output transformer for a heathkit w5m

[Shadowdog]

This isn't 10K primary like the originals , but has a 6.6K which is used in fixed bias for the 6L6/KT-66.

It should have more interleaving (likely similar to Dynaco) than Edcors so should be a better buy. You can also buy them from the manufacturer.

http://www.triodeelectronics.com/40-18080.html

On the other end of the scale Hashimoto has a 10K 40W one @ about $360 ea. Not sure on UL tap specs , but I think they will maybe build to order that way. If you want original sonics and used are too pricey they may be the way to go. Likely last with no worries about used dying !

 

[multichamp12]

I am curious too. To do it right, will require a significant amount of interleaving of the windings, as that is how the original was manufactured. That process takes much more time (and is therefore much more costly) than simply winding a complete primary winding, and then a complete secondary winding.

Multi -- Did you ever pass a 10 kHz square wave signal through your Williamson's to check the supersonic performance of your Heyboer transformers (overshoot, ringing, stability, etc.)? That would really tell the story as to how well they copied the original design. The A 8072 is a more difficult transformer to reproduce as its UL operation requires very tight coupling between the primary halves.

Dave

Dave, Unfortunately I have not checked the performance on my pilot aa901 clones, since I am slowly trying to piece together funds to get a scope and a signal generator. I wish I had someone who lived nearby who had the equipment and could show me how to do it. I'm glad I have this forum though.

The stancor a8072 transformers I have are originals, I've been lucky enough to get my hands on a few over the past few months.

As most know, the OT is the most important part of the Williamson amplifier and must be of very high quality, which makes me skeptical about trying to reproduce something like a peerless or acrosound OT. I think for a true Williamson amp, seeking the vintage OTs is the way to go.

 

[Tubologic]

Can we get back to the original reason I made this thread. Does anyone know a good replacement output trans for the w5m. One that doesn't cost more than amp is worth. And one that is readily available.

The rest of the service upgrade document for the w5m would be greatly appreciated also.

You won't find a (used) Peerless 16309 for cheap and even if you do it will be an old failure prone part which can die at any moment. (aggravating your losses) As you have probably noticed, rewinding is also an expensive option, without any certitude or guarantee that the original (mostly unknow) specs will be met. High-end and other "exotic" transformers costs more than your amp worth, then...
At $ 68.88 the EDCOR CXPP30-10K is probably the cheapest option and surely worth a try. It will work and the slight difference (40%-50%) in the screen taps ratio should not have a great impact (provided that the original is 50%, which isn't even sure). Maybe you'll need to experiment and slightly change the value of the NFB and 1st stage HF compensating caps ( 240pF and 270 pF respectively) for best HF stability and no ringing/overshoot, and rise the value of the NFB loop resistor (4K7) if any tendency to unstability is noticed. How well (or badly) the EDCOR will compare sonically to the original PEERLESS I can't say but there aren't really other cheap options available. Everybody here will tell you that compared to the Peerless the Edcor is an inferior quality transformer and the sound will be degraded etc... but if you don't actually try it you'll never know...
Also, I didn't checked if mechanically it will fit the W-5M and maybe some work will be needed in this area.
Another issue which hasn't been mentioned yet and contributes to the OPT premature failure in the W-5M (not the only cause, but surely a contributing factor) is the VERY CLOSE proximity of the OPT to the heat generating output tubes: with the original recommended KT66's the tubes enveloppe are nearly touching the transformer ! The two output tubes are dissipating about 42 watts at iddle and this extra heat doesn't do any good to an already over stressed old transformer. I would recommend mounting a heat screen (a simple L shaped metal bracket) between the output tubes and the transformer to prevent overheating it.

 

[cocobolo95]

When you mentioned the heat problem, I was thinking of mounting the new ot on a separate chassis. It won't be far enough away to cause problems, but far enough from the output tubes to keep it away from the heat.

That way I won't have to modify the original w5m chassis for a larger ot.

