Bottlehead Forum

General Category => All things MagneQuest => Topic started by: Doc B. on June 30, 2014, 12:46:06 PM

Title: TL-404 production run
Post by: Doc B. on June 30, 2014, 12:46:06 PM
Mike pinged me last week to let me know that he is preparing to do a very limited run of TL-404 autoformers. For me this is one of Mike's two or three top products, a killer headphone output transformer. So you just might want to get in on the run...acrosound at aol dot com
Title: Re: TL-404 production run
Post by: johnsonad on June 30, 2014, 01:24:32 PM
He has 6 pair available. No low Ohm taps. Get them while you can!
Title: Re: TL-404 production run
Post by: mcandmar on June 30, 2014, 01:50:29 PM
These sound interesting, wish i had the knowledge to put them to good use.
Title: Re: TL-404 production run
Post by: xcortes on June 30, 2014, 01:58:55 PM
You got your pair Aaron?

Pity about the no low ohm taps. I will one day build speaker amps with these (mi three pairs DO have low ohm taps).
Title: Re: TL-404 production run
Post by: johnsonad on June 30, 2014, 02:03:03 PM
Xavier, I didn't buy a pair as they can't drive the Altec's.
Title: Re: TL-404 production run
Post by: Doc B. on June 30, 2014, 04:54:13 PM
My guess is those are maybe coils left from the very first run Mike did, as my early TL-404s that went into the big amp did not have lower impedance taps, but later runs did.
Title: Re: TL-404 production run
Post by: johnsonad on July 02, 2014, 12:57:38 PM

Can you talk a little bit about the best application of the TL 404 please?  I would like to use a pair with 4 Ohm taps in a 45 amp build.  Is there a disadvantage to this and if so, can you please explain?


Title: Re: TL-404 production run
Post by: Lobo on July 09, 2014, 09:55:57 AM
Has anyone tried the TL-404 in the Mainline amp and could give feedback if/how this could be done? I learned from the forum that the Mainline is a design derived from a "big headphone amp" which apparently incorporated the TL-404, so they may also be of benefit in the Mainline. From a response by Doc B. to a similar question regarding incorporating the TL-404 in a Smack, I learned however that it may be not as easy as just swapping the TL-404 for an OT-3. There seem to be several issues; differences in grounding and loosing the option for balanced output beeing two of those.

So I would be really thankful to get feedback beforehand if it would likely be worth taking the challenge to replace an OT-3 by a TL-404? If yes, and if it was not as easy as soldering wires/drilling holes/enlarging case, which I could easily do myself, and understanding that more complex electronic consulting may take quite a bit of valuable time by the bottlehead designers, I would also be happy if could "consult" bottlehead in order to get the right instructions - if this was a project that could be realistically done on the basis of the stock design and be likely worth a try.

Thanks in advance for any help on this!
Title: Re: TL-404 production run
Post by: Paul Joppa on July 09, 2014, 10:30:40 AM
Mainline has a correctly grounded primary so an autoformer such as the TL-404 could be used. The output switching will of course be quite different, and as observed there is no way to get a balanced output. Be aware that magnetic hum pickup in the output transformer is a possibility - obviously we have not experimented with layout and orientation to find a good arrangement. You could probably reduce the parafeed capacitor to something in the range 5uF-8uF, if you are replacing it with another type.

There are no instructions available from Bottlehead, and to the best of my knowledge we have never compared the transformers directly. The '404 is a great unit so I would expect it to sound better than the OT-3, but we have no data to test that assumption.
Title: Re: TL-404 production run
Post by: fullheadofnothing on July 09, 2014, 10:45:22 AM
Also Mike posted on the MQ forum that the run has sold out. It's unclear from your post if you actually ordered or not, so bear that in mind.
Title: Re: TL-404 production run
Post by: Paul Birkeland on July 09, 2014, 10:52:34 AM
Also, the TL-404 is freakin huge. 
Title: Re: TL-404 production run
Post by: xcortes on July 09, 2014, 11:03:00 AM
The TL 404 is awesome. The Big Pre with 437As and my Sony R10s is the best sound you can get. The best. Period.

I have three pairs of 404s with low ohm outputs to build something for my horns someday. Could be SR45s or maybe I try transconductance. Right now there's no time though.
Title: Re: TL-404 production run
Post by: Lobo on July 16, 2014, 08:27:51 AM
Thanks for the replies regarding my question about integrating the TL-404 into a Mainline amplifier! I have indeed ordered a pair of these transformers, so I will definitely give this project a try and report about my experiences (both technical and soundwise) on the forum. I am still a beginner (with a background in physics  though, but with no experience in electrical circuit design), so this will well take some time. Additionally, I am just building the Mainline, so I want to spend some time with the stock design first before changing it. (I am also feeling guilty to think about a relative layman doing modification of the OT-3 output, which is built on purpose, sophisticated and thought through, even before having actually heard it. It was just that the TL-404 seem so rarely on offer, are so beautifully built and get such rave reviews. And experimentation is half the fun of DIY...)

