|
General Vintage Technology Discussions For general discussions about vintage radio and other vintage electronics etc. |
|
Thread Tools |
9th Jan 2021, 11:26 am | #21 | |
Octode
Join Date: Dec 2008
Location: Wincanton, Somerset, UK.
Posts: 1,784
|
Re: Transformer calculations app
Quote:
My thoughts also. The dimensions (or more accurately, proportions) of E-I style lams vary quite considerably (historically), in part driven by the relative cost of steel v copper. I suspect many modern transformers are "iron-rich", and it may be your app has this implicitly built-in. Which is not to denigrate it in any way, just add a note of caution regarding its universal applicability. John Last edited by John_BS; 9th Jan 2021 at 11:39 am. |
|
9th Jan 2021, 11:38 am | #22 |
Octode
Join Date: Dec 2008
Location: Wincanton, Somerset, UK.
Posts: 1,784
|
Re: Transformer calculations app
Eg, compare patterns 74 and 293.
|
9th Jan 2021, 1:02 pm | #23 | ||
Hexode
Join Date: Mar 2019
Location: Edinburgh, Scotland, UK.
Posts: 350
|
Re: Transformer calculations app
Quote:
For me, my lab tablet is the target device, it already has some useful free electronic apps on it. The app I'm hoping to build would be focused on ease of use, eg the ability to load presets with a single click. Eventually it will have the features to aid me in winding/re-winding mains power transformers and eventually output transformers/chokes for my hobby, valve based audio gear. Doug |
||
9th Jan 2021, 3:12 pm | #24 | ||
Hexode
Join Date: Mar 2019
Location: Edinburgh, Scotland, UK.
Posts: 350
|
Re: Transformer calculations app
Quote:
|
||
9th Jan 2021, 6:10 pm | #25 |
Hexode
Join Date: Mar 2019
Location: Edinburgh, Scotland, UK.
Posts: 350
|
Re: Transformer calculations app
Here's a pic of the winder with a 30x54mm bobbin, bit of play on the shaft though due to not having any bearings, secure enough for medium sized jobs. Not enough meat on the frame to drill out and fit bearings.
The plastic gears are only used for the 8x mode, for 1x everything is metal. The plastic handle was rough/sharp round the edges, I had to smooth it with 400 grit wet/dry sandpaper. The handle tends to bump into the other shaft and needs to be set carefully. When in 1x mode I had to file a flat edge on the shaft to secure the handle better. The shaft has a 10mm metric thread, has a centred hole at the end so could be secured there. |
10th Jan 2021, 12:33 am | #26 |
Heptode
Join Date: May 2005
Location: Adelaide, South Australia
Posts: 541
|
Re: Transformer calculations app
I've wound many transformers both mains and output using recycled laminations and did for a while use a program I had written in VB.NET, I now have an Excel spreadsheet that does the same thing but is a bit more customisable. For a hobbyist like me like it is not possible to have a complete range of wire sizes so there is often a compromise that needs to be made when it comes to the design - it needs to be made to fit the wire sizes I have.
In using a recycled core the quality of the lams is a bit of an unknown so I do a test to determine the max and optimum turns per volt for a given core, you can't always go by the original turns per volt - the transformer may have been designed to run close to saturation. Once you pull a few apart you find the current density in the wire varies from around 3A/sqmm to 5 or even 7A/sqmm - those transformers that have come from 40's, 50's and 60's gear are generally around 2A/sqmm While you do not not have to it is advisable to remove ALL of the old varnish from the lams - using solvent this is a messy and unpleasant job, I have heard you can throw the old transformer into a good fire and retrieve the lams once to fire has cooled, I haven't tried this my self but in theory it should work - the bobbin is destroyed but it is not that hard to make another. Getting the wire to fit ! I aim for a fill of 75%, I find if my fill calc goes much over 80% then it's almost certain the windings won't fit. I have a homebuilt winder that uses a stepper to drive the wire traverse, I have learnt to live with it's failings and now I've had some practise I can if I need to, wind transformers the 'old' way - no bobbin - paper insulation between each layer. Good luck with your windings, I'm sure you'll find it rewarding |
10th Jan 2021, 12:36 am | #27 |
Banned
Join Date: Jun 2015
Location: Mareeba, North Queensland, Australia
Posts: 2,704
|
Re: Transformer calculations app
Laminations are called "lossless" because when they are stamped, the knockout from a standard lamination becomes the I part of the E's and I's. Lamination 74 is a standard lamination. 293 is a wide lamination and is made that way usually for high voltages and it enables the many turns to fit with the increased insulation between layers, and between primary and secondary.
