Hi Peter
I haven't dipped a transformer yet😉 so no debate, it was just a recommendation.
But I have protected stuff with varnish and so I did expect that it wouldn't air dry inside your transformer, as you seem to have found.
Fine for coil surface protection, I don't doubt that your motor rewind company use it successfully, but not what you need.
I have done a bit of research into low viscosity epoxy for another project.
Some low viscosity products just have some solvent/thinner added, not ideal for your requirement, or air quality.
But the manufacturer can actually use different resin and hardener chemicals with lower viscosity. There is quite a bit of freedom in the chemistry, only the "epoxy" link between the two components is locked in.
So low viscosity solvent-less epoxy is available.
I expect the viscosity would be sufficiently low, it's presumably possible to add a little epoxy thinner to make it even runnier.
If you choose an epoxy with a slow cure then you should have plenty of time. Slow cure epoxies also usually have better properties.
Or choose one that needs a little heat to cure, even more freedom to soak and then solidify.
Best wishes
David
Hi David;
Thanks for your patience with me. I have more or less committed to the electrical grade varnish. Someone mentioned that once it stops smelling, the varnish has dried. One could argue that the varnish in the deeper crevices can't dry. At the same time, it has be doubted that varnish penetrated deeply.
It's been 4 days since the first dip. It dripped for 1 day, smelled for 2 days, but I will wait a full week from the first dip, to dip again. (I believe) this first layer is the only layer that has contact with the enamel wire, so I will be extra patient with the dry time.
To remove my blinders for a moment, and review what I have learned;
A: The Crest may be noisy because it was not a major design factor, amp has noisy cooling fans anyway.
B: Line voltage may be too high. Amp and Toroid are labeled 120 VAC, and my line voltage is 124 VAC.
C: There may be noise or DC on the line. (Amp gets noisier when Microwave is used.)
I have two concerns;
My Yamaha P2200 (with toroid) is quiet in the same conditions, and my DC Blocker circuit have absolutely no audible effect.
May I ask again; can anyone suggest a method of measuring DC offset on the line?
Regarding safety issues, I have years of audio and film projection equipment installation and service experience.
Thanks for your patience with me. I have more or less committed to the electrical grade varnish. Someone mentioned that once it stops smelling, the varnish has dried. One could argue that the varnish in the deeper crevices can't dry. At the same time, it has be doubted that varnish penetrated deeply.
It's been 4 days since the first dip. It dripped for 1 day, smelled for 2 days, but I will wait a full week from the first dip, to dip again. (I believe) this first layer is the only layer that has contact with the enamel wire, so I will be extra patient with the dry time.
To remove my blinders for a moment, and review what I have learned;
A: The Crest may be noisy because it was not a major design factor, amp has noisy cooling fans anyway.
B: Line voltage may be too high. Amp and Toroid are labeled 120 VAC, and my line voltage is 124 VAC.
C: There may be noise or DC on the line. (Amp gets noisier when Microwave is used.)
I have two concerns;
My Yamaha P2200 (with toroid) is quiet in the same conditions, and my DC Blocker circuit have absolutely no audible effect.
May I ask again; can anyone suggest a method of measuring DC offset on the line?
Regarding safety issues, I have years of audio and film projection equipment installation and service experience.
I also have mains power that is substantially in excess of nominal: 253 V rather than 230 V or 240 V,
I connected my CRO to a small shunt resistor to look at the transformer current waveform, very distorted from core into non-linearity.
The shunt resistor blew from a turn-on transient due to transformer saturation before I checked for DC.
A current probe would have been safer.
So I feel your pain😉. I hope the varnish works for you and am very curious to see if it helps.
Best wishes
David
I connected my CRO to a small shunt resistor to look at the transformer current waveform, very distorted from core into non-linearity.
The shunt resistor blew from a turn-on transient due to transformer saturation before I checked for DC.
A current probe would have been safer.
So I feel your pain😉. I hope the varnish works for you and am very curious to see if it helps.
Best wishes
David
Last edited:
If you have a DC blocker on hand, connect an induction motor or incandescent lamp load through it. Then measure the DC voltage across the blocker with a hand held DC meter.