As for Paul Joppas comments (great, thank you!), I have done some thinking and reading and would like to post my thoughts (hoping some of this may be helpful to others, as well as perhaps getting some new ideas back):

1. magnetic hum-pickup(1): From pictures of other Bottlehead kits, I deduced that different transformers inside one device always seem to be oriented with the axes of the coil's windings perpendicular to each other. I interpret this as a simple consequence of basic physical laws: The magnetic field of a coil (with current flowing through the windings) should be strongest in the direction of the axis (that runs central through the wound core). The voltage this field induces in another coil is bigger if the magnetic flux lines intersect this second coil's wound core perpendicularly, rather than lying in the plane of the second coil's windings. So the induced voltage should be largest if the coils are close to each other and have the same orientation (like birds on a wire) and smallest if they are farthest apart and their axes are perpendicular. Since the orientation of the TL-404 coil axis relative to the mounting holes seems to be different than in the OT-3 (as far as I can see from images of the TL-404), I presume I will have to at least mount it rotated by 90°.

2. magnetic hum pickup(2): To experimentally find (or verify, see above) the best position for the tl404, I got the idea that I may just take one of the TL-404 transformers, hook it up with it's full (autoformer) winding to a voltmeter ac input and use it as an antenna to pick up the stray field of the power transformer. I could move it around close to the live amplifier (without big hazard, since it is not connected to the circuit) to get ideas at which position/in which orientation it is least influenced by the power-transformer's field.

3. new value of parafeed capacitor for use of TL-404 in Mainline: As I understand it, the coupling capacitor, output transformer, plate load and tube form a "network" that will let only certain frequencies pass to the ouput (i.e. it is a filter). I have read in a post by Paul Joppa (thread is here:,35.msg79.html#msg79 (,35.msg79.html#msg79) and there are serveral more for this topic) that there is a formula to at least estimate a good candidate for the parafeed capacitor value in case a plate choke is used. However, there is a constant current source in the Mainline, and (as I understand it) Paul states that there is no simple way of calculating an appropriate capacitance in this kind of circuit. So I assume the suggested 5-8uF for the parafeed capacitor in combination with the TL-404 is either from experience or from a more sophisticated model (which I think I have read somewhere Paul has). From my simple understanding of an LCR-circuit (oscillator circuit), the resonance frequency is proportional to the product of L and C. Since the TL-404 has a 5000 winding primary, and the OT-3 seems to have a 4000 winding primary, I expect the inductance L of the TL-404 to be higher. So this may be an approximative simple explanation for using a smaller capacitance C - to have the product of L and C not changed too much? In several posts by Paul Joppa, I read that also capacitance values between double and half the value that is calculated by the simple (yet in this case not applicable) formula should yield good performance. I also read that in the end, quantitative measurements and critical listening are necessary to find the optimum. So perhaps I should experiment with 2 or 3 reasonably priced capacitors of different capacity before probably getting a more "boutique part" with just the right capacity.

I do not want to "clog" the post with too much theoretical/hypothetical considerations and do not want to go off-topic. I am happy about feedback and promise to let you know about my practical findings as soon as I have verything ready and have had the chance to experiment.
Title: Re: TL-404 production run
Post by: Paul Joppa on July 16, 2014, 06:45:05 PM
Your observation about perpendicular magnetic axes is spot on. If you visualize the magnetic field lines around a solenoid, you can infer that this works best if the locations are also centered on each other. This latter effect is kind of variable in magnitude though. If you examine the SEX and Stereomour, you will see that the transformers and plate chokes are closely centered fore-and-aft, but less carefully in the vertical (axial for the output transformer and plate choke). This flexibility is so far entirely an experimental result for me.

Your experimental approach (item 2) is exactly how I have done it. It takes a lot of patience but is more informative than any words ever written!

The problem with parafeed circuits is getting the damping right. The output transformer has an iron core which dominates its performance, and its low-frequency impedance (without load) is more like a semi-inductance, with a phase angle usually between 30 and 60 degrees. This is obviously dependent on the particular core material and geometry. When there is an airgap (as in a plate choke) the inductance is reasonable dominant and when the load is resistive as in an amplifier driving a speaker well below its resonance frequency then a pretty good formula ties the choke inductance, parafeed capacitance, and load resistance together.

With a current source instead of a plate choke, there is not much of a resonance - the transformer's self impedance ps partly resistive, and the load (headphone) makes it more so. To get the lower capacitance estimate I just scaled the capacitance by the nominal load resistance squared without consideration of different coil and core parameters. For the Mainline, Doc B and PB did a lot of listening. I'm not a headphone guy so I sat those sessions out. You are well advised to experiment with capacitor values! I always suggest using affordable caps to experiment with, listening to the bass only. Once you find the right value then you can look at better capacitors to clean up the treble.

But even then, remember to give the caps some break-in, usually 50-100 hours of actual music with an actual load (so the cap sees significant AC current). I have no physical explanation for this phenomenon but far too much direct experience to ignore it.