since I'm in the market for replacement power supplies, I was thinking of replacing the two laptop supplies with a single supply, like the Meanwell LRS-200-24, but the ripple is rated at 150mv p-p. Is that high?
I was also thinking to add a filter (CRC? LC?) between power & amp in the hope it would reduce the ripple.
Kind regards,
Drew
I was also thinking to add a filter (CRC? LC?) between power & amp in the hope it would reduce the ripple.
Kind regards,
Drew
If you are using the 2A rated inductors, you may want to consider the 15 volt version of the LRS set to 18 volts out. I used separate LRS-100-15 supplies that way for my monoblocks and they work fine. Actually, I can get a solid 18.2 volts from each supply. No hum or noise, no cycling off, no overheating. No ripple noise but my speakers are only about 88dB. The amp channels sound good. I am running about a 1 volt drop across the inductors.
The LRS-200-24 minimum voltage setting is 21.6 volts. If I'm doing the math correctly, 1.6A through the inductor (1V drop) is about 27.5 watts output device bias dissipation at 18.2V supply and 33 watts dissipation at a 21.6V supply. If your heat sinks can handle the extra 5 watts (or more if you run a higher supply voltage) then you should be OK. My device temp is around 60 degrees C, which is the most I am comfortable with.
Balanced MoFo Zeus like
Maybe I'm hijacking this tread a bit. But I'm looking for some advice. Two years ago I've build a 159ZC based MoFo. Running with some big IXYS part @ 2.2 amps. This has been my daily amplifier (my bloody expensive KT88 triode amp has moved to the basement).
Of course I always want to "improve" things. And as mentioned several times in this tread (also by me): balancing seems a nice way to dump the output capacitors (pages 38 and 201 have some discussions on this).
A few months ago I found a "crashed" DIY amp. It had some huge heatsinks, 12 OPA627's, 12 FQA48N20's, 7 toroids and a lot of other stuff. All for 50 bucks. 2 of these toroids restarted my balanced MoFo quest. A few weeks later I found 6 nice Lundahl transfomers, including 2 LL1539 (these are big, high DBU transfomers).
The components used for my new amp are: FQA48N20 as Mosfet. 2SK170 as input Jfets (I have quit a lot of these). The mentioned LL1539 are used as input transformers. And the output "chokes" are the salvaged 500 VA toroids (2 *40 VAC output used as choke). The resistance is around 0.15 ohm per secondary. And I'm running it at 1 amp per Mosfet (but you can run close to zero amps: this coupled choke runs nicely Class B). Max output is currently around 25 Watts at 8 ohm and 35 at 4 ohm (and even a bit more at 2 ohm). Output impedance (measured) is depending on the idle current. With the 1 amp it is around 0.7 ohm. The Mosfets are matched, so no need for a DC balancing (one channel has a DC offset of 7 mv, the other channel 2 mv).
The 2SK170's are running at their IDSS. In this case 22 ma for the pair. This 22 ma is also the limit of these Lundahls: they are not intended to be used with DC current: running more of these Jfets in parallel will generate distortion because of saturation. And running these Lundahls with less current will limit the low frequency response (the LL1539 needs a low impedance). I tried using the primaries in parallel, but then you need even more current.
After thinking, simulating, building, thinking, simulating and building, I came up with this amp. This version is working nicely. The bass is really good. More pressure compared to the original MoFo (lower damping or higher current?). The soundstage is big and there is absolutely no fatigue.
But: the original MoFo has a better "imaging". Every instrument has air around it. This is a really, really nice feature of this amp. Addicting. And there is more sparkle. I cannot measure this (the balanced version goes also really high). But the original MoFo is a bit brighter.
My "new and improved" balanced MoFo doesn't do this. At least: not that good. Yes, it has more bass. And it is a bit less fatiguing (might be caused by the brightness) . But I really miss this pin point accuracy. Once you have heard this absolute 3D imaging, you miss it when it is gone.