Noise was a prime consideration in the design of the Yamaha P series amplifiers. The fans are variable speed and only on demand. Played quietly with no significant heat load they don't run at all. The trafos are dead quiet - even when our line gets to 127 or higher which is often. That was the main reason I bought the P3500 for the home system - dead quiet unless the music is very loud.
Hi guys,
Sorry if I am putting you through a full circle (again). I did measure the voltages across my blocker with the amp as a load. I only saw 90 mV AC and maybe 1 mV DC, even had my wife turn on the microwave. I could retest with a resistive load (bulb).
wg ski; I just googled both Yamaha amps, The P3500 is (I believe) almost 20 years newer than the P2200. I like the old class AB amp because it seems to be very 'well behaved'. It is also not fan cooled but has large heat sinks, 4 rack spaces and (only) 230 Watts / ch. I use it to drive my 16 planar tweeters
Sorry if I am putting you through a full circle (again). I did measure the voltages across my blocker with the amp as a load. I only saw 90 mV AC and maybe 1 mV DC, even had my wife turn on the microwave. I could retest with a resistive load (bulb).
wg ski; I just googled both Yamaha amps, The P3500 is (I believe) almost 20 years newer than the P2200. I like the old class AB amp because it seems to be very 'well behaved'. It is also not fan cooled but has large heat sinks, 4 rack spaces and (only) 230 Watts / ch. I use it to drive my 16 planar tweeters
Attachments
Just a minor update;
The toroid has been dipped 4 times now. I was told by the very experienced teck at the shop, that one Vacuum impregnation is equivalent to 7 dips at atmosphere. I understand that there are benefits to Vac that can not be duplicated regardless of how many dips at atmosphere.
Not wanting to look a gift horse in the mouth (or just forgot), I did not get more info on the Electrical varnish that I received. It is obvious at this point, that it dries (cures) at room temperature, and that it is solvent based. From the visual results, my guess is that I will be dipping this thing at least 10 times. At 4 dips, I'm seeing nice filets between the wires, but it is far from the solid lump of varnish that I was hoping for.
There was a concern expressed that the interior would remain a 'gooey mess'. There are still voids in this toroid; the last bubble was detected an hour after the last submersion. My argument is that, as opposed to a closed paint can, the vanish found its way in, so will the oxygen, eventually. I am using an electric heater to accelerate drying, but am not allowing temperatures to threaten the external wires and connectors.
The toroid has been dipped 4 times now. I was told by the very experienced teck at the shop, that one Vacuum impregnation is equivalent to 7 dips at atmosphere. I understand that there are benefits to Vac that can not be duplicated regardless of how many dips at atmosphere.
Not wanting to look a gift horse in the mouth (or just forgot), I did not get more info on the Electrical varnish that I received. It is obvious at this point, that it dries (cures) at room temperature, and that it is solvent based. From the visual results, my guess is that I will be dipping this thing at least 10 times. At 4 dips, I'm seeing nice filets between the wires, but it is far from the solid lump of varnish that I was hoping for.
There was a concern expressed that the interior would remain a 'gooey mess'. There are still voids in this toroid; the last bubble was detected an hour after the last submersion. My argument is that, as opposed to a closed paint can, the vanish found its way in, so will the oxygen, eventually. I am using an electric heater to accelerate drying, but am not allowing temperatures to threaten the external wires and connectors.
Attachments
Maybe the fumes are getting to me but here's another couple of issues I'm struggling with;
A: Is the varnish dissolving layer(s) of previous dips? At this point I feel that the tranny should remain 'in the drink' until it stops bubbling. (An hour last time). This may be more understood by the 10th dip!
B: The Toroid was wrapped toroidally with plastic film. It is impossible to wrap this way again (see pictures) because the epoxy centre core remains. It was then wrapped around the circumference with a (vinyl) 2 inch wide adhesive tape. My varnish fumes enhanced thinking is leading me to believe that this amp will hopefully be back in use before the interior is 100% solvent free. The adhesive on the tape is dissolved be solvent, so a different out protection will have to used. I am thinking; some kind of 'sock' or stocking (can't seem to keep my wife out of this project), stretched over the transformer and saturated in varnish. Also; some kind of insulating disks top and bottom, between the windings and the cabinet, but I feel that windings should not be sealed, trapping any remaining solvent.
