Sometimes an amp designed for +/-24v when operated at higher voltages will exhibit a change in the harmonic distortion profile. It may have lower overall distortion but higher relative H3/H2. Lower voltage may have higher H2/H3, hence sweeter sound. Zman01 is right, the way to get to the bottom of this is to measure.
Here’s how:
Howto - Distortion Measurements with REW
Here’s how:
Howto - Distortion Measurements with REW
Now that i m happy with how the M2X sounded, i will proceed to house it into a proper chassis. Then i will pull out the Focusrite and log some readings.
Thanks X for the thread on Distortion Measurement with REW.
Thanks X for the thread on Distortion Measurement with REW.
A 38vac trafo makes 53vdc no load. Probably 49v under load, so with zero safety margin it might work but you might also get a big explosion at some point.
But 38vdc is fine for 50v caps. I think you need 64v caps for a 38vac trafo.
But 38vdc is fine for 50v caps. I think you need 64v caps for a 38vac trafo.
I actually misread the panel on the Trans. It was 36 vac that the transformer outputs. This :
https://catalog.triadmagnetics.com/...oroidal-medical-power-transformers/vpm36-6940
I connected it to power supply with 50 vdc volt 10000 micro farad power supply and smoke went up. I am ordering 63vdc ones.
Imgur: The magic of the Internet
https://catalog.triadmagnetics.com/...oroidal-medical-power-transformers/vpm36-6940
I connected it to power supply with 50 vdc volt 10000 micro farad power supply and smoke went up. I am ordering 63vdc ones.
Imgur: The magic of the Internet
Hmm....
Something doesn’t sound right, if hooked up correctly I don’t think you should see any smoke with a 36vac trafo and 50vdc capacitors. Could there be a short somewhere?
With 36VAC transformer very much you push those 50V caps to their limit. Also, it depends on who produced those capacitors. You should have at that voltage min 10V or so extra tolerance.
Double check how the primary side is wired up and measure your mains voltage. Was a cap wired up with its polarity reversed?
Using a variac to slowly ramp up the voltage while monitoring the output is the safest method to start up a new design. There is also the incandescent lightbulb method and the use of current limiting resistor for slow/safety start ups.
My final PSA would be to wear eye protection when your first fire up a new power supply. Explosions of hot electrolyte in the face won’t be pleasant...
Using a variac to slowly ramp up the voltage while monitoring the output is the safest method to start up a new design. There is also the incandescent lightbulb method and the use of current limiting resistor for slow/safety start ups.
My final PSA would be to wear eye protection when your first fire up a new power supply. Explosions of hot electrolyte in the face won’t be pleasant...
. There are two primary windings. red/black is primary. which is the correct way to wire it for a bipolar power supply like SLB?
1 - to use only one of the primary winding and leave the other pair of red/black loose
2 - to connect the two reds together and connect the two black together and then to connect them to the 115v mains?
Last edited:
thank you.
i havent been able to find any other mistakes. but i intend to double and then triple check everything before firing it up again.
Last edited:
Having built up 2 sets of SLB and had bench tested it at 3A resistor dummy load, it performed so well on the oscilloscope with very little ripples.
However, when I hook the amp up with the SLB, I got slight hums and noise on my F6 and USSA5, on M2X its worst. Having reached my wit's end, I seek help from Xrk. He shares with me this diagram of his wiring setup.
I followed his method and Voila!
All 3 amps are very quiet, M2X surprises me the most, it used to be the worst hum among all and now so quiet.
SLB is awesome for ClassA applications. Thanks, X!
However, when I hook the amp up with the SLB, I got slight hums and noise on my F6 and USSA5, on M2X its worst. Having reached my wit's end, I seek help from Xrk. He shares with me this diagram of his wiring setup.
I followed his method and Voila!
All 3 amps are very quiet, M2X surprises me the most, it used to be the worst hum among all and now so quiet.
SLB is awesome for ClassA applications. Thanks, X!
Hi,
Is the BOM project labeled "Smooth_Like_Butter_BOM_V1.2" the correct BOM for the dual rail SLB boards on Etsy? There are two choices for dual rail boards on Etsy, one labeled V1 and one with no version label. I ordered the boards that did not have a V1 label, and want to make sure this is the correct BOM for the board I ordered.
