For sure - I see that now. 😅
I selected those caps a while back. I started off thinking, as I often do, "I'll do the best of everything"! That was hubris. I have since slowly discovered (again, as I often do), that doing 'the best thing' requires more knowledge than I possess, and perhaps Mr Pass has that knowledge and that's why he's the guy designing these things and assembling the nice kits to make it easier for folks like me. If I had bought the full store kit and selected the 300VA Antek transformer I was supposed to I think I would be a lot further along than I am now. 😂
On the bright side, at least I give myself plenty of opportunities to learn.
That's actually pretty clever. I'll let wiser minds give it a look, but at first glance I see no critical issues.
pros - shorter mains AC wiring. AC still under the perforated base plate.
potential con - Mains AC proximity to DC. Still the current pulses through the secondaries (may???) have more energy than even the mains, so is that really a concern?
This is me trying to act like I know what in the world I'm talking about. My gut says that if we're concerned at all about EMI from mains wiring coupling to the DC lines (rectifier to filter boards) that the proximity to the secondaries literally connected to the same rectifiers would be more of an issue. So, is it a complete nothingburger of a concern?
So... again, I think you've come up with a clever layout, but I'd let a few people that "really" know their stuff chime in, particularly since I am pretty sure you don't have a scope to nose out any potential noise.
In my case, it's because I'm A) Lazy ... B) Not smart enough to have considered your layout.
Unless someone chimes in to say ... DON'T... I'd try it.
Yes... nicely done. 🙂
No. 18AWG is totally fine. I updated the guide to be a little less cavalier on the subject. Wire gauge definitely does matter, I was only trying to convey that colors and # of strands per bundle aren't things that I'd recommend agonizing over once a suitable wire gauge was chosen. I like 16 / 18 for all mains / DC supply / Speaker. I like a nice, flexible multistrand with an almost soldering-iron-proof insulation. Some people like solid core, and can manage to get a gorgeous connection without melting their PVC insulation.
I like repurposed CAT6 / CAT5 for signal. If I were to recommend buying wire vs. using scrap from things I've used in the past - I'd probably buy 24AWG solid core. Why? It's not so tiny that I want to kick something when I inevitably break it. Small enough to (maybe) get those advantages I read about from using 'finer' gauge solid-core wiring for small signal. I have to be honest though, I've tried a few different wires for line-level I/O. With the exception of choosing a nicely shielded wire (like some of the Mogami / Canare options) ... I truly don't understand all the hoopla. YMMV. That's why I am trying to keep my opinions out of the guide, but I'll happily share them here in the forum. Yeah... it's odd, I know.
It's a relatively short run.
Can I with any level of 'true' knowledge say to use the link below, no. Have I used it in the past to give myself the (perhaps false) confidence that 18AWG wire would be perfectly sufficient in that location / in those circumstances, yes.
https://www.omnicalculator.com/physics/dc-wire-size
Last edited:
I just consult the "Maximum amperes for chassis wiring" column from this table: https://www.powerstream.com/Wire_Size.htm and then do a little arithmetic
MyPrimaryAmpsMax = (TransformerVA) / (MainsVoltageRMS)
MySecondaryAmpsMax = MyPrimaryAmpsMax * (Vprimary / Vsecondary) . . . . . . . divided by "N" if there are "N" identical secondaries as in the Antex AS-3218
Then I look in the table linked above, to find the wire AWG number whose "Maximum amperes for chassis wiring" is at least 2X greater than MySecondaryAmpsMax.
example: 678VA transformer, primary=230VRMS, two identical secondaries each 49VRMS
MyPrimaryAmpsMax = (TransformerVA) / (MainsVoltageRMS)
MySecondaryAmpsMax = MyPrimaryAmpsMax * (Vprimary / Vsecondary) . . . . . . . divided by "N" if there are "N" identical secondaries as in the Antex AS-3218
Then I look in the table linked above, to find the wire AWG number whose "Maximum amperes for chassis wiring" is at least 2X greater than MySecondaryAmpsMax.
example: 678VA transformer, primary=230VRMS, two identical secondaries each 49VRMS
- MyPrimaryAmpsMax = 678 / 230 = 2.95 amperes
- MySecondaryAmpsMax = [ 2.95 * (230 / 49) ] / 2 secondaries = 6.93 amperes per secondary
- DoubleItForSafety = 13.85 amperes per secondary
- TableSays: AWG-19 has max amps for chassis wiring = 14. I would use AWG-19 or larger diameter. Voila, the diyAudio default value of AWG-18 happens to be even safer in this case!
