F5 Listening Impressions & Discussion

Member
Joined 2004
Paid Member
DC100 darker

PS. An interesting note which speaks to the subjectivity and sensitivity of all of this stuff to the specific and particular system & implementation is that in the main F5 thread a poster said that his implementation of Borbely's DC100 amplifier was "darker" than his F5. My own experience with the DC100 is the opposite, the F5 seemed to have less sibilance than the DC100 - which BTW is a rather nice , clean, and somewhat warmish sounding design (properly implemented).

I will revisit the F5 vs DC100 (2sk176/2sj56 3 pair/ch) 3again before it's conversion to use the F5 driver. I have done a test using the B60's 2sk135/2sj50(one pair/ch) outputs with the F5 driver and found the Hitachi's a little more detailed/resolving. By darker I refer to the space in which the performance was recorded. With the F5 the hall/wall reflections are revealed and the spaces between performers, like the stage lighting was turned up to illuminate more? The detail was also greater with the F5. The DC100 does have a WIMA 3u3F cap at its input, I've been wondering if it might be responsible for the darker sound. The DC100 driver is built with the Old Colony Boards/Kits from 1987. I have found that when I have it running dual mono it has greater detail/resolution as does increasing capacitance. Also inserting a 2m2H coil between the first and second caps results in better resolution. Currently setup with a single tranny.
No surprises there. So, yes running the DC100 as dual mono with a macho cap bank would increase resolution. ;)
 
Darker is usually used to refer to less bright sounding...

I'm wondering what you heard WRT sibilance between the two amps?
And ar eyou putting the choke in the ground lead between the cap sections?

You can remove the 3.3ufd input cap (bypass) IF you have no DC of any significance going to the input, since the amp will amplify DC. It will sound more better.

Or you can put a "botique" cap like a film/oil in and see if that does anything...

I think I'd just get some more mosfets before stripping out the DC100, unless maybe it is sitting on a plywood board or something akin to that.

_-_-
 
So no one else has any listening impressions about any of the F5 variations?? :Ohno:

Good or not good, have no fear...

Please don't feel that it isn't reasonable to not like this amp more than some other amp...

Going back to what I said at the top, the local experience so far is that the power supply implementation appears to have a significant influence on the perceived character of the amp.

I am quite curious to know what others have experienced thus far.

_-_-bear



PS. also going back to the top, the lack of capacitors in the signal path is an attractive feature to me, and might make it possible (given the simplicity) to identify the "source" of certain sonic "factors".
 
So no one else has any listening impressions about any of the F5 variations?? :Ohno:

Good or not good, have no fear...

Please don't feel that it isn't reasonable to not like this amp more than some other amp...

Going back to what I said at the top, the local experience so far is that the power supply implementation appears to have a significant influence on the perceived character of the amp.

I am quite curious to know what others have experienced thus far.

_-_-bear



PS. also going back to the top, the lack of capacitors in the signal path is an attractive feature to me, and might make it possible (given the simplicity) to identify the "source" of certain sonic "factors".

x2.
 
I have a Quad 606/Quad 34/Harbeth SHL5 and a diy F5/B1/Harbeth3ES. I have made some upgrades to the Quad 34 using Burson op-amps and Mundof ZN/Blackgate NX caps. Quad was recognised for its clarity and uncoloured sound in the 1980s.

The F5/B1 was a diy-project 3 months ago when my friend highlighted this project. The F5/B1 was a real surprise - its effortless presentation, spot-on soundstaging and clarity with a hint of warmth. I spend more time listening to the F5 now.

The man is a GENIUS! Thank you Mr Pass.
 
Patrick,

Thanks I have been reading the F5 thread, but that thread is mostly about the contruction and circuit design. I thought it would be useful to separate out the subjective result from that part. Also the few subjective reactions that are there have been spread sparsely throughout a very very long thread, hard to pick them out in a hurry.

So, it is not early on, what is your impression at this point? And, have you experimented with A) different brand outputs B) power supply variations C) resistor variations??

_-_-bear
 
I have not been listening for over a year, so my impressions have not changed.

I have built 4 F5-balanced in total, 2x with Fairchild and 2x with Toshiba MOSFETs. And I do like Toshiba more. But everyone here knows that and I am thus not entirely neutral. Toshibas just sound smoother to me, especially at the HF end. They have the magics of the laterals at HF and yet still have the control of the IRFs at LF. This summer I sold the 2 Fairchild boards, fully tested.

I did not play with power supplies, just CRCRC as I had lying around. Also resistors were as they were shown in the photos -- PRP 0.5Ws / 1Ws in parallel (up to 4x), and MPC74s for source degeneration. I don't think you have much more to gain by changing to Caddocks, etc. But Peter Daniels must be the expert on these.

