NO !
The Mains Bulb Tester goes into the mains feed before reaching the equipment.
For a low power item like a pre-amp use a very low wattage bulb. try <=28W 240Vac
No.
A fuse takes a LONG time to rupture for small levels of overload.
Even at twice rated current, they can take over a minute to rupture.
If you are using a 50VA transformer on a 230Vac supply, then the close rated fuse should be the rounded up value of 50/230 = 217mA Try a 250mA fuse.
Do you have one in stock? What will you use instead? A 1A fuse, or a 3A fuse? How long will they take to rupture?
The bulb filament resistance acts as a soft start that is on the cusp of instability. A 28W 240Vac filament starts off cold @ ~ 8% to 10% of 2057ohms. i.e. cold resistance = 165r to 206r
This rises rapidly as the transformer starts up (but still delivering no secondary current) and the filament glows dimmly indicating that the resistance has possibly risen to 30% to 50% of the hot temperature resistance. The transformer is now presenting a significant impedance to the mains supply and the current demand drops, allowing the filament to cool and become dim, or even appear out.
If there is a fault that makes the transformer impedance lower than expected, the current demand does not drop just as the secondary starts to deliver current. Instead the filament sees the high current demand and goes to almost full hot.
With this high resistance of the hot filament the bulbs drops most of the mains voltage, maybe 90% to 99%. The remaining voltage is applied to the transformer.
A 230:30Vac 50VA transformer with only 5% of mains voltage becomes an 11.5:1.5Vac 2.5VA
How much energy can a 1.5Vac secondary rated at 2.5VA put into a faulty amplifier?
And all this happens automatically in a second, or so.
Use a Mains Bulb Tester.
A fuse takes a LONG time to rupture for small levels of overload.
Even at twice rated current, they can take over a minute to rupture.
If you are using a 50VA transformer on a 230Vac supply, then the close rated fuse should be the rounded up value of 50/230 = 217mA Try a 250mA fuse.
Do you have one in stock? What will you use instead? A 1A fuse, or a 3A fuse? How long will they take to rupture?
The bulb filament resistance acts as a soft start that is on the cusp of instability. A 28W 240Vac filament starts off cold @ ~ 8% to 10% of 2057ohms. i.e. cold resistance = 165r to 206r
This rises rapidly as the transformer starts up (but still delivering no secondary current) and the filament glows dimmly indicating that the resistance has possibly risen to 30% to 50% of the hot temperature resistance. The transformer is now presenting a significant impedance to the mains supply and the current demand drops, allowing the filament to cool and become dim, or even appear out.
If there is a fault that makes the transformer impedance lower than expected, the current demand does not drop just as the secondary starts to deliver current. Instead the filament sees the high current demand and goes to almost full hot.
With this high resistance of the hot filament the bulbs drops most of the mains voltage, maybe 90% to 99%. The remaining voltage is applied to the transformer.
A 230:30Vac 50VA transformer with only 5% of mains voltage becomes an 11.5:1.5Vac 2.5VA
How much energy can a 1.5Vac secondary rated at 2.5VA put into a faulty amplifier?
And all this happens automatically in a second, or so.
Use a Mains Bulb Tester.
Last edited:
use the MBT to power ON the transformer alone (no rectifier, no smoothing capacitors, no amplifier).
Measure the two sets of secondary voltages.
Are they what you expect?
Power OFF.
Short together the two middle leads.
Power ON
Measure the three sets of output voltages.
Are they what you expect?
Power OFF.
report back.
Measure the two sets of secondary voltages.
Are they what you expect?
Power OFF.
Short together the two middle leads.
Power ON
Measure the three sets of output voltages.
Are they what you expect?
Power OFF.
report back.
Finally, I got my B1 up and runnig. AndrewT I followed you advise and nothing blew up ;-).
However, I wired it up with my power amps and a audio source and there is no sound. I followed Salas advice in "HOW TO KNOW IF THE CIRCUIT IS WORKING PROPERLY" to check whether it is working.
So, here are my results:
LEDs all seem to work.
I connected a mesmerize I-Select for input selection. I enabled the first input by shortwiring the according contacts. When it is powered on the input is going through to the inputs of the B1. So, this should not cause the problem.
Any advice?
However, I wired it up with my power amps and a audio source and there is no sound. I followed Salas advice in "HOW TO KNOW IF THE CIRCUIT IS WORKING PROPERLY" to check whether it is working.
So, here are my results:
1. See that you have 2V+/-0.3V across R1 combinations. That says the shunts CCS currents are normal.
At my 10 Ohm resistors (hot-rodded version) I measure 1.62 and 1.60 volts.
2. Check that you hear the output relay clicking one time 3-4 seconds after power on.
It is definitely clicking, yes.
3. See that you have less than +/-5mV DC offset on the audio outputs.
Strangely my multimeter shows 0.000V.
4. See that you have +Vout a bit more than - Vout.
+Vout is 9.65 and -Vout is 9.86.
LEDs all seem to work.
I connected a mesmerize I-Select for input selection. I enabled the first input by shortwiring the according contacts. When it is powered on the input is going through to the inputs of the B1. So, this should not cause the problem.
Any advice?
Attachments
1. Still a high enough CCS result to go by without issues. Very good match though.
2. Good sign
3. Weird. Make sure you measure each channel's output pad at DCmV DMM scale VS the middle output (ground) pad.
4. +V>-V is expected in the Mezmerize. You have a Hypno so any mismatch could go either way due to IDSS & VF tolerances
If (3) is still 0mV try at the 220R resistors next to the relay VS ground. That's the signal path to the relay. So to know if offset or test signal exists. If yes, the relay blocks. If no life signs before the relay, check there is ~10V D-S voltage on each signal FET in the quartet and their correct orientation.
