Deltalab Effectron voltage rails question

[TVC15]

Hello,
I have this effectron iii which had serious battery acid damage (along with some other problems) and I've got it almost back up and running BUT
the corrosion ate one of the secondary leads all the way up to tap basically and the wire is like the diameter of a human hair so I was not able to get a lead soldered to it. The unit has a toroidal transformer with 3 secondary windings one at 18V one at 10V and the one that got eaten which I assume is 24V or less since that one is supposed to supply the +12 analog and +12 digital rails. The 18v winding supplies the -12 rail, and the 10v winding supplies the +5 digital and +5 analog rails.
What I'm wondering is can I just use the 18v winding to supply the +12 analog and +12 digital rail as well as the -12, by jumping the 18v input from the -12 bridge rectifier over to the +12 A+D bridge rectifier input? Or is there some good reason not to do that?
I really just don't want to have to buy a new transformer and this unit is only supposed to pull 100 mA of current max according to the fuse.
The attached schematic is not the exact model but the power supply is identical, I've drawn in the two jumpers that I proposed adding.
Thanks for looking!

 

[JohnRoberts]

If you could, they might have done that in the initial design.

It looks like the 5V rail is the high current draw.

The +12V supply is two separate regulated rails one for analog and the other perhaps dedicated to just digital(?) +12D

The -12v rail is a little unusual, the 7812 (a positive regulator) has its + output connected to ground to generate the -12V supply. Connecting the two diode bridges together will prevent the floating minus unregulated from developing a suitable negative voltage.

Perhaps capacitor coupling the red 1,2 winding to 5,6 bridge with two large electrolytic capacitors (respecting proper voltage polarity) might (?) work but no guarantee. Another possible gotcha (?) is there enough current to spare from that 1,2 winding to drive another supply rail?

1/10 A at mains voltage will be over one amp when stepped down to 12V

JR

 

[radardoug]

Not a good plan. Maybe replace the one transformer with three, because you probably wont find that one. Or do a major supply redesign.

 

[TVC15]

RadarDoug- I like the idea of using multiple transformers if I could find them in a small enough form factor. It's only 1RU , even the original toroidal is the slim kind, which makes finding a direct replacement even more unlikely. Maybe a small transformer and a little perfboard circuit with some LM723's would fit.

John- thanks for responding, I'm a fan of your loft delay units.
the Red 1,2 winding for the +12 analog and +12 digital is the one with the corroded wiring - part of the reason it was eaten by the acid is because it is such a tiny gauge, so I think it must be low current consumption.
The black 5,6 winding for the -12 rail is decent gauge so perhaps that winding would provide enough current to feed the 1,2 bridge through a pair of electrolytics as you suggested. I'm going to read up on the P = V × I relationship.
Revised image:

 

[radardoug]

This is where a bench supply is your friend. Drive the positive rail with the bench supply, and find out what current is drawn by that supply. Then maybe get a tiny transformer to do just that supply, and find a space for it. Leave the existing transformer to do the other supplies.

 

[TVC15]

Yes, Great idea, very helpful, I'll definitely do that - I have a dual output bench power supply, which was very useful in this project already, because there was initially a short between the -12 and ground, so I used it to inject a low voltage into the rail and then used a thermal camera to find the culprit, which was a 4558.

 

[TVC15]

Ok - the +12 Analog and Digital rails pull only 110mA. I found a 18vac transformer made by zettler that is pretty tiny, bv301s18023 that is rated for 128mA secondary current. 15 bucks for 4 on ebay, not too bad.
It's 27mm on it's side, and the height available in the enclosure is 28.5mm.
I really appreciate the help in finding a solution!

 

[radardoug]

Good luck!

 

[NOON]

Another option might be the meanwell triple output switch mode modules. They tend to have 5v at high current and +/- 12v or 15v rails. I've found them to be very quiet and usually about the same price as a toroidal through the major distributors.