With questions like "where should the AC wires go?" "where should the rectifiers go?" "where do the primary wires go?" "where do the secondary wires go?" "where does..?" "where does...?" .... I like to answer with another question. WWPD? (What Would Pass Do?) On the factory FW amps (million picture examples on the interwebz..), you almost always see the noisiest devices / wiring all the way up front, near the front panel. This is far far away from everybody else that may not like the company of those noisy neighbors.
The FW PS boards use some wide central PCB traces as a convenience bridge / connection point to transfer the incoming AC to the donut primary wires. I've never seen that arrangement on any of the DIY PS boards, so just running the sleeved / secured AC mains wiring beneath the steel base plate seems to make the most sense on the typical build.
Just a note regarding chassis wiring ampacity. The insulation of wire will have a temperature rating. Ampacities in the NEC are given for 60C, 75C and 90C. That said, most of the PVC hookup wire I run across is listed at 105C. PTFE is much higher, typ 200C.
I apologize for the length of this email in advance.
I finished building my 5FM about a week and half ago and getting the bias set around .52. Initially, I struggled with equipment and speaker matching as I found the F5M to be way too bright for me. That surprised me as in my rotation of speakers is a DIY open baffle speaker with a super tweeter and a Dahlquist DQ-10 both of which are not detail shy. Neither of these is a good match for the 5FM as they are low efficiency, especially the Dq-10, so I did not drive these with the F5M. I did test the 5FM on a high efficiency 15 inch full range speaker which I regularly pair with a flea watt amp with great success, but this was a bad match as well. The highs seemed "grainy" to me, for lack of a more articulate way of saying this.
After about a week I eventually pulled out my pair of CSS Criton speakers which I don't love as I find them to be a bit dull and low on treble detail. I also didn't think they would be a good match as they are not efficient. Much to my surprise it is a great match. However, I thought things still sounded muffled. As a last resort I removed my Belles A21 preamp, which I love, and replaced it with Korg B1 Preamp, which is a nice preamp, but honestly doesn't compare to the Belles, and I was wowed by the results. Great tone, great levels of detail, nothing veiled about the sound, etc. I also retested the bias which was now .57 and stable. It had been stable at .52 for a week, so this surprised me. The amp was still was not uncomfortable to touch, so I left it.
I know this is a lot of detail, but here are a few questions for everyone:
Thanks for reading this tome.
T
I finished building my 5FM about a week and half ago and getting the bias set around .52. Initially, I struggled with equipment and speaker matching as I found the F5M to be way too bright for me. That surprised me as in my rotation of speakers is a DIY open baffle speaker with a super tweeter and a Dahlquist DQ-10 both of which are not detail shy. Neither of these is a good match for the 5FM as they are low efficiency, especially the Dq-10, so I did not drive these with the F5M. I did test the 5FM on a high efficiency 15 inch full range speaker which I regularly pair with a flea watt amp with great success, but this was a bad match as well. The highs seemed "grainy" to me, for lack of a more articulate way of saying this.
After about a week I eventually pulled out my pair of CSS Criton speakers which I don't love as I find them to be a bit dull and low on treble detail. I also didn't think they would be a good match as they are not efficient. Much to my surprise it is a great match. However, I thought things still sounded muffled. As a last resort I removed my Belles A21 preamp, which I love, and replaced it with Korg B1 Preamp, which is a nice preamp, but honestly doesn't compare to the Belles, and I was wowed by the results. Great tone, great levels of detail, nothing veiled about the sound, etc. I also retested the bias which was now .57 and stable. It had been stable at .52 for a week, so this surprised me. The amp was still was not uncomfortable to touch, so I left it.
I know this is a lot of detail, but here are a few questions for everyone:
- I have not heard a First Watt amp before, but it appears to me from reading through this thread that the First Watt amps all produce a lot of detail. What is enough or too much detail is a subjective thing and depends upon system pairing, personal preference, etc. so my question is simply is my impression correct? If not, then something may be off with my build.
- I have no sense as to how burn in impacts the sound of this amp. (And no, I am not looking to start a flame war on the topic, which I know has two distinct points of view.) I had this amp on a lot over the first week and a half and am curious how material a role this might have been in the sound improvement. Same question for the slight bias shift.
- Lastly, the Belles works great with my Parasound A21 so it makes no sense to me that it would not do the same with F5M. Not sure that this is really question.