Another concern; The (mu-metal) outer shield was held in place with the same adhesive tape, which had let go, just with age. I believe the amp is '97 vintage. I will have to come up with a better installation/insulation method for that too.
A: Is the varnish dissolving layer(s) of previous dips? At this point I feel that the tranny should remain 'in the drink' until it stops bubbling. (An hour last time). This may be more understood by the 10th dip!
B: The Toroid was wrapped toroidally with plastic film. It is impossible to wrap this way again (see pictures) because the epoxy centre core remains. It was then wrapped around the circumference with a (vinyl) 2 inch wide adhesive tape. My varnish fumes enhanced thinking is leading me to believe that this amp will hopefully be back in use before the interior is 100% solvent free. The adhesive on the tape is dissolved be solvent, so a different out protection will have to used. I am thinking; some kind of 'sock' or stocking (can't seem to keep my wife out of this project), stretched over the transformer and saturated in varnish. Also; some kind of insulating disks top and bottom, between the windings and the cabinet, but I feel that windings should not be sealed, trapping any remaining solvent.
Another concern; The (mu-metal) outer shield was held in place with the same adhesive tape, which had let go, just with age. I believe the amp is '97 vintage. I will have to come up with a better installation/insulation method for that too.
Attachments
drill some big holes in the centre plug.
It will collapse enough to let you push it out since it will not adhere well to the Mylar insulation.
Then wrap new Mylar insulation ensuring you get overlap at the outer edge.
If you aopt 10% overlap you will need to go around twice to ensure you have TWO layers of insulation over your windings.
If you adopt 55% overlap, you only need to go around once, since this guarantees at least double layer insulation.
If you were insulating the primary I would recommend 55% overlap AND go around twice to get 4 layers of Mylar insulation.
The middle gets massive overlap compared to the outer perimeter and you end up with 6 or more layers insulating the middle. That's where your varnish will not get through to the interior windings.
It will collapse enough to let you push it out since it will not adhere well to the Mylar insulation.
Then wrap new Mylar insulation ensuring you get overlap at the outer edge.
If you aopt 10% overlap you will need to go around twice to ensure you have TWO layers of insulation over your windings.
If you adopt 55% overlap, you only need to go around once, since this guarantees at least double layer insulation.
If you were insulating the primary I would recommend 55% overlap AND go around twice to get 4 layers of Mylar insulation.
The middle gets massive overlap compared to the outer perimeter and you end up with 6 or more layers insulating the middle. That's where your varnish will not get through to the interior windings.
The potting in the center should however, prevent any buzzing in that area.
When I rewind a secondary, I insulate with mylar tape and a couple layers of friction tape (the cloth kind).
When I rewind a secondary, I insulate with mylar tape and a couple layers of friction tape (the cloth kind).
There's your problem. You need a pair of real JBL 2445😉
Should not be too hard since at least you are in North America.
I was very happy to find a pair, quite rare in Australia.
There is a distinction here.
Some protective coats simply dry, the solvent just evaporates and if solvent is reapplied then they redissolve or soften.
Some materials just cure, solvent-free epoxy for instance.
Some materials do both, they have a solvent that dries and then they cure, solvent based enamel paint for instance.
Do you know the chemistry of your varnish?
Cure reaction may not rely on oxygen at all, some cross link anaerobically, some of the one-pack polyurethanes are catalysed by trace moisture.
This will affect the chance you have a runny mess in the centre, or at least how much time it will take.
This depends on the answer to the question above.
Real mu-metal? It's dammed expensive, maybe just steel?
I have had a bit more time to study this problem, found some information from industrial manufacturers, more likely to be real than "audiophile" claims.
You have tried to immobilize the copper wire layers.
But the majority of the noise is typically from the silicon steel core itself.
It is difficult reach the core because the inner layer of wire is, of course, insulated - practically always by toroidally wrapped tape.
This insulation is vital for safety because it separates the primary and secondary, so it is usually liberally applied, substantial overlaps and several thicknesses.
This is why they usually use both vacuum and then pressure to force the varnish/resin to fully soak the core.
Will be very educational to discover how much reduction you achieve.
Blocked up air path should help, and probably better damped mechanical noise.
Best wishes
David
Thanks everyone for keeping things interesting while I'm 'watching paint dry'.
