I think you are good there, I 've used my BA-3 FE to drive my BA-3 Power Amp into high efficency speakers. I had plenty of gain, but
it wasn't like that. usually something is mis wired or resistor value off somewhere. Seems like I've had tha ( volume way to sensitive) when I miswired a volume pot. Hope you figure it out.
Russellc
I did some testing regarding my tapping/pulsing noise with some measurements and I got strange results. I am still getting 34-36v DC coming out of the SLB power supplies, which is high as I was going for 32v. As a test, I disconnected V+, V- and Speaker Out from the RIGHT BA3 output board (D In was still connected to the FE), and left the other channel connected, so I could compare values.
One both sides, I have about .02v DC before the output cap on the FE boards, but on the outgoing side, the RIGHT channel has about +4 to 6v of DC! Wha?? How?? The LEFT channel (which is connected to the powered output board) has about -2.5DC on the output side of the cap. Both sides have about 30mV of AC on the output of the caps. I also repeated this test with two different brands of MM, two different wires, and tried from different ends of the amp in case I was picking up stray voltages.
Thinking of that, could I even be picking up stray voltages of that level? Could the pulsing noise be coming from the power supply caps as stray noise?
One both sides, I have about .02v DC before the output cap on the FE boards, but on the outgoing side, the RIGHT channel has about +4 to 6v of DC! Wha?? How?? The LEFT channel (which is connected to the powered output board) has about -2.5DC on the output side of the cap. Both sides have about 30mV of AC on the output of the caps. I also repeated this test with two different brands of MM, two different wires, and tried from different ends of the amp in case I was picking up stray voltages.
Thinking of that, could I even be picking up stray voltages of that level? Could the pulsing noise be coming from the power supply caps as stray noise?
That is odd.
Apologies for not reading all the way back. You said SLB supplies vs. supply.
- unloaded?
The right FE section is suspect. If you have 4V+ of offset, one or both of the MOSFETs was potentially being overdriven, but it may be within SOA. Was it for a long period of time? Any smoke?
I'd dial your FE bias back down to 0 on the right channel. Make sure that P1 and P2 still have the desired overall affect and "re-zero them". Then, swap the PSU being used for the good left channel to the right, and very slowly ramp it back up to see if the right will bias properly with the known-working-good PSU. Alternatively, you could use a bench supply if you have one. If you don't have a bench supply, and if you have a Variac, that would be very helpful to slowly ramp things up, so you're not switching on / off ... on / off if something gets hot or you see an unexpectedly high voltage in critical place.
That may help narrow it down to: Issue with FE and/or the PSU. After that, you can troubleshoot further.
Hopefully others will chime in with other suggestions / clarification. Good Luck!
Apologies for not reading all the way back. You said SLB supplies vs. supply.
- Dual mono?
- What is each positive and negative rail reading for both supplies?
- unloaded?
- What transformers are you using? Have you measured the VAC on the secondaries? Any oddities?
- Do you see any AC (other than a bit of ripple) on the rails?
The right FE section is suspect. If you have 4V+ of offset, one or both of the MOSFETs was potentially being overdriven, but it may be within SOA. Was it for a long period of time? Any smoke?
I'd dial your FE bias back down to 0 on the right channel. Make sure that P1 and P2 still have the desired overall affect and "re-zero them". Then, swap the PSU being used for the good left channel to the right, and very slowly ramp it back up to see if the right will bias properly with the known-working-good PSU. Alternatively, you could use a bench supply if you have one. If you don't have a bench supply, and if you have a Variac, that would be very helpful to slowly ramp things up, so you're not switching on / off ... on / off if something gets hot or you see an unexpectedly high voltage in critical place.
That may help narrow it down to: Issue with FE and/or the PSU. After that, you can troubleshoot further.
Hopefully others will chime in with other suggestions / clarification. Good Luck!
Sounds like a great idea. I have always wondered how this sounds with full bias and all 12 specified transistors per channel, it is a mid to long term plan of mine too.
I would not use a 2U. In fact, I would choose 4U. Specified dissipation for 6 deep config, like you want to build, is 300 watts. Divided by two, you get 150 watts per channel. Divided by 2, that is 75 watts per sink. I would say 4U, and no less than 3U. In a 4U, you can even squeeze a wee bit more Iq out of each device
Also, and you may have thought about this allready, I would suggest you buy your own devices and match up yourself, enabling you to reduce source resistor values a bit. The need for this of course dictated by speakers.
Edited to add: I can match up a close set of transistors for you and ship them over, Vishay not harris for the P channel part. Let me know if you are interested.
I wish you luck!
Regards,
Andy
I‘m always a bit insecure about the difference between ba-3 OS and f4 (is there any?).
I built my f4 into a minidissipante 4U 300.
The heatsink is about as big as a 3U 400 (but with better geometry/better temperature-distribution), and it gets around 50°C on full bias (275mV)…
There’s a fellow builder who managed to successfully stuff a full f4 into a 3U 350, but that’s said to be quite an oven (choose caps appropriately, avoid avoidable heat-sources etc!), and it is questionable if getting the pre-section into this too…
I built my f4 into a minidissipante 4U 300.