Thanks
Is the BOM project labeled "Smooth_Like_Butter_BOM_V1.2" the correct BOM for the dual rail SLB boards on Etsy? There are two choices for dual rail boards on Etsy, one labeled V1 and one with no version label. I ordered the boards that did not have a V1 label, and want to make sure this is the correct BOM for the board I ordered.
Thanks
V2 is just the production board with more Faston connectors for star ground. Other components are the same. V1 are the verification build boards 1.6mm thick, 1oz copper, no ENIG. They work well though.
Hello Meanie,
Can you post the wiring diagram for benefit of other folks?
Thanks
Balaji
Can you post the wiring diagram for benefit of other folks?
Thanks
Balaji
Having built up 2 sets of SLB and had bench tested it at 3A resistor dummy load, it performed so well on the oscilloscope with very little ripples.
However, when I hook the amp up with the SLB, I got slight hums and noise on my F6 and USSA5, on M2X its worst. Having reached my wit's end, I seek help from Xrk. He shares with me this diagram of his wiring setup.
I followed his method and Voila!
All 3 amps are very quiet, M2X surprises me the most, it used to be the worst hum among all and now so quiet.
SLB is awesome for ClassA applications. Thanks, X!
Fun with the SLB
Since I've already found the LT4320-based rectifier to be worthwhile, and have been contemplating CFP style capacitance multipliers in my power supplies for some time, it seemed natural to try the SLB power supply for an F6 clone build that is currently in progress. I was fortunate to get in at just the right time, and shipping was very prompt.
This is not my first rodeo with CFP-based capacitance multipliers. I have designed them into professional medical imaging (Ultrasound) equipment, as well as used a variation in two of my old Class A/B favorites, the Hafler DH-200 and DH-220. So if it looks like I'm taking a couple small liberties with the design, yes I am, but informed by previous experience.
I purchased two of the SLB boards to be used in a dual-mono F6. I have built one so far, to check out the design with my tweaks applied. Here is a list of the changes that were made, referenced to the SLB-APRIL-11-2019 schematic:
R1, R2: 18 Ohm
Q1 – Q8: IPP029N06N
C7, C8, C13, C14: SLPX183M050H9P3 (18,000 uF, 50V)
Q9, Q11: KSC2690AYSTU, KSA1220AYS
Q10, Q12: MJW1302AG, MJW3281A
R17, R18: 1.8 Ohm
I've been using the IPP029N06N Mosfets for my other ideal rectifier builds and they seem to work well. The FOM values look good on paper, and the devices worked favorably in my M2x. I haven't tried other parts yet, and the specified STP110N8F6 should be very comparable. I prefer to provide adequate supply filtration prior to a capacitance multiplier. Though the pass transistor(s) appear to multiply the capacitance at their base, it works better if they have some real capacitance to back them up. Hence the substitution of 18 mF caps for the 15 mF caps.
Besides having high gain, the pass devices need to have low Vce Saturation performance. This helps to maintain gain at the low voltage dropout which we prefer to use for this implementation of the CFP. The KSC2660 and KSA1220 are superior in this respect to the BD139 and BD140. Likewise the MJW1302 and MJW3281 are higher performance in this area. It may also be worthwhile trying the newer Toshiba replacements, TTA1943 and TTC5200. Finally, I prefer to keep emitter degeneration of the controlling transistors (Q9, Q11) to a minimum, thus reducing the 10Ω values to 1.8Ω. There are other means of controlling oscillation or other instability in the CFP, but they don't seem to be necessary here.
On my test bench, I hooked up the SLP to an Antek AS-3222 and used it to drive a 15 Ohm load on each of the outputs (PVout & NVout). The transformer performed as specified driving the inputs of the ideal rectifier with a solid 22.2 Vac. Given the 1.7 Amp load, the CFP section exhibited a dropout of 2.1 Volts for a final Vout of 26 Volts with 2 mV ripple indicated on my DMM. A scope with AC coupling on Vout showed nothing remarkable, and supported the 2 mV ripple reading.
If there is any surprise here, it was that the rectifier + filter section only managed to produce about 28 Volts at the input of the CFP. I might have expected higher, but am still somewhat new to the LT4320-based rectifiers. Overall, I am pleased with the performance of the SLB, and plan to build the second board the same way as the first. What remains to be determined is whether additional bulk capacitance on the outputs of the SLB is beneficial or not. This needs to be tested by ear.