With lamination 293 the knockout becomes two I,s, and it makes them expensive. Joe |
10th Jan 2021, 10:42 am | #28 | |
Hexode
Join Date: Mar 2019
Location: Edinburgh, Scotland, UK.
Posts: 350
|
Re: Transformer calculations app
Quote:
I think the use of 'modern' silicon steel with its inherent higher flux density shows up when you re-wind the older transformers. The older lams seem to be a lot softer too, I think they were annealed that way. Once you have built with lams stripped of laquer, to you re-pot them? Doug. |
|
12th Jan 2021, 12:40 am | #29 |
Heptode
Join Date: May 2005
Location: Adelaide, South Australia
Posts: 541
|
Re: Transformer calculations app
I would think that modern commercial transformers would have better quality lams so they could save on copper which is more expensive than steel, if you are doing your own winds then it may not be that critical as you can always use a larger core if you have it. I usually repot in varnish only the windings and not the complete thing, I use ordinary floor varnish which may or may not be a good idea - I haven't had any issues yet. In breaking down transformers I have come across those that were not potted and also have seen plenty where the varnish has failed to penetrate right into the windings.
I repaired a Laney Guitar amp that had a failed output transformer - a partial short in the secondary, the owner was in a bit of a bind and wanted the amp back for a gig so I stripped the transformer down and rewound over a weekend, I stripped it down carefully to count the turns and note the winding layout, expecting to to see some sort of damage/burn marks but found nothing that resembled a short, I also found that the varnish had only penetrated partially - most of the inner windings were free of varnish. This lead me to think that maybe the short in the secondary was caused by the wires vibrating and rubbing through the enamel insulation. |
12th Jan 2021, 2:26 am | #30 |
Banned
Join Date: Jun 2015
Location: Mareeba, North Queensland, Australia
Posts: 2,704
|
Re: Transformer calculations app
Modern good quality lams will for sure handle higher flux density resulting in less copper. Less copper means higher losses and more heat. You only have to check some cheaper transformers in some imported equipment that run class D or E specification, which is way over 100C temperature rise. SO !!! what you gain on the roundabout, you lose on the swings.
Many modern transformers are "dipped". That means they are heated to about 100C in the oven, then simply dropped into a bucket of transformer varnish at room temperature. As the core/windings cool they cause a weak vacuum and suck some of the varnish into the assembly. They are then drained and baked in the normal way. A GOOD quality transformer today is made the same way it was fifty years ago: Remove most of the air from the assembly, then dip it into the varnish. As there is no/little air to cause air pockets the varnish flows into all of the assembly. Its expensive!!, its good quality, it retains more heat in operation, so usually they are conservatively rated. Conservative means more materials in all aspects of construction so a "GOOD" transformer can cost perhaps three times as much to produce. One of my biggest bugbears today with cheaper transformers is no interleaving between layers. ( They do have insulation between winding sections). If decent wire is used ( heavy formvar varnish) it is not quite so bad, but with any high voltage traffos, the best medicine is "to do it once, do it properly", and insulate it correctly. Its safer. In the long run it will be cheaper. ( no warranty claims). Many of you have seen old English made transformers, especially some of those that have almost 1/4" of deep varnish over the whole assembly. Some were made before WWII, and still beat megger tests against many modern "equivalents" with infinity or close too megohms. Joe |
12th Jan 2021, 3:48 pm | #31 |
Pentode
Join Date: Feb 2018
Location: Nottingham, Nottinghamshire, UK.