Play a strong 300HZ sinewave from a lab gen or a digital source with no power amp connected to the pre and trace it from RCA input through the modules until it vanishes so to find the bottleneck point. Use the DMM in ACV if you haven't got a scope. There could be some wrong wiring arrangement or a bad part. There are many website signal generators or free software gen applications for PC audio and mobile phones.
2. Good sign
3. Weird. Make sure you measure each channel's output pad at DCmV DMM scale VS the middle output (ground) pad.
4. +V>-V is expected in the Mezmerize. You have a Hypno so any mismatch could go either way due to IDSS & VF tolerances
If (3) is still 0mV try at the 220R resistors next to the relay VS ground. That's the signal path to the relay. So to know if offset or test signal exists. If yes, the relay blocks. If no life signs before the relay, check there is ~10V D-S voltage on each signal FET in the quartet and their correct orientation.
Play a strong 300HZ sinewave from a lab gen or a digital source with no power amp connected to the pre and trace it from RCA input through the modules until it vanishes so to find the bottleneck point. Use the DMM in ACV if you haven't got a scope. There could be some wrong wiring arrangement or a bad part. There are many website signal generators or free software gen applications for PC audio and mobile phones.
You cannot accurately measure 0, or 1, or 2mvdc output offset if you use a digital voltmeter that has a resolution of 1mVdc (0.001Vdc)
You should select a range that is ~ 10times more sensitive than the measurement you expect.
If you are trying to measure around 1mVdc, then use the 199.9mVdc voltmeter range. The LSD is 0.1mVdc (LSD = Least Significant Digit).
That will give reasonably accurate measurements of offsets down to <0.5mVdc and might be good enough down to 0.2mVdc
That is a general rule that applies to many measurements. Your resolution should be ~ 10% of the lowest reading you want to see.
Hey guys,
sorry for my late reply.
I have the DCB1 working now and I must say I am impressed. It sounds superb. I never expected a pre-amp to have such an effect. Everything sounds clearer and I can hear more details. Next step will be the case of the amp. Thanks for your help again!
sorry for my late reply.
3. Weird. Make sure you measure each channel's output pad at DCmV DMM scale VS the middle output (ground) pad.
You were right of course, I measured 1-1.3 mV.
Play a strong 300HZ sinewave from a lab gen or a digital source with no power amp connected to the pre and trace it from RCA input through the modules until it vanishes so to find the bottleneck point. Use the DMM in ACV if you haven't got a scope. There could be some wrong wiring arrangement or a bad part. There are many website signal generators or free software gen applications for PC audio and mobile phones.
Very good advise as well. I did as you said and found a wrong wiring of the audio signal. Last project I worked on was a mono amp and that seemed to have confused me.
I have the DCB1 working now and I must say I am impressed. It sounds superb. I never expected a pre-amp to have such an effect. Everything sounds clearer and I can hear more details. Next step will be the case of the amp. Thanks for your help again!
Hi Salas,
I got one DCB1 laying around that I want to use as -/+ 15V voltage regulator.
Is it ok to just add two red leds on both sides or I need to change some e.g.: resistor values additionally?
Is the performance of DCB1 used as voltage -/+ 15V regulator comparable to Salas BiB (in terms of noise)?
Regards,
Tio
I got one DCB1 laying around that I want to use as -/+ 15V voltage regulator.
Is it ok to just add two red leds on both sides or I need to change some e.g.: resistor values additionally?
Is the performance of DCB1 used as voltage -/+ 15V regulator comparable to Salas BiB (in terms of noise)?
Regards,
Tio
Yes you can add extra Leds to boost the Vrefs and also use an adequate voltage trafo. No resistors manipulation is needed when using only Leds as voltage setting elements.
I would use three Leds* in series on thumbnail sized matrix boards and wire the whole things end to end at one of the original five Leds pads on the main board per side. Leaving one empty if first populating or pull out one original Led from each side if its an already assembled board.
Noisewise its comparable to the BiB when using same Vref filter capacitors (at the Wimas area there is space and pads for film or electrolytic). Currentwise its more capable than an unmodified BiB due to its big MOSFETS.
*Maybe mixing some green with the reds also, depends on what brand of Leds are in hand. If its important to reach the 15V goal exactly or to match the other side well, it will take some iterations in Leds picking. There are the usual VF and IDSS tolerances at play here too. If its a Mezmerize board it will also need a small easy trick not to oppose your matching efforts of its voltage sides. Because its rigged to force a small asymmetry. In that case let me know to explain more. If its a Hypnotize board do nothing further.
I would use three Leds* in series on thumbnail sized matrix boards and wire the whole things end to end at one of the original five Leds pads on the main board per side. Leaving one empty if first populating or pull out one original Led from each side if its an already assembled board.
Noisewise its comparable to the BiB when using same Vref filter capacitors (at the Wimas area there is space and pads for film or electrolytic). Currentwise its more capable than an unmodified BiB due to its big MOSFETS.
*Maybe mixing some green with the reds also, depends on what brand of Leds are in hand. If its important to reach the 15V goal exactly or to match the other side well, it will take some iterations in Leds picking. There are the usual VF and IDSS tolerances at play here too. If its a Mezmerize board it will also need a small easy trick not to oppose your matching efforts of its voltage sides. Because its rigged to force a small asymmetry. In that case let me know to explain more. If its a Hypnotize board do nothing further.
Thanks Salas for explanation. I do not need exact 15V I need matched -/+ @ ~15V on both rails. I know how to match leds, so I will try to do it first and then let you know. Thanks!
No can do, ordered it from Toroidy and they fabricate only on order :/ but I think can use it as aux psu in other projects. Quite expensive trafo to power fans etc