What could be the reason for this? And is there a solution? The 2SK170 are original and used as intended (I know they can swing up to 44 volts, but they survive). The output choke is of course not a real choke, but I wouldn't expect difference in the 3D imaging (only in bass). So my guess is: the Lundahls are the limiting factor. They are not the most expensive ones (not cheap either). Would replacing them with higher grade transformers help? I don't know what sound difference there is between transformer grades.
I like my current Tube 6N6P SRPP preamp. Not balanced. But I don't intend to change this one. I can also try to replace the Lundahl with some other phase splitter, but this splitter must be able to handle a lot of voltage (without influencing the quality).
Does anyone have some advice (or opinions)?
Maybe I'm hijacking this tread a bit. But I'm looking for some advice. Two years ago I've build a 159ZC based MoFo. Running with some big IXYS part @ 2.2 amps. This has been my daily amplifier (my bloody expensive KT88 triode amp has moved to the basement).
Of course I always want to "improve" things. And as mentioned several times in this tread (also by me): balancing seems a nice way to dump the output capacitors (pages 38 and 201 have some discussions on this).
A few months ago I found a "crashed" DIY amp. It had some huge heatsinks, 12 OPA627's, 12 FQA48N20's, 7 toroids and a lot of other stuff. All for 50 bucks. 2 of these toroids restarted my balanced MoFo quest. A few weeks later I found 6 nice Lundahl transfomers, including 2 LL1539 (these are big, high DBU transfomers).
The components used for my new amp are: FQA48N20 as Mosfet. 2SK170 as input Jfets (I have quit a lot of these). The mentioned LL1539 are used as input transformers. And the output "chokes" are the salvaged 500 VA toroids (2 *40 VAC output used as choke). The resistance is around 0.15 ohm per secondary. And I'm running it at 1 amp per Mosfet (but you can run close to zero amps: this coupled choke runs nicely Class B). Max output is currently around 25 Watts at 8 ohm and 35 at 4 ohm (and even a bit more at 2 ohm). Output impedance (measured) is depending on the idle current. With the 1 amp it is around 0.7 ohm. The Mosfets are matched, so no need for a DC balancing (one channel has a DC offset of 7 mv, the other channel 2 mv).
The 2SK170's are running at their IDSS. In this case 22 ma for the pair. This 22 ma is also the limit of these Lundahls: they are not intended to be used with DC current: running more of these Jfets in parallel will generate distortion because of saturation. And running these Lundahls with less current will limit the low frequency response (the LL1539 needs a low impedance). I tried using the primaries in parallel, but then you need even more current.
After thinking, simulating, building, thinking, simulating and building, I came up with this amp. This version is working nicely. The bass is really good. More pressure compared to the original MoFo (lower damping or higher current?). The soundstage is big and there is absolutely no fatigue.
But: the original MoFo has a better "imaging". Every instrument has air around it. This is a really, really nice feature of this amp. Addicting. And there is more sparkle. I cannot measure this (the balanced version goes also really high). But the original MoFo is a bit brighter.
My "new and improved" balanced MoFo doesn't do this. At least: not that good. Yes, it has more bass. And it is a bit less fatiguing (might be caused by the brightness) . But I really miss this pin point accuracy. Once you have heard this absolute 3D imaging, you miss it when it is gone.
What could be the reason for this? And is there a solution? The 2SK170 are original and used as intended (I know they can swing up to 44 volts, but they survive). The output choke is of course not a real choke, but I wouldn't expect difference in the 3D imaging (only in bass). So my guess is: the Lundahls are the limiting factor. They are not the most expensive ones (not cheap either). Would replacing them with higher grade transformers help? I don't know what sound difference there is between transformer grades.
I like my current Tube 6N6P SRPP preamp. Not balanced. But I don't intend to change this one. I can also try to replace the Lundahl with some other phase splitter, but this splitter must be able to handle a lot of voltage (without influencing the quality).
Does anyone have some advice (or opinions)?