Posts: 229
|
Re: Transformer calculations app
Just wondering, as capacitance is inversely proportional to distance, in an audio output transformer, is there a practical point at which the negative effects of inter layer capacitance can be offset by using thicker insulation between layers?
|
13th Jan 2021, 9:34 am | #32 |
Banned
Join Date: Jun 2015
Location: Mareeba, North Queensland, Australia
Posts: 2,704
|
Re: Transformer calculations app
Capacitance is the dread of audio transformer designers!! Yes, we can make 20 sections and then have to deal with HUGE ( relatively ) interwinding sectioning capacitance.
You can ( as I do) use extra insulation between sections to reduce this, BUT it then needs a bigger core to fit the windings into the available window space. It comes back the the fundamental transformer theory ( and I quote Radiotron Designers Handbook ) A large core of ordinary steel laminations, is far better than a tiny core of very special laminations. A couple of bods in England have my output transformers, and as yet I have had no negative feedback about their performance!!. If you want " hi falutin kill em all performance" make em big!! Use LOTS of iron. Use BIG copper. section the traffo as if you were making it for yourself!!. GOOD engineering is just that. Microphone transformers are a special case. Especially today!! As are really good moving coil cartridge transformers. ANY capacitance will kill the high frequencies!! IF the capacitance is not calculated to suit the microphone or cartridge loading, performance will also effect performance!!. Its NOT a black art at all, BUT it does need a certain amount of pixie dust!! After all with carts and mics, mu-metal laminations are expensive. As is copper tape to wind primaries for the aforementioned transformers. Joe |
14th Jan 2021, 8:58 am | #33 |
Dekatron
Join Date: Jul 2009
Location: Sleaford, Lincs. UK.
Posts: 7,669
|
Re: Transformer calculations app
This link works for the exe.file download - https://app.box.com/s/w7jvgt2lrocmjo2qyojd9l2c68mrj7ii I've used the program in the past and it works ok, it's also useful for quickly calculating unknown trmrs power rating.
I take my hat off to you Doug for updating the program,nice work. One thing I did notice is you've entered 220v for your primary voltage,surely that should be 240v? 245v is more typical. Andy.
__________________
Curiosity hasn't killed this cat...so far. |
14th Jan 2021, 9:37 am | #34 | |
Hexode
Join Date: Mar 2019
Location: Edinburgh, Scotland, UK.
Posts: 350
|
Re: Transformer calculations app
Quote:
The mains voltage is usually 240-245 in my lab, I've set the default to 240 in my new app. The defaults will eventually be loaded from an editable xml file so can be set and save by the user. This week I've been testing the low cost winder I bought and I'll build a some test transformers using the old and new program. I'm also thinking about adding some extra features, eg calculations to estimate the resistance and voltage drop under load. If all goes well I'd like to add a choke option and then a OT option. I've had a look at the program OPT_da-322 by Yves Minmagnon for audio transformers, there is also a user guide in French (heureusemement je parle francais) at www.dissident-audio.com Doug. |
|
19th Jan 2021, 8:21 am | #35 |
Dekatron
Join Date: Jul 2009
Location: Sleaford, Lincs. UK.
Posts: 7,669
|
Re: Transformer calculations app
That's an interesting looking bit of software, I'll check it out.
Going back to Silvio's original tfmr software it is a bit crude, I've never used it to wind both primary and secondary but did this weekend, am sure there's not something quite right.looking into it a bit deeper the program used a turns per volt of 4.6 for both pri & sec,which seems odd, what about pri -sec losses? Yet to wind the secondaries will see how it goes. Have sent you a PM Doug about your program. Andy.
__________________
Curiosity hasn't killed this cat...so far. |
19th Jan 2021, 9:00 am | #36 | |
Dekatron
Join Date: Nov 2010
Location: Oxford, UK.
