current. More resistance will raise the Vout but get the proper JFET when you can for better CCS stability also.
I am building a preamp (lsk pre or ba3 se) and would like to have the simplistic phono stage in the same case, with a separate case for the transformers, simplistic power supply board and possibly the bib board.
I am considering placing the bib near the ba3 board, but am afraid that the heat generated will affect the phono stage.
Since the bib has sensing on the output, is it ok to place it in the psu enclosure and run an umbilical with about 1 metre? This umbilical would carry the dc for the preamp and also the dc for the phono stage, use twisted pairs of 22-24awg wire and some sort of 6 pin connector.
I am considering placing the bib near the ba3 board, but am afraid that the heat generated will affect the phono stage.
Since the bib has sensing on the output, is it ok to place it in the psu enclosure and run an umbilical with about 1 metre? This umbilical would carry the dc for the preamp and also the dc for the phono stage, use twisted pairs of 22-24awg wire and some sort of 6 pin connector.
I measured the voltage drop across the 3k resistors and obtained 4.89V. It suggested that the Idss is about ~1.6mA. I replaced one of the 3k by a 10k. After switching on for 30 min, the Vout is about 30.6V and the voltage across the R301 is 1.28V, lower than the value (1.4V) before replacing the 3k.
I cannot obtain 2SK117 in HK. But I do have some 2SK170, forget whether they are GR or BL. Are they better than 2SK30A in this application?
I cannot obtain 2SK117 in HK. But I do have some 2SK170, forget whether they are GR or BL. Are they better than 2SK30A in this application?
Good for Vgs and noise and Yfs. Driftier than K117GR and more capacitive. You may spot a tendency for the output voltage to rise rather much with ambient temperature rise especially in a hot reg with a mainly passive resistive Vref and K170. But much better suited than K30ATM in mid term. Get some K117s on line to ultimately use there or ask Tea Bag to send you some confirmed original ones. He is quite a helpful chap.
Andrew & Salas
thank you for the input.
I observed that the voltage across R301 started to rise first and climb to about 2.xV in the first 1~2mins and then started to drop until 1.28V about 30mins. Is it normal?
Another observation: in the first trial, I got Vout=22V and 1.4V across R301. In the second attempt, I got Vout=30V but the voltage across R301 drop to 1.26V. Why there is a drop in voltage across R301?
The leds D302~D304 are brighter than D305~306. The current past thru Q302 was more than that past thru Q303?
thank you for the input.
I observed that the voltage across R301 started to rise first and climb to about 2.xV in the first 1~2mins and then started to drop until 1.28V about 30mins. Is it normal?
Another observation: in the first trial, I got Vout=22V and 1.4V across R301. In the second attempt, I got Vout=30V but the voltage across R301 drop to 1.26V. Why there is a drop in voltage across R301?
The leds D302~D304 are brighter than D305~306. The current past thru Q302 was more than that past thru Q303?
Last edited:
Those are the Vgs thermals of the associated MOSFET and a little delta from the associated LEDS. Normal, it mostly cycles back.
The VDS changing across Q101 MOSFET due to Vin-Vout changing (that is what Q101 has across) changes its Vgs just a little and its capacitance. Burns different Watts with different VDS when its current is almost constant. This CCS has 2 pillars. The resistor value and the LEDsVftotal-Vgs voltage. Shake one a little, the roof dislodges to a different position so to speak.
The VDS changing across Q101 MOSFET due to Vin-Vout changing (that is what Q101 has across) changes its Vgs just a little and its capacitance. Burns different Watts with different VDS when its current is almost constant. This CCS has 2 pillars. The resistor value and the LEDsVftotal-Vgs voltage. Shake one a little, the roof dislodges to a different position so to speak.