- I built this with one Antek transformer even though I bought the dual power supply kit as that was what I had on hand and Antek was sold out on the same transformer. How much would adding the second transformer and converting this to dual power supply make on the sound?
- I have 4.7 Ohm Thermistors to replace the ones that came with the kit. With a Modushop Deluxe 4 U Aluminum how high can I safely take the bias in concept? I know I need to monitor temp regardless of what "should" work. Is the change in sound linear as the bias is increased?
- Add in scrubber boards which I have on hand.
Thanks for reading this tome.
T
Last edited:
“Burning In”. : with a M2x 2 years ago and a Ship of Theseus recently, i felt there was indeed an effect after 10 hours or so of listenning.
Treble was a bit bright then better.
It is a feeling
(There might be other things at play but… don t know how to monitor them)
Treble was a bit bright then better.
It is a feeling
(There might be other things at play but… don t know how to monitor them)
Cheers. Yeah, I had the same inkling of a concern re: the mains and the secondaries, but like you say, it could be a nothing burger. I have trouble telling Nothingburgers from Whoppers. 😂
I haven't had much luck with solid core CAT5. That insulation is not meant to see heat, and melts if you stare at it too long. Bottlehead kits utilise the stuff a fair bit, and it does my head in. I tend to replace it with 24awg solid core Tensility brand stuff, as that insulation is a lot tougher. That said, for amp wiring, I sorta wish they made silicone insulated test leads in a variety of colours - the lengths and gauges would work well in an amp.
I'm thinking for the inputs I might just repurpose some shielded RCA cable I have lying around. It's the same stuff as this. I used shielded coax on my ACP+ inputs and outputs, and shielded solid core CAT5 (despite my feelings on its meltiness) on the input selector switch to the board, and that thing is so quiet that when I pin the volume on my most sensitive headphones I hear nothing at all. So both do the job very well. I am pro-shielding of some flavour on signal wires vs naked twisted pair.
Thanks for that, Mark. This is sort of the thought I had, but my math was wrong in that I didn't factor in the mains voltage before doing the rest. It seems like, following that formula, 18awg doesn't quite cut the mustard in my use case.
500VA/230VRMS = 2.17A
[2.17*(230/18)]/2 = 13.86A
Doubled is 27.72a
And the table says 15AWG.
Now Tortech, the manufacturer, have specced the secondary wires as 14awg, so no problem on their end. For my part, my question is "is that sort of amperage possible yet alone likely anywhere else in the build"? My thought was that, if anywhere, you might see it between the bridge rectifiers and the supply board as the caps charge, as it would be seen by the transformer as a near dead short for a moment. But then, the thermistor resistance (10 ohms before the transformer, 40 ohms after it) should slow that considerably, thereby reducing the draw through those wires on inrush. Is that right? Is it even in the ballpark?
I hear you. I initially wanted to do that Omega mount on the transformer and have it vertical, and move everything noisy forward a lot, but the manufacturer (Toroid) doesn't ship here as far as I can see. I also thought about bolting the AC hookup and bridge rectifiers to the back of the faceplate to really move them forward, but decided that tapping threads into (but more likely ruining) my nice aluminium faceplate wasn't a headache I needed. 😂 I was also concerned about the amp being too front heavy and awkward to carry, so I settled on leaving the transformer flat and hoping the mag shielding does its job. I was immediately convinced that mains wiring under the base plate was ideal, not only because it's the shortest path, but also because the plate shields everything from the AC a bit. I've also considered running two gal steel dividers either side of the transformer and power supply board to further protect the signal stuff, but then that would impede airflow a bit. Maybe some of the modushop perforated internal risers? Hmmmmm.
I had seen that, yeah. I do wonder if that's the temp at which deformation starts, or whether it's the point at which PVC falls apart completely. I'd be inclined to take whatever the max is and halve it. That reminds me, this post on an EE forum was maybe a touch tongue-in-cheek, but it also helped me understand what the thought process should be: https://www.edaboard.com/threads/ho...18-awg-wire-and-a-20-awg.399865/#post-1721403
Anyway, thanks all for your help!
On your CAT5 cable, get a 'plenum' version and the individual wires will generally have teflon insulation and solve your melting problem
The 3.2A figure mentioned in the thread is for power transmission with allowed wire length far in excess of any chassis wiring and also factors in being enclosed in conduit and/or being bundled with other cable. The 16A figure I gave was for a 90C rated wire ran within an enclosure but NOT within raceway or tightly bundled with multiple current carrying conductors. Now, I have no idea what arbitrary length was used for the chassis ampacity ratings but 1' of 18ga wire has a resistance of less than 0.007 ohms... you'd drop 0.112V while pulling 16A and so not enough to heat the wire noticeably above ambient temps during normal operation
Nice!