The heatsink is about as big as a 3U 400 (but with better geometry/better temperature-distribution), and it gets around 50°C on full bias (275mV)…
There’s a fellow builder who managed to successfully stuff a full f4 into a 3U 350, but that’s said to be quite an oven (choose caps appropriately, avoid avoidable heat-sources etc!), and it is questionable if getting the pre-section into this too…
No difference in OS. Difference is in input buffer, BA-2 OS being without it iot have modulary approach. Principal is the same: only current gain, no NFB, and in original circuit IRFP9240s provided some uncorrected sweetness/H2 at lower volumes and H3 at higher output levels. Nice and soft top end, slightly rounded, full and massive bottom end, and clear but a bit sweet midrange.
BA-2 OS being modulary, one can choose 3 deep config, meaning F4 OS basically. Or 6 deep, doubling the OS. Iot keep the devices in the same area of linearity, meaning keeping Iq the same for more devices, one needs bigger chassis. That’s not to say it won’t sound good with less Iq per device, but probably a little bit different
6 deep in big chassis can drive even lower impedances without burping.
Andy,Sounds like a great idea. I have always wondered how this sounds with full bias and all 12 specified transistors per channel, it is a mid to long term plan of mine too.
I would not use a 2U. In fact, I would choose 4U. Specified dissipation for 6 deep config, like you want to build, is 300 watts. Divided by two, you get 150 watts per channel. Divided by 2, that is 75 watts per sink. I would say 4U, and no less than 3U. In a 4U, you can even squeeze a wee bit more Iq out of each device
Also, and you may have thought about this allready, I would suggest you buy your own devices and match up yourself, enabling you to reduce source resistor values a bit. The need for this of course dictated by speakers.
Edited to add: I can match up a close set of transistors for you and ship them over, Vishay not harris for the P channel part. Let me know if you are interested.
I wish you luck!
Regards,
Andy
I tested both versions, the original and 6 deep, adding two modules on heat sinks into a 4U/400 frame.
From the beginning I designed my amp to test 3 and 6 deep.
Really, I don't need more power, I was just curious.
As usual my focus is the sound quality and also the capacity to drive low impedance speakers.
Basically, in my amp, the SQ doesn't change, confirming the outstanding performances in woofers control.
So I would definitely recommend 4U, or even 5U (pico‘s macho-box?)No difference in OS. Difference is in input buffer, BA-2 OS being without it iot have modulary approach. Principal is the same: only current gain, no NFB, and in original circuit IRFP9240s provided some uncorrected sweetness/H2 at lower volumes and H3 at higher output levels. Nice and soft top end, slightly rounded, full and massive bottom end, and clear but a bit sweet midrange.
BA-2 OS being modulary, one can choose 3 deep config, meaning F4 OS basically. Or 6 deep, doubling the OS. Iot keep the devices in the same area of linearity, meaning keeping Iq the same for more devices, one needs bigger chassis. That’s not to say it won’t sound good with less Iq per device, but probably a little bit different
6 deep in big chassis can drive even lower impedances without burping.
I like to run as much Iq as possible, so for 6 deep mono blocks, my choice would be 4U to achieve safe dissipation of 150 watts per channel. But I guess 3U will do, though haven’t checked it’s specs.
I think Picos was 4U500. Or did he have two types made?
Howdy! Happy New Year!
I'd be very careful using those dissipation specs. Gianluca published a study (in Pico's thread) showing how those numbers are generated. They're at the saturation of the sinks at some truly astonishing temperatures not conducive to my listening room.
My personal advice is to roughly double the dissipation spec (or de-rate the sink) for a useful number to get you around 29C above ambient... => 50C roughly in my situations. Better yet, just use ZMs rule of big toe (or thumb).
I'd be very careful using those dissipation specs. Gianluca published a study (in Pico's thread) showing how those numbers are generated. They're at the saturation of the sinks at some truly astonishing temperatures not conducive to my listening room.
My personal advice is to roughly double the dissipation spec (or de-rate the sink) for a useful number to get you around 29C above ambient... => 50C roughly in my situations. Better yet, just use ZMs rule of big toe (or thumb).
I meant 75 per sink, meaning 6 transistors per sink, in mono block config. Sorry, just went out cross country skiing. The forest always makes me think. And I thought: maybe I should have stated that clearer.
To subject starter: my suggestion was to use 2x4U400s for 6 deep mono blocks. Ending with 300 watts total dissipation, split by four, meaning 75 watts per sink.
Pics of what exactly? I can take pics tonight.
Yes, dual mono and I will post the voltages of both supplies soon as I have a moment to check them.
Transformers are Antek 3228s and yes, the voltages of the secondaries are right on point. 28 volts, give or take a few 1/10ths. I was aiming for 32volts DC assuming the transformers would sag under load, but as of yet, there is no sag so I am getting 34-36 volts out of the power supplies.
Happy new year to you all too!Howdy! Happy New Year!
I'd be very careful using those dissipation specs. Gianluca published a study (in Pico's thread) showing how those numbers are generated. They're at the saturation of the sinks at some truly astonishing temperatures not conducive to my listening room.
My personal advice is to roughly double the dissipation spec (or de-rate the sink) for a useful number to get you around 29C above ambient... => 50C roughly in my situations. Better yet, just use ZMs rule of big toe (or thumb).
Aha! This issue clears up a little. I had the impression that those numbers weren‘t a proper definition, but more because the emitting values (n Watts) were, uhm, not sufficient—there had to be a complicated formula to calculate that stuff 🥸…
@andynor - Hope you enjoyed your ski.
75W per sink (of that rough size) is much more realistic.
all good