To address your last points first; The outer (steel) strip is only a concern for me as far as re-installing it without risk of making electrical contact anywhere. The original method of self adhesive tape, was not sufficient. Certainly won't work now with trace levels of solvent still present. As far as "mu metal", I can't be sure, but there is (was) absolutely no hum issue in the speaker output. I envision wrapping the toroid in several (cotton) cloth bands and locking the toroid, cloth bands and together with...........well you know, more varnish.
BTW, when I bought the amp, the adhesive tape had loosened, as mentioned and the band had expanded very slightly. I want an installation method that will prevent this, but would prefer not to permanently attach it to the windings...still mulling this over.
Regarding the less efficient RCF LF18X401. I am actually happy with the drivers. They have a P to P cone excursion of 50 mm, so will move more air than the JBL, just need monster horse power! My ideal driver choice would be the JBL 2242H. The JBL's are more than 3 times the cost. I can buy new RCF's for less than recone KITs for the JBL. I know you get what you pay for, but I believe JBL products pass through too many hands before they get to to final user.
Back to the Toroid; I have a read extensively now that core laminations are very often a noise source. This can't be confirmed on my toroid but I remain hopeful regarding my dipping for two reasons;
The epoxy centre (which is beyond my patience and skill level to remove successfully) did not appear to vibrate, according to my homemade stethoscope. I am convinced that the noise was emanating from or through the outside windings.
Secondly, I don't expect the varnish to penetrate the layers of steel in the core, however, even if this is the source, I hope that encapsulating it (sealing in or damping or both) may reduce the noise.
I was almost tempted to put the AC to it, to see if there has been any improvement, but instead; it's back in the drink for dip number 5! I have to leave some suspense to an otherwise tedious project.....maybe I should have bought the Crown!
Thanks again everyone for your responses, Peter
To address your last points first; The outer (steel) strip is only a concern for me as far as re-installing it without risk of making electrical contact anywhere. The original method of self adhesive tape, was not sufficient. Certainly won't work now with trace levels of solvent still present. As far as "mu metal", I can't be sure, but there is (was) absolutely no hum issue in the speaker output. I envision wrapping the toroid in several (cotton) cloth bands and locking the toroid, cloth bands and together with...........well you know, more varnish.
BTW, when I bought the amp, the adhesive tape had loosened, as mentioned and the band had expanded very slightly. I want an installation method that will prevent this, but would prefer not to permanently attach it to the windings...still mulling this over.
Regarding the less efficient RCF LF18X401. I am actually happy with the drivers. They have a P to P cone excursion of 50 mm, so will move more air than the JBL, just need monster horse power! My ideal driver choice would be the JBL 2242H. The JBL's are more than 3 times the cost. I can buy new RCF's for less than recone KITs for the JBL. I know you get what you pay for, but I believe JBL products pass through too many hands before they get to to final user.
Back to the Toroid; I have a read extensively now that core laminations are very often a noise source. This can't be confirmed on my toroid but I remain hopeful regarding my dipping for two reasons;
The epoxy centre (which is beyond my patience and skill level to remove successfully) did not appear to vibrate, according to my homemade stethoscope. I am convinced that the noise was emanating from or through the outside windings.
Secondly, I don't expect the varnish to penetrate the layers of steel in the core, however, even if this is the source, I hope that encapsulating it (sealing in or damping or both) may reduce the noise.
I was almost tempted to put the AC to it, to see if there has been any improvement, but instead; it's back in the drink for dip number 5! I have to leave some suspense to an otherwise tedious project.....maybe I should have bought the Crown!
Thanks again everyone for your responses, Peter
The outer guass band does not contact any other conductive surface.
The secondary insulation prevents contact in that direction. You just need to add outer insulation to prevent any outside electrical contact.
The strip should make electrical contact to itself as you wrap it around. The flux runs around the transformer. It's that ability to take escaping flux around that attenuates flux peaks that could affect nearby equipment.
The secondary insulation prevents contact in that direction. You just need to add outer insulation to prevent any outside electrical contact.
The strip should make electrical contact to itself as you wrap it around. The flux runs around the transformer. It's that ability to take escaping flux around that attenuates flux peaks that could affect nearby equipment.
The wet varnish should be a fairly good insulator.
