Fairchild 255A

Brice

Super Member
Hi guys,

A quick post on a pair of Fairchild i just finished last week.
I didn't know Fairchild before, and never heard one of their product. To me it was a new adventure.

First, let's start of how they sound: fantastic !
Those little guys are extremely dynamic. They have a perfect integration top to bottom. Very clear and detailed sound, perfect for Jazz and vocals. They may lack some 3D, but the punch they have and the accuracy compensate. And they are loud. Small but loud.
I really enjoy listening to them since last week and I am sure they will make it often to the amp rolling addiction I am suffering.

I equipped the amps with Jan GE 6AB4, RCA clear top 12AU7, EH 6CA7EH, Sovtek GZ34.

What I have done to them:

1) Rebuild the bias supply and removed one selenium rectifier. I increased a bit the caps there.

2) Rebuild the PS with additional capacity. I used PSUD to find the best choice of caps. My findings was to use a KTL25 Authenticap as the big can, and additional film caps for the B+2 B+3 supplies. I used 50uF off the rectifier (30+20) and a 1uF film cap in parallel, then the rest of the KTL25 so 120uF (80+40), then 4.7uF film and finally also 4.7uF film. The original design had 40 40 30 30, and now I have 51, 120, 4.7 4.7. I also used high quality resistors for the supply. I added 2 diodes before the rectifier for added safety. I also added a fuse 250 mA sloblo on the B+ rail. The ripple has been lowered to less than a volt, as of before it was over 3 V.

3) I did not use the gain pot, so I re purposed it as a DC tube balancing pot, as on this model, there was only a bias and an AC balancing pots. So instead of the gain pot, I now have a 10K bias pot adjustable with a screw driver, feed by the bias supply, and on each side a 100K to ground, and with a 47K to each coupling caps. I had to tweak the last resistor of the bias supply to compensate the changes, so the 10K was boosted to 39K. The bias range is now -41V to -31V.

4) As mentioned, I don't use the gain pot. The RCA input goes straight to the grid of the driver tube via a 10K resistor and is grounded by a 470K. I removed the input cap as there is no need really as there is no DC present there (and of course my preamp does not leak DC :) )

5) Nothing has been done to the amp itself, other than replacing some out of spec resistors. I tried to re-use CC resistors where I could.

6) I use the amp on a bucking box as the voltage were higher to specs. Both power transformers are getting hot after a 4 hours session, so I prefer to use a small USB fan on them. Works perfectly. Both amps consume 77 watts at idle.

So in conclusion I now understand why Fairchild has such a good reputation. These are top quality amps. Well build, solid, and simple. And I think all the magic is happening in the Fairchild output transformer.

Here are my pics. Enjoy.

Brice.

20180606_173715_HDR.jpg

Original (almost)
20180513_115631.jpg

Now:
20180602_180744.jpg

THD full power test:
20180602_172827.jpg

1K square:
20180605_155748.jpg

10K square20180605_155802.jpg
 
Last edited:
@radioactive will like this post.

Brice those mono blocks look very nice, good job.

I have been enjoying some VAC vintage tube mono blocks of late. It's nice to find some power amps that integrate into a mac system and actually make it sound better with some speakers.
 
Nice work, Brice--nifty amplifiers! Square waves look good, too. I'm working on a pair of Bell 2200C's for a guy, so your work is inspiring me! Further inspiration for me to get acquainted with PSUD and use it for some other work I'm doing as well. I've heard about using film caps instead of electrolytics in power supplies--what does that get you sonically? I always do use films to replace electrolytics in speaker crossovers and feel the sound really benefits from this, but these are directly in the audio path.
 
Last edited:
Nice work Brice! I am drooling over these, just too cool. The EH6CA7’s are great bang for buck tubes, but I know my 8b went from sounding pretty good to sounding incredible when I swapped out the EH’s with some NOS Mullard XF2’s. Would love to hear these with my 248 pre someday!
 
Thanks guys.
Yes film caps help in the PS. Here is the PSDU simulation before and after the changes:
Fairchild-255-original.jpg

and after:
Fairchild-255-upgraded.jpg
 
Nice job Brice! Have you run a frequency response test on them? It would be interesting to know how those big 'formers do!

Dave
 
Thanks Dave.

No I haven't. What tests would you like to see?
Are you also interested in supersonic frequencies?
 
Great work Brice !!!nice to see another pair of Fairchild amps brought back to life with improved performance.
 
Brice -- Mainly supersonic, but from 1 kHz up to 20 kHz, and then on up to (say) 80 kHz or so all in 10 kHz steps.

Dave
 
Thanks radioactive, IPAMAN and all others.

Dave, I'll do it...when I take them out of my listening system.
For now I like them so much that it may take me some time to do it ;)

I'll report.
 
Thanks radioactive, IPAMAN and all others.

Dave, I'll do it...when I take them out of my listening system.
For now I like them so much that it may take me some time to do it ;)

I'll report.
ill be waiting with some:lurk: and:beerchug:.
 
So looks like you're approaching 2 db down at 20 kHz, with a smooth roll-off after that with no sub-peaks. That indicates that the unit should have very good HF stability. No doubt you could tweak the HF shaping circuits to produce a HF response that is much closer to flat at 20 kHz, while still having a smooth roll-off after that. In any event, the transformers have a very nice supersonic characteristic. Sometime, it would also be nice to know what the undistorted power response of the amplifier is as well.

Thanks for the info!

Dave
 
Sure no problem. It just took me more time than expected as I got side tracked with the Acro 20-20 (again :rolleyes:).
Yes this is a good response.
You mean playing the the NBF network to get a better response?
For the power response you mean measuring the max power produced by the amp before clipping (or a set %THD with a distorsion analyser) based on the frequency?

The square response at 1K and 10K were very good also:
20180605_155748.jpg 20180605_155802.jpg
 
The 10 Khz square waves show the effect of the steep roll-off in supersonic response. As a result, the rise time is compromised, which translates to the speed with which the amplifier addresses HF transient information. Ideally, for an acceptable compromise between rise time and stability, a HF response that is down no more than about 1 db at 40 - 50 kHz is desirable. With a measured response that is showing to be down 6-8 db at those frequencies, that is why the 10 kHz square waves show the significant rounding on the leading edge. The wave tops are nice and flat however, which again speaks to a high level of HF stability. The HF performance characteristics of an amplifier will always reflect some element of compromise. The trick is to combine the individual performance parameters into the best overall compromise package so as to reflect the best overall performance from the amplifier.

Undistorted power response is an indication of the band of frequencies over which full power can be developed while meeting the manufacturer's distortion specifications. It is often specified as being within +/- 1 db of the rated power output.

Dave
 
Back
Top Bottom