help with CMOQ-2S

[buildafriend]

I have a CMOQ-2S output transformer
When I tried to use it, the 312 card did not work. Then I popped (almost typed pooped) in a new CMOQ-2S and the card worked. Through deductive reasoning, I think that this one is broken.
when I was moving I must have smushed a winding.. doh!
http://cinemag.biz/output/PDF/CMOQ-2.pdf
I repaired the leads and I think that they are okay. But inside the transformer might be a different story.
I used to have a signal generator and old tektronics scope, but I broke them.. at least they died in battle.
So, I grabbed my DMM and got cracking. It looks like brown and red don't show continuity.
DCR values:
brown to red - can't get a good reading but my DMM says something like 2M
orange to yellow - 18.4Ohms
green to blue - 18.5ohms
violet to grey - 18.5Ohms
Continuity tests
brown to red - open
orange to yellow - shows continuity
green to blue - shows continuity
violet to grey - shows continuity

Is there any way I can try to repair this?
What methodology is there for testing transformers on a DIY tech bench type budget?

Up next, I'll be probing around a broken EA2622.

 

[joaquins]

I'd just use it as 1:2 transformer, leaving out the brown red winding, for now days applications, usually 1:3 output in those devices aren't really useful, you will have 3.5dB less headroom but it isn't a big deal when going to a dad since most of the time they can't take the headroom of a 312, which is around +30dBu.

I don't know if it's the correct transformer for this circuit, API preamps uses a Si GO lams, this Ni lams gives more inductance (higher impedance) which is not needed for driving it from a DOA as 2520 and they saturate sooner, it's true that distortion is lower in Ni lams for lower levels, for higher levels in with impedance sources Si can have less distortion than Ni, for the same core config. Also be aware of the ringing on these transformers, load them properly, you should make some measurement to over 200kHz where the ringing is happening. The original 312 has a ringing on the low end caused by the output cap and the transformer inductance, which is solved with the resistance in parallel with the cap, but has no contemplation on HF resonance at the transformer caused by parasitic elements on it, a proper termination as a 600Ω load shown on the data sheet or maybe better a zobel network. For the low frequency resonance you should look and tune the resistor value, cause the inductance on this transformer is probably higher than the original.

JS

 

[buildafriend]

Thanks JS.

Good to know that I still have an option that makes this usable.

I lost my patience and started hacking. Hopefully I won't cause more damage than I already have. If I break it more, I'll just buy another one. I'll feel better knowing I tried to fix this, and learned in the process.

It looks like the outer windings are still making it to the leads. This makes me think that the damage is closer to the core. I'm a little apprehensive about going further than this.. I have no idea how to get the bobbin out from the lams without getting destructive.

 

[joaquins]

Is not easy, you certainly will destroy the windings and I don't think you will be able to find quadfilar wire, and with luck you would be able to make a trifilar winding in the space they do quadfilar, so you'll probably end up with what you have now, and you need 3 wires tied parallel by hand the whole process, not nice or easy, I've done that and ended up with a couple of nice transformers but I don't know if I would recommend, certainly not with this unless you just want to learn.

To take that apart you have to take the lams apart one by one, usually there is some varnish on it so you need to heat it up to be able to do it without braking them, but too much heat will change the properties of the lams which isn't good either... Maybe boiling water, I never did this with Ni lams so I don't know. once you got the lams apart, you could try to look if there is some wire opened but hard, if you unwind it you should count the turns, so you know it for rewinding, if not I would advice to throw your winding apart, get a new plastic bobbin and wind on it a new winding, maybe a 3 wire for the same wire gauge wind with all the turns you could fit in, I did this for both I made, I plan to do more, the second one I made it too tight and it blended the plastic so the lams didn't fit nicely so I had to take it into the chuck and start to force it in, it resulted pretty nicely and pretty close to a real API transformer in inductance and inter winding capacitance, but only with 3 wires instead of 4. I have some of those measurements around, but I put them on a topic where I write about a seminar with Saul Walker, where I measured one, and there is another topic where I write about my discrete mixer where I put something about those transformers I builded. You could try to make a quadfilar winding, but probably you will end up with less inductance so an option is to go with a smaller gauge of wire, but then you increase the winding resistance, which makes the same problem at the end of the day, both increases the low end freq and low freq distortion, and as I said before I think 1:2 is quite good for the application.

JS

 

[MatthisD]

That is the S (steel) version so if you remove the laminations it wouldn't be a big deal to damage one or two of them. Once they are out you can unwrap the tape and check the breakout leads for connection.