Now the question is, do I have to buy a 305M pull box. 😂
This is a good reality check for me, thank you. I did know that this was all related to resistance/conductance, but somehow I lost sight of that fact at some point.
TomR:
If you are willing to try something more extravagant, I might suggest buying a set of regular F5 boards from the store. Then you could try the version that I wrote about in the earlier F5m thread. The use of high transconductance Fets in the output stage, in combination with a selected set of gate stopper resistors can be transformative.
The PSU you have already is likely sufficient, but I have also found that the use of motor run capacitors to augment each channel's supply capacitance to be a noticeable improvement.
If you are willing to try something more extravagant, I might suggest buying a set of regular F5 boards from the store. Then you could try the version that I wrote about in the earlier F5m thread. The use of high transconductance Fets in the output stage, in combination with a selected set of gate stopper resistors can be transformative.
The PSU you have already is likely sufficient, but I have also found that the use of motor run capacitors to augment each channel's supply capacitance to be a noticeable improvement.
Last edited:
Re transformers mounted verticaly...
L bracket from the store next door,
Re transformer mounted verticaly + bridges aswell...
L bracket + metal sheet from the store next door. Small side of the L on the base plate, long side of the L carries sheet and transformer, sheet is sandwiched between transformer and L bracket and the transformer bolt clamps everything. Bridges on the metal sheet.
As it happens, I am currently just bolting this together (well, with 2 transformers and 4 bridges) for an other amp...
So double what you would need... total cost is less than 20E (with all the screws and belleville washers), solidity beyond any doubt - the baseplate of the amp would be the next weakest part.
Not that you really need that for this build, but just in case you have a solution without needing to rework the front plate...
I hope this helps
Claude
L bracket from the store next door,
Re transformer mounted verticaly + bridges aswell...
L bracket + metal sheet from the store next door. Small side of the L on the base plate, long side of the L carries sheet and transformer, sheet is sandwiched between transformer and L bracket and the transformer bolt clamps everything. Bridges on the metal sheet.
As it happens, I am currently just bolting this together (well, with 2 transformers and 4 bridges) for an other amp...
So double what you would need... total cost is less than 20E (with all the screws and belleville washers), solidity beyond any doubt - the baseplate of the amp would be the next weakest part.
Not that you really need that for this build, but just in case you have a solution without needing to rework the front plate...
I hope this helps
Claude
And it looks like this (in my case) for 2 x 400VA Toiroidy transformers.
Metal sheet is at the back (will be cut to dimension and bored for the transformers and bridges bolts), brackets need some minor rework because my build (NON F5m!) is very tight... and ignore the plexi sheet with the metal sheet - they are for an other section of this amp.
Voilà!
Claude
...
Say that you buy 500$ worth Kiradashi knife
Suddenly you get an idea to write to involved, well known Japanese Artisan, you having an idea of implementing every possible thingie from biggest Swiss Knife .......
....
F. he will politely reply - **** off, you stupid Gaijin
choose your option; personally, I was given with enough of F, after some time I started thinking that maybe they all are not so funny, maybe funny is on my side
...
after that, you can only expect that your e-mail addy is added in block list titled "ZM and other Idiots"
Aah, Kimosabe... you missed another thread, where actually Nelson himself pointed out that in his next batch of projects he's separating the FE from the OP on his projects. It makes a lot of sense when you are designing amps.
Now, whether or not that becomes commercial, it's likely not something that the rich guys buying PL amps would care... but perhaps it might trickle down to the DIY world?
Driving the bridged F4s with a BA3 and a preamp is an indication....
BTW, I bought a very nice sharpener that uses belts. I got three guides so I can sharpen all the way from a jamon serrano knive, through Japanese blades all the way to Standard Western BIG chef's knifes... Even my BIG Chinese Cleaver.
Chop! Chop!
Well, I don't know why you have so much stuff in your bedroom.... we only got a big TV up on the wall and the Big Marantz, Speakers and DAC/Android stuff... As it is, and this bedroom has two dedicated AC lines, when I turn on the Marantz the LED night lights flicker. It's the big 125 watter...
But my home office... HAH, I got THREE systems in it now. And I managed to have two systems shared on TT in the living room.
The way it's going, if I ever build another house, I'll just build one HUGE SPEAKER and I'll move in inside of it.