 

[trobbins]

You won't find a (used) Peerless 16309 for cheap and even if you do it will be an old failure prone part which can die at any moment. (aggravating your losses).

That's the advantage of a restoration that removes the chances of other parts failing (eg. coupling cap leakage that then stress the output transformer), and add extra protection if a part failure does occur (eg. secondary side fusing). An original output transformer can then whistle along for another 60 years of use.

 

[Tubologic]

That's the advantage of a restoration that removes the chances of other parts failing (eg. coupling cap leakage that then stress the output transformer), and add extra protection if a part failure does occur (eg. secondary side fusing). An original output transformer can then whistle along for another 60 years of use.

IF... the original transformer hasn't be already weakened by years of use in the unrestored donor amp with all the know issues. Untraceable, and too many unknows. I've seen more than one old second hand 16309 OPT's failing after a few days/weeks of use in a perfectly restored and controlled W-5M. It was already a failure prone part right from the start 60 years ago (Heathkit were aware of the issue and opted to change for a more reliable OPT), and is much more today. It's a gambling game and nobody will give you any long term warranty on a 60+ years old used part. At $ 25 I would take the risk, but certainly not at $ 150-200

 

[trobbins]

Are we confident of the failure mechanism(s), or just that they 'failed' ? The term 'weakened' implies that a failure mechanism involving some form of degradation was known about.

 

[Tubologic]

Are we confident of the failure mechanism(s), or just that they 'failed' ? The term 'weakened' implies that a failure mechanism involving some form of degradation was known about.

To be fair, I really have no idea about the exact failure mechanism involved and I didn't investigate any further but all three OPT's developped an open half primary winding which happened after awhile on restored amps where all the measures were taken to prevent any overload condition. (new coupling cap's, increased cathode resistor to limit static bias to 40 mA per tube, protection fuse in the main B+ line, etc...). I suspect that the gauge of the wire used for the primary winding was insufficient for the rated current and/or some long term chemical/corrosion effect "weakened" it up to the breaking point, but it's only a educated guess... open is open (broken wire or connection).

 

[cocobolo95]

On page 16 of the Williamson amp pdf that's available to download, it states his amp with kt66 output tubes requires an OPT with a 10K resistance. He also states his amp requires a larger OPT than other amps of the same wattage.

The part that scares me is when he says that an OPT that doesn't have the correct specs will not sound good.

I'm in the frying pan and about to fall into the fire. On one side, I can buy a 50+ year old used OPT and hope it holds up.

On another side, I can buy a rewound 16309 and hope the person winding it knows what he's doing and the proper specs for the OPT.

Finally, I can buy a new OPT and hope it is good enough to give good sound. Can't spring for a custom wound as it would blow my entire budget just for the OPT.

Anyway I look at it, I end up falling out of the pan and into the fire.

 

[soundmotor]

I just bought a heathkit w5m that is missing the output transformer. Does anyone know of a good readily available substitute for the peerless 16458, which amp originally came with.

Do you just have the single amp?

 

[cocobolo95]

Yes, just one w5m. Will do mono at first until another w5m comes my way at a good price. And I already have 2 Heathkit wa-p2 preamps to use with the w5m.

 

[Shadowdog]

Maybe a pair of the Hashimoto or similar wouldn't be so bad a deal for what the W5 amps and parts are going for and given the transformers will eventually fail. You might find another W5 for very little $ if OPT dead or missing or just replicate it if you have
the skills.

 

[cocobolo95]

I was reading an older thread from audio karma and one member said he used a Hammond 1650F OPT with good results. Any thoughts on this?

 

[primosounds]

Yes, just one w5m. Will do mono at first until another w5m comes my way at a good price. And I already have 2 Heathkit wa-p2 preamps to use with the w5m.