Posts: 4,993
|
Re: Transformer calculations app
Quote:
It has several windings, the most challenging is 1400 turns of #40AWG wire (0.08mm). It was a success with care, and three breakages which I terminated on spare pins on the bobbin I used. Impregnated with a mix of paraffin and bees wax. They aren't bad for the money, but demand care in setting up, and dealing with the spool of winding wire you are using. Craig |
|
19th Jan 2021, 1:09 pm | #37 | ||
Hexode
Join Date: Mar 2019
Location: Edinburgh, Scotland, UK.
Posts: 350
|
Re: Transformer calculations app
Quote:
For a proper job you have to unwind (throw away the wire) and start again. 6k turns is normal, I've gone as high as 9k. The breaks can be repaired if careful as there are only mV in the coils. It's possible to solder both ends of the 42SWG to a small piece of wire taken from a multi strand hookup wire, then coat with nail varnish. On the other end of the wire gauge, I've ordered some 18SWG to wind some air core inductors... Doug. |
||
28th Jan 2021, 12:07 pm | #38 | |
Dekatron
Join Date: Feb 2007
Location: Lynton, N. Devon, UK.
Posts: 7,088
|
Re: Transformer calculations app
Quote:
It arrived. 'You Get what you Pay For' is the phrase which comes to mind - quality of the casting is appalling. The handle is loose on its shaft. It's crude. But, it does do the job. I modified mine, the spindle had a long M10 threaded portion with a couple of cones and a knurled nut, to clamp a mandrel between the cones. I cut it off and threaded the end 3/8"-UNF to take a drill chuck. Greasing the (plain) bearings makes a difference, there are oil holes in the casting but I squirted grease, thinned with paraffin (which evaporates once in place), down the holes as oil being flung around into windings which may be impregnated isn't a good idea. I've wound an output transformer (1,600 turn primary) and to be fair, crude as it is, as stated above it did the job. And it is small and compact. I made a de-reeler for the spool of wire with a wooden frame, a length of 8mm steel rod, and a couple of ball bearings. Friction to stop it spilling everywhere is provided by a cloth pad against the spool flanges - and when winding, tension and traverse is by finger techniques. |
|
28th Jan 2021, 9:10 pm | #39 | ||
Hexode
Join Date: Mar 2019
Location: Edinburgh, Scotland, UK.
Posts: 350
|
Re: Transformer calculations app
Quote:
Doug. |
||
5th Mar 2021, 10:09 pm | #40 |
Hexode
Join Date: Mar 2019
Location: Edinburgh, Scotland, UK.
Posts: 350
|
Re: Transformer calculations app
An update, my 1st 'proper' transformer wound using my app for the calculations:
I entered the size of the core (width of E plate tongue x height of lams stack) into my valve amp power transformer program and then added the windings I required (for taps, leave the current required blank) The HT has multiple taps 0-120-220-360v @ 160mA, a bias supply of 50v @ 10mA and 2 x 6.3v: one for 3040mA (enough for 4 x EL84) and 3000mA (I entered these at 6.9v to account for the voltage drop under load). I plan to add a voltage drop feature to my app to calculate this based on the length of wire. The transformer laminations/hardware were kindly donate from a solid state amp (a Peavey 80 watt combo, I picked the amp and 2 guitars including a 1960s Japanese model for £20 on Gumtree a couple of years ago) Weighs in at 5lb 2oz / 2.3kg I measured the ac voltages with no load on the HT and bias, today my mains voltage was a wee bit lower at 235v (should be 240v) 0-120-218-359 and 48.5v So close enough, might have been spot on if the mains was at the usual 240v. On the heaters side, I measured the ac voltages without a load and with the load of a single EL34 drawing 1500mA (as per data sheet) 1) 6.87v no load and 6.3v under load 2) 6.86v no load and 6.2v under load photos attached Doug |