What happened to the input transformer?

 

[buildafriend]

thanks guys

im exploring ways to take it apart now. Since the lams are varnished together and go inside of the bobbin it's hard to get them off. I gently heated up an outer lam with a heat gun and got it off but it got destroyed. boiling seems like a decent option but it would make the thing pretty hard to hold while taking the lams off. it's looking like at this point ill end up ordering a new output transformer and just use this one for learning. dave at CM said that usually when things like this happen people just buy new ones, they are so cheap that it's not worth investing too much time repairing them.

i'm yet to mess with the input transformer.

 

[buildafriend]

I went with the boiling method, and said a prayer goodbye to mr sort of useable transformer as I sent it to its death. after boiling I beat it with a hammer. at this point im only taking this apart for educational purposes. soon im going to start writing stuff down so I can remember the actual anatomy. I wish I had CJ's coil counter/winder..

Im noticing that the lams support the transformer by entering the bobbin. do the lams essentially create the core? so if the lams are steel then it's a steel core transformer? 

 

[buildafriend]

moving along..

 

[buildafriend]

looks like the leads definitely made it out.. any damage must be from within the transformer.

 

[joaquins]

Too bad it had to die, too bad for the transformer, good for your learning process... How did you count 6 layers? You have to start unwinding and count then, there is no way to count layers before, at least no easy way with this particular one... Wait, we usually call layers two things, the amount of winding one over the other, and the amount of parts for pri/sec, for the last one it has only one, it's multicore wire. For the first one I would expect  more than 6... Unwrap it and start counting, turns layer per layer.

JS

 

[buildafriend]

It seemed like just 1 clean spool of four evenly wound wires. I wonder if I'm wrong with that or not.. I remember people saying something about 312 xfrmrs being interleaved or something. It was 475 turns total. I didn't count layers because I was expecting the layers to be separated by some kind of tape.

What gauge and kind of material/wire is this? I'm going to try re wrapping. If that fails I can probably just re wrap again with new wires. I wonder if I can take this opportunity to attempt to make a higher gain 312 op xfrmr. 

My lams are toast. Can I find replacement lams? Maybe something experimental would be fun.

 

[buildafriend]

unwound bobbin

 

[joaquins]

I wouldn't try to reuse the wire, but you can try, I did build something alike as I said, using what I could get Si GO lams, I think it was about 0.3mm wire and 3 wires winds as tight as I could as many turns as I could, don't remember exactly but something around 300 turns, maybe a little more, would be nice to know the thick of that wire, my project has 25 percent less turns, roughly and ended in something about 6Ω DCR so guessing something around .2mm

I don't know the 2503 isn't interleaved, it's quadfilar, which tells you isn't interleaved, primary and secondary are wounded together. Maybe the 2623 is interleaved but I don't think so, look at jensen transformer web, they manufacture now days transformers used in api equipment, they do so since they came out, I think there isn't any exact part number but they are there in some way remarked for api, look the closer ones in winding type and core, API uses Si lams, jensen use Ni mostly, I don't know if they sell directly any transformer with Si lams.

JS

 

[buildafriend]

wow I learned so much today.

I need to sleep on this .. it's a lot of information.

Thanks JS.

 

[joaquins]

Good night then, I should be sleeping too but wake up so late so it's going to be hard...

JS

 

[buildafriend]

okay so I decided I want to attempt to re wind this transformer with whatever nearly matching materials I can get. I'm not going for a perfect copy, just close.

There has to be some way to make this quadfiller. I'll pull it from 4 rolls of wire if I have to.

It seems like edcor sells everything I would need to get started (besides a DIY coil winder..thats a whole different mission).

I just don't know what sizes/gauges/etc to order.

How can I ***** the bobbin to order a similar one?

How can I measure and order the same size lams?

What size magnet wire should I get?

What kind of varnish should I use to hold the lams together?

 

[joaquins]

First measure the core size, the mid leg of the E, then look in a table the closest standard size, for the bobbin you only need to measure the inner sizes, if they are the same (or very close) you just ask for a square bobbin for that lam size, if not you just need to add the height of the stack. For the wire, if you want to ask the same size wire you should measure it, wiping off the varnish, and with a micrometer, my guess was close to 0.20mm. I would ask something smaller in order to fit the needed turns inside the bobbin, maybe 0.15mm. Forget the varnish now, once you have it working you put that to finish it and make it reliable.

JS