I don't think that having one amp with a different output trans is a good solution. You need 2 of the same OPT. Also unless your planning to have the WP2 preamps on a seperate power supply plugging them into the original power supply socket will add more stress to your amplifier putting it at greater risk. I would use a different preamp for your w5 instead of planning to power it off the main amp. Besides, unless the preamps are already rebuilt you would need to address the same rebuilding concerns with them as well.

 

[primosounds]

I was reading an older thread from audio karma and one member said he used a Hammond 1650F OPT with good results. Any thoughts on this?

If your original plan was to listen to a w5 with the altec transformers then using Hammond opt will not be the same. If you just wanted to build an amp that has kt66 or 6l6gc pp tubes then that would be a good choice. There are many options out there and one can get rather overwhelmed with the combinations available. The vintage circuits for 6l6 or kt66 are all proven classic designs and will give excellent quality audio. So, you must decide whether you want to wait or buy what is available and start building. Depending on your abilities, it might be prudent to learn the ropes on some not so esoteric parts before working on rarer components.

 

[Tubologic]

That original Williamson circuit wasn't UL like the Heathkit. It was triode strapped. Acro/Dynaco did a lot of research to develope their replacement trannies for Williamson amps and most of the builders back then went with the UL specs that came from that. I wouldn't put too much focus on the original 10K spec Williamson used unless you want to revert to an original triode strapped version.

This is a valid point. Actually no reliable informations can be found regarding the electrical specs of the (16309) OPT used in the W-5M and (afaik) these were never published in any Heathkit litterature. The only source (which has been repeated many times) is from a short mention by Mike Lefebvre (Magnequest) who says: 10K P-P primary with 50% screen taps. Could anyone here who actually measured a 16309 confirm this ? Or are there any other available informations to crosscheck ? Though the amp will probably work with any transformer with a primary impedance ranging from 6K to 10K and 40 to 50% U.L screen taps, these are still important basic datas needed to select the best substitution OPT in order to maintain the original performances of the amp.

 

[stanko]

This is a valid point. Actually no reliable informations can be found regarding the electrical specs of the (16309) OPT used in the W-5M and (afaik) these were never published in any Heathkit litterature. The only source (which has been repeated many times) is from a short mention by Mike Lefebvre (Magnequest) who says: 10K P-P primary with 50% screen taps. Could anyone here who actually measured a 16309 confirm this ? Or are there any other available informations to crosscheck ? Though the amp will probably work with any transformer with a primary impedance ranging from 6K to 10K and 40 to 50% U.L screen taps, these are still important basic datas needed to select the best substitution OPT in order to maintain the original performances of the amp.

I will see if I have any W5's with a working 16309. If I find one I will measure the impedance and post it. I will also check the two bad 16309's to see if perhaps the primaries are good (doubtful) and will measure one of those if I get lucky. Not today, I'm pretty sick from something going around up here, but ASAP.

 

[Tubologic]

I will see if I have any W5's with a working 16309. If I find one I will measure the impedance and post it. I will also check the two bad 16309's to see if perhaps the primaries are good (doubtful) and will measure one of those if I get lucky. Not today, I'm pretty sick from something going around up here, but ASAP.

That would be great to help clearing up the doubt for good. Thanks

 

[Steve O]

I've measured a few examples of both the 309 and 458. Both are nominally ~ 10K ohms A-A w/50% taps to 4, 9 (not 8) & 16 ohms. Actual is closer to 9.5 - 9.6K inc pri DCR factored in. Ratio meas made using 60Hz and sufficient primary voltage to produce 4VAC @ 16 ohm tap.

Another reference is John Atwood's "One Electron" site. He used an actual "universal" PP amp to make meas and comes up with somewhat different values than I or Lefevre. FWIW, Atwood frequently comes up with values different from mfgrs published specs so I consider his results open to ah.....interpretation. YMMV

 

[stanko]

That would be great to help clearing up the doubt for good. Thanks

You guys got my curiosity up, so I don't mind doing it. Steve O seems to have done the work already in the previous post, but I will see what I come up with, it's good practice anyway and easy to do.