Since I've already found the LT4320-based rectifier to be worthwhile, and have been contemplating CFP style capacitance multipliers in my power supplies for some time, it seemed natural to try the SLB power supply for an F6 clone build that is currently in progress. I was fortunate to get in at just the right time, and shipping was very prompt.
This is not my first rodeo with CFP-based capacitance multipliers. I have designed them into professional medical imaging (Ultrasound) equipment, as well as used a variation in two of my old Class A/B favorites, the Hafler DH-200 and DH-220. So if it looks like I'm taking a couple small liberties with the design, yes I am, but informed by previous experience.
I purchased two of the SLB boards to be used in a dual-mono F6. I have built one so far, to check out the design with my tweaks applied. Here is a list of the changes that were made, referenced to the SLB-APRIL-11-2019 schematic:
R1, R2: 18 Ohm
Q1 – Q8: IPP029N06N
C7, C8, C13, C14: SLPX183M050H9P3 (18,000 uF, 50V)
Q9, Q11: KSC2690AYSTU, KSA1220AYS
Q10, Q12: MJW1302AG, MJW3281A
R17, R18: 1.8 Ohm
I've been using the IPP029N06N Mosfets for my other ideal rectifier builds and they seem to work well. The FOM values look good on paper, and the devices worked favorably in my M2x. I haven't tried other parts yet, and the specified STP110N8F6 should be very comparable. I prefer to provide adequate supply filtration prior to a capacitance multiplier. Though the pass transistor(s) appear to multiply the capacitance at their base, it works better if they have some real capacitance to back them up. Hence the substitution of 18 mF caps for the 15 mF caps.
Besides having high gain, the pass devices need to have low Vce Saturation performance. This helps to maintain gain at the low voltage dropout which we prefer to use for this implementation of the CFP. The KSC2660 and KSA1220 are superior in this respect to the BD139 and BD140. Likewise the MJW1302 and MJW3281 are higher performance in this area. It may also be worthwhile trying the newer Toshiba replacements, TTA1943 and TTC5200. Finally, I prefer to keep emitter degeneration of the controlling transistors (Q9, Q11) to a minimum, thus reducing the 10Ω values to 1.8Ω. There are other means of controlling oscillation or other instability in the CFP, but they don't seem to be necessary here.
On my test bench, I hooked up the SLP to an Antek AS-3222 and used it to drive a 15 Ohm load on each of the outputs (PVout & NVout). The transformer performed as specified driving the inputs of the ideal rectifier with a solid 22.2 Vac. Given the 1.7 Amp load, the CFP section exhibited a dropout of 2.1 Volts for a final Vout of 26 Volts with 2 mV ripple indicated on my DMM. A scope with AC coupling on Vout showed nothing remarkable, and supported the 2 mV ripple reading.
If there is any surprise here, it was that the rectifier + filter section only managed to produce about 28 Volts at the input of the CFP. I might have expected higher, but am still somewhat new to the LT4320-based rectifiers. Overall, I am pleased with the performance of the SLB, and plan to build the second board the same way as the first. What remains to be determined is whether additional bulk capacitance on the outputs of the SLB is beneficial or not. This needs to be tested by ear.
Hi Tungsten,
Thanks for the detailed account of your build. Sounds fun indeed. R17 and R18 sometimes depends on the amp you are driving. I have heard a screeching sound on certain Pass Class A amps with the SLB when the degeneration is too low. But sounds like you tried it on an M2? Anyhow, 1.7A AF 26v and 2mV ripple is super performance. Nice!
I was able to get 1mV ripple at 37v and 4A. So it may actually work better at higher currents? Although I was running a 3v drop across the pass transistor.
Cheers,
X
Thanks for the detailed account of your build. Sounds fun indeed. R17 and R18 sometimes depends on the amp you are driving. I have heard a screeching sound on certain Pass Class A amps with the SLB when the degeneration is too low. But sounds like you tried it on an M2? Anyhow, 1.7A AF 26v and 2mV ripple is super performance. Nice!
I was able to get 1mV ripple at 37v and 4A. So it may actually work better at higher currents? Although I was running a 3v drop across the pass transistor.
Cheers,
X