;)


Patrick
 
Last edited:
Patrick,

Thanks I have been reading the F5 thread, but that thread is mostly about the contruction and circuit design. I thought it would be useful to separate out the subjective result from that part. Also the few subjective reactions that are there have been spread sparsely throughout a very very long thread, hard to pick them out in a hurry.

So, it is not early on, what is your impression at this point? And, have you experimented with A) different brand outputs B) power supply variations C) resistor variations??

_-_-bear

Hi bear,

I've built 2 F5's.

F5 1.0 - full circuit on CViller's PCB, tech-diy kit parts, Fairchild mosfets.
(CRC) 300VA antec, panasonic power caps, home depot copper wire, 4.7uF sonicap bypass, custom bridge rectifier w/ 10R-0.1uF snubbers.

F5 2.0's - Fairchild mosfets. Monoblocks with 300VA antec, mundorf HC caps, 16g silver wire, custom bridge no snubbers(yet), no bypass caps (yet). Point to point with similar grade parts as the tech-diy kit. Deleted thermal compensation and protection circuit portions. Also left out R9 and soldered the jFets directly to the female RCA, although do this with caution as leaving out R9 may cause oscillations depending on your source.

Impression:
2.0 has a substantial (as in no way its a placebo effect) improvement in detail and smoothness. Much better separation of instruments. Ambient detail also had a richer quality. I could practically hear the dust coming off the floor with the foot stomping on "Grandmother Song" by Vienna Teng.:)
Obviously too many changes to determine the value of each.
 
  • Like
Reactions: 1 user
Hi bear,


F5 2.0's - Fairchild mosfets. Monoblocks with 300VA antec, mundorf HC caps, 16g silver wire, custom bridge no snubbers(yet), no bypass caps (yet). Point to point with similar grade parts as the tech-diy kit. Deleted thermal compensation and protection circuit portions. Also left out R9 and soldered the jFets directly to the female RCA, although do this with caution as leaving out R9 may cause oscillations depending on your source.

Alazira,
Thanks for sharing your F5 impressions. If it is not too much bother, could you post some pictures of your build? I'd be interested in your PCB and PS component layout.
Thanks,
David
 
I should not comment as I seem to have reached the limits of my speakers and sources, but since not many others are, I will.

I have built two sets of F5's, one with Fairchild parts as designed and one with two pairs of IRF parts and slightly higher rail voltage, slighlty lower bias and a generally larger power supply. And a little more attention to detail.

I honestly can not hear a difference between them except at the limits, they both sound slightly better than my vintage SS gear, but doing blind A/B tests between either of them and my Adcom GFA-555ii - I can't tell a difference, unless I go beyond the limits of the F5. Anytime I think I can hear a difference I can switch and not hear a difference. I have the impression that the single output Fairchild F5 is sounds a little bit better, but I can not quantify it and again, any time I swap the other one sounds just as good. I can identify which of my preamps is in use pretty quickly and accurately, but not these amps. It seems like I should be able to hear a pretty big difference between a class A mosfet and a class AB BJT - but . . .



I have tried a number of different speakers, but nothing really high dollar - I suspect I need to change my speakers and/or my room before any more amp tweaking. My daily use speakers they are connected to are a set of Infinity QLS 5's (F5), and a set of Infinity RSIIIa's(F5+). I have also tried using my ZaphAudio SR71's that are used downstairs and a set of Infinity Beta 20's. My guess is that none of these speakers are revealing enough for me to hear the difference between the different output mosfets. Each speaker set has a very distinct signature, but the amps all sound very much the same through them.

The one thing that I have learned from this is that a good low power amplifier is totally sufficient for daily listening and higher than expected SPL. Unfortunately with my larger F5 I am drawing close to 300 watts at idle which kind of cancels out the typical advantages of a low power amp - ie low power draw and low A/C bill.

Another problem I have found is trying to find really well recorded music.
 
Last edited:
F5 built with Cviller boards, Tech DIY kits with Fairchild outputs, Antek 400va torroid, Vishay rectifiers, PSU-Panasonic TS caps & ERG resistors, wired with 14ga. copper and double shielded input cable, Vampirewire chassis connectors.

Listening with Yamaha C-2a preamp, Yamaha DVD/SACD as primary sources, Fostex 126e BLH, others.