You could apply power to the primary (well fused) to warm up the toroid. That will expand air which due to pressure increase will do it's best to escape. Then after a few hours let it cool down for another few hours. The atmospheric pressure will force in new varnish to replace the contracting air. This would be a poor man's equivalent to VAC/VAC injection.
You can repeat this wet heat/cool/heat/cool without removing the transformer from the dip tank.
This may be more effective than removing and draining 522times.
You could apply power to the primary (well fused) to warm up the toroid. That will expand air which due to pressure increase will do it's best to escape. Then after a few hours let it cool down for another few hours. The atmospheric pressure will force in new varnish to replace the contracting air. This would be a poor man's equivalent to VAC/VAC injection.
You can repeat this wet heat/cool/heat/cool without removing the transformer from the dip tank.
This may be more effective than removing and draining 522times.
Hi Andrew
Thanks again. I may be using the heat/cool effect, to some extent, because the toroid has usually hung in front of my portable heater for a day before subsequent dips in the cooler varnish.
Observations after dip 5; less bubbling from the interior, secondly; I will have to eventually shorten the submersion times because it appears that one hour in fresh varnish seems to dissolve previous layer(s)
For now, I am encouraged by a reduced bubbling the last dip. Do you have any suggestions regarding (readily available) material suitable for secondary insulation?
Thanks, Peter
Thanks again. I may be using the heat/cool effect, to some extent, because the toroid has usually hung in front of my portable heater for a day before subsequent dips in the cooler varnish.
Observations after dip 5; less bubbling from the interior, secondly; I will have to eventually shorten the submersion times because it appears that one hour in fresh varnish seems to dissolve previous layer(s)
For now, I am encouraged by a reduced bubbling the last dip. Do you have any suggestions regarding (readily available) material suitable for secondary insulation?
Thanks, Peter
Had an idea I would like to check.
The suspicion is that the transformer is a bit close to saturation, so what is it rated at (VA) and what mass?
I hope you put it on the scales before you soaked it, to check how much varnish it absorbed.
Or could you put it on now and estimate how much varnish has been added?
They are nice but optimised more for sound reinforcement - robustness, efficiency, and use in substantial numbers (so they couple).
The 4645 cabinet first used the 2245, which should have flatter response when used in just ones or twos.
That's what I have, typo in my previous post.
Best wishes
David
The suspicion is that the transformer is a bit close to saturation, so what is it rated at (VA) and what mass?
I hope you put it on the scales before you soaked it, to check how much varnish it absorbed.
Or could you put it on now and estimate how much varnish has been added?
They are nice but optimised more for sound reinforcement - robustness, efficiency, and use in substantial numbers (so they couple).
The 4645 cabinet first used the 2245, which should have flatter response when used in just ones or twos.
That's what I have, typo in my previous post.
Best wishes
David
Hi Dave and everyone;
Thanks again for taking the time to respond, it is great to have access to your vast expertise. The Diyaudio forum, and all the people who contribute, is a tremendous resource!
I spent considerable time comparing t/s parameters of many 18" drivers and (jumped) to the conclusion that the RCF was the best value. I think the Fs was 1 Hz higher but that 50 mm cone excursion was irresistible! IIRC, the pole pieces are 15 mm thick as opposed JBL's 12 mm, so a longer linear cone displacement also.
Dip number 6 is delayed, I'm off to look at a couple of dead power amps. Idle hands are the devil's service department.
Regards, Peter
Thanks again for taking the time to respond, it is great to have access to your vast expertise. The Diyaudio forum, and all the people who contribute, is a tremendous resource!
I spent considerable time comparing t/s parameters of many 18" drivers and (jumped) to the conclusion that the RCF was the best value. I think the Fs was 1 Hz higher but that 50 mm cone excursion was irresistible! IIRC, the pole pieces are 15 mm thick as opposed JBL's 12 mm, so a longer linear cone displacement also.
Dip number 6 is delayed, I'm off to look at a couple of dead power amps. Idle hands are the devil's service department.
Regards, Peter
the thickness of the pole piece does not define the stroke.
Xmax = (Voice Coil height - pole gap height) / 2
The thickness of the poles pieces does not even have to equal the gap height.
Some manufacturers do not use that formula and add on an extra factor to make their Xmax look bigger than their competitors.
- Status
- Not open for further replies.