Attachments
Last edited:
It is a lot of text. So coffee might be needed.
My Title says "Balanced MoFo Zeus like". I think I know Susan Parkers website by heard..... Although my version is not copy-paste. She uses the output toroids also as transformer (so both primary and secondary). And she uses a small toroid as input transformer (for her Toroid version at least).
Interesting remark btw: you think that this "pin-point-accuracy" might have something to do with the THD spectrum? Didn't think of that. I'm shunting one of the secondaries already with a RC combination for a decent square wave response. Without this you will get a big overshoot. And shunting both secondaries will make it worse, in stead of better. I first listened to the amp without this RC, and it sounded a bit "sharp" (don't know if this is correct English). But the original fine tuning was done mono.
Last edited:
Slightly OT: Part number differences?
Hi all,
Apologies for the slightly OT post, but I don't see a NoobCorner anywheres.
I'm going to order more parts for my MoFos (which blew up), and one vendor shows two models for the IRFP250 mf'd by Infineon:
I compared the spec sheets for both, but I don't see any differences.
Can anyone else help with the differences?
Kind regards,
Drew
Hi all,
Apologies for the slightly OT post, but I don't see a NoobCorner anywheres.
I'm going to order more parts for my MoFos (which blew up), and one vendor shows two models for the IRFP250 mf'd by Infineon:
- IRFP250NPbF
- IRFP250MPbF
I compared the spec sheets for both, but I don't see any differences.
Can anyone else help with the differences?
Kind regards,
Drew
Drew,
only two cent, it seems Infinion/IR used the same chip in both for easier production/costs.
The Vishay is different.
Other members here can much better explain, but i will try:
- Compare Rds on
- Forward Transconductance - S
- Reverse Transfer Capacitance - Crss
also luck at: Vgs(th); Ciss; Coss
In that way to other interesting candidates could be: FQA28N15 (little better than IRFP)
And a more robust FQA40N25, which i have biased to 2,8A in my "Test" MoFo.
MoFo is a good base to try out different Preamps, Transistors in different bias loads and
different PSU configuration and listening to effects.
Mine is sitting aside because other projects for the while.
only two cent, it seems Infinion/IR used the same chip in both for easier production/costs.
The Vishay is different.
Other members here can much better explain, but i will try:
- Compare Rds on
- Forward Transconductance - S
- Reverse Transfer Capacitance - Crss
also luck at: Vgs(th); Ciss; Coss
In that way to other interesting candidates could be: FQA28N15 (little better than IRFP)
And a more robust FQA40N25, which i have biased to 2,8A in my "Test" MoFo.
MoFo is a good base to try out different Preamps, Transistors in different bias loads and
different PSU configuration and listening to effects.
Mine is sitting aside because other projects for the while.
Low voltage MoFo?
The write up indicates one could use a 12 volt power supply that would clip at a lower output. With 4 ohm speakers, how many watts could I expect with 12 volts and biased somewhere between 2 to 3 amps? Also, how high an input voltage could a 12 volt MoFo swing?
Thanks for any responses.
Jazzzman
The write up indicates one could use a 12 volt power supply that would clip at a lower output. With 4 ohm speakers, how many watts could I expect with 12 volts and biased somewhere between 2 to 3 amps? Also, how high an input voltage could a 12 volt MoFo swing?
Thanks for any responses.
Jazzzman
ideally, with inductive load , thingie can swing double rail voltage
however , real world thingie is swinging less
I wouldn't use 12V, even with follower construction , mosfet capacitances are so-so linear starting from higher voltages
so , for power calculations , count on double Iq and double rail as voltage envelope , then decrease that for at least 20%
output voltage is slightly lesser than input one , being source follower
however , real world thingie is swinging less
I wouldn't use 12V, even with follower construction , mosfet capacitances are so-so linear starting from higher voltages
so , for power calculations , count on double Iq and double rail as voltage envelope , then decrease that for at least 20%
output voltage is slightly lesser than input one , being source follower