This is one fine amp.
Listening to a 30bit Gold Master Disc of "Sketches of Spain"-Miles Davis, on the first track, I could hear sheet music falling to the floor when someone made an inept page turn. I hadn't heard that, previously.
Listening, initially, was hung up a bit by a previously unnoticed peak in the Fostex mid-range but these horns never fail to change. Getting past that, I noticed that neither bass or treble was missing or imbalanced in most any way. Resolution, speed, and all other attributes used to judge amplifiers seem kind of paltry. The F5 goes beyond all of that because, it seems, it does all of it so well. :grouphug: :D
 
F5 built with Cviller boards, Tech DIY kits with Fairchild outputs, Antek 400va torroid, Vishay rectifiers, PSU-Panasonic TS caps & ERG resistors, wired with 14ga. copper and double shielded input cable, Vampirewire chassis connectors.

what specifically is the 14ga connecting? i'm trying to figure out what gauge wire is needed in the PS for the F5 to perform at it's best - i'm talking about current. i have 12ga coming off my TX into block rectifers, and 14ga connecting the rectifiers to the PS. all wiring in the PS is done with solder-soaked stranded 18ga. i will probably use dry 18ga to connect the PS to the amp PCBs.
 
Hi reviewers,
I have built 3 x F5. Doubles monos, 2 x47000uF per side, no limiters.
For me F5 is "The" amp.
I will only mention i have noticed an improvement when using shottkies instead of common bridges.
Nex thing worth to try, i think, is a capacitance multiplier as PSU.(Lower ESR).
 
To add to the discussion on bandwidth:

Today I just completed F5 and here is the measurement I did on frequency response. The measurement was done on my HP 3577 network analyzer and is pretty same as Nelson's published specs DC to -0.25Hz at 200 KHz. I would just say this is awesome - flat to DC and -1 db at 252KHz in my measurement! What a amp. Only one thing I wish for - little bit more power.
 

Attachments

  • F5_fq_response.jpg
    F5_fq_response.jpg
    172.2 KB · Views: 2,710
Alazira,
Thanks for sharing your F5 impressions. If it is not too much bother, could you post some pictures of your build? I'd be interested in your PCB and PS component layout.
Thanks,
David

Hi David,

You can see pics here

F5 version2 pictures by Charlestonhood - Photobucket

or in the F5 thread, Post 5553

http://www.diyaudio.com/forums/pass-labs/121228-f5-power-amplifier-556.html#post1969043

No PCB used, although for the second monoblock channel I did use a teflon sheet. The only wire used on the circuit was 16g silver between the drains for the output. Check the signal path of 1.0 vs 2.0:

F5 v1.0 - RCA - solder point - wire - solder point - PCB trace - solder point - 1k resistor - solder point - PCB trace - solder point - jFet.

F5 v2.0 - RCA - solder point - jFet

Respected members Zen Mod & Magura have said a CLC power supply with an L >= 2mH offers a benefit as long as you stay in class A. Something I will try in the future (after trying Hitachi/Rensasis, Toshiba, and IRF cascode versions):) endless fun with the F5
 
There’s nothing special about the build, in fact its ugly as hell. Peter Daniel boards, K170/J74 and IRF Mosfets. This amp is by far the best I had in my system for years. Beats my AlephX hands down. There’s rolling and shaking bass, there’s high’s to die for and there’s silk sweet mids to make one drool. My 800's are –unlike other B&W’s- not demanding or power hungry. Like AR2, the only thing I could whish for is a bit more power.
 

Attachments

  • DSC00618.JPG
    DSC00618.JPG
    132.5 KB · Views: 2,719
F5channel2jFetsonRCA.jpg
Deleted thermal compensation and protection circuit portions. Also left out R9 and soldered the jFets directly to the female RCA, although do this with caution as leaving out R9 may cause oscillations depending on your source.

Hello Akazira,
Thanks for the reply and great pictures. Could you or someone else explain the purpose of R9? According Pass' F5 article, R9 is there to suppress parasitic oscillation of the input JFETs. How does R9 do this? I like the idea of elimination of parts if the result is better sound, but if the cost is possible failure then maybe not. What is the nature of the source that would cause oscillations or what should the source look like to ensure that oscillation will not occur?
Regards,
David
 
R9

Hello Akazira,
Thanks for the reply and great pictures. Could you or someone else explain the purpose of R9? According Pass' F5 article, R9 is there to suppress parasitic oscillation of the input JFETs. How does R9 do this? I like the idea of elimination of parts if the result is better sound, but if the cost is possible failure then maybe not. What is the nature of the source that would cause oscillations or what should the source look like to ensure that oscillation will not occur?
Regards,
David

Hi David,

As the F5 article says, the gate resistors for the jFets (R9) and mosfets (R13&R14) are there to prevent parasitic oscillations. I can't tell you why or how as I don't know myself.

I too like deleting things so I decided to try it sans R9 because I use a lightspeed attenuator, which is basically a variable resistor, so putting R9 in series with it seemed redundant as long as I keep the LSA from dipping below 1K (presumably).

For testing purposes you might try a 1K trim pot (shorting the middle leg to either side) for R9 and see at what resistance, if any, oscillations occur in your system.