 

[TVC15]

Very interesting, I saw something like that on the bradthx youtube channel when he couldn't find the fault with the factory switch mode power supply on a Roland SP-700 and just replaced it lock stock and barrel with one of those types of modules. Looks like a Meanwell RPT-60B could fit if I removed the original transformer.

 

[CScotty]

Ok - the +12 Analog and Digital rails pull only 110mA. I found a 18vac transformer made by zettler that is pretty tiny, bv301s18023 that is rated for 128mA secondary current. 15 bucks for 4 on ebay, not too bad.
It's 27mm on it's side, and the height available in the enclosure is 28.5mm.
I really appreciate the help in finding a solution!View attachment 119228

Running the numbers with that transformer you risk burning it out. I used power supply designer II PSUD2

110mA average on the DC side translates to 147mA RMS on the transformer secondary. I took the effective impedance of the transformer at the secondary as 104 ohms and the open circuit voltage as 31.4V RMS see here; https://www.zettlermagnetics.com/pdfs/EI30_23.pdf

Since you have four coming do you have room to wire two in parallel?

 

[TVC15]

Darn, errors galore! I did the RMS calculation for the voltage but not for the current.
To answer your question, CScotty, yes - I think I could fit two of these little guys paralleled.
However your question had me questioning my current draw measurement. I had injected the voltage at the input of the +12 analog voltage regulator because the inputs to both the +12 analog and +12 digital regulators are in parallel - that way I could feed both rails with one power source - but I failed to account for the voltage drop of the regulators. With my bench power supply set for 12v I was actually only getting about 10 and a half volts to each rail. So I just repeated the process but with the bench psu set for 13.5v so that I was actually getting 12v after the regulators and the current draw became 130mA. So I really need like 180mA RMS on the secondary.
If I understand the rules for paralleling transformers correctly, two of the aforementioned zettlers would provide 256mA RMS and I'd be good to go, right?

 

[CScotty]

Yes, double the current for two transformers.

With two of those transformers you will find the average voltage on the capacitor to be around 26V which is rather high. Therefore you will be dissipating about 3.25 watts in the two regulators and I don't know how this is shared between them or if they have heatsinks. Just check they don't run too hot.

 

[TVC15]

No heatsinks - just a large ground plane (that I spent a long time scrubbing copper (II) hydroxide off of). Although I have some small heatsinks from Tayda that I could add to those regulators if needed. I need to add some kapton tape under the regulators for the -12 and +5A, and put in a replacement battery. The white wires are an attempt at using the slim remains of the Red 1,2 winding that didn't pan out. I was planning on moving one of the large caps to make room for the transformers. But I'm keeping the Meanwell triple output psu that Noon suggested in my back pocket if gets too crowded in there.

 

[TVC15]

I should also put a power switch in this thing since it has none

 

[pvision]

I don't think you'll get a Meanwell to fit in the case. As you know the transformer is mounted to the PCB so the depth available is limited. A chassis-type, uncased, SMPS might fit

I don't think you'll be able to find room for additional transformers, either

I'd build an external psu, or look for one of the many non-working Effectrons out there. I've not had a bad transformer on one so far, but the overall failure rate with them is about 80%

Overall I'd be careful as I suspect they are susceptible to damage from static

 

[TVC15]

I concede that it will most likely require modification to fit, but I'm going to give it a try - as the zettler 18v transformers I had mentioned were not actually available from that ebay seller.
The Meanwell rpt-60b hasn't arrived yet, but the datasheet says it's 29mm from the top of the tallest cap to the bottom of the pcb, but then states that the solder side of the board needs 3mm maximum additional clearance. There is also already a large hole in the pcb under the original transformer that could be expanded to accomodate the 2x4" meanwell psu footprint.
I'd Dremel the rectangle out, then zip tie from each corner of the rpt to a corresponding hole drilled in the mainboard.
I'm going to be on the road this week, but I'll update this post when I get back and have all the materials in hand.

 

[JohnRoberts]

I saw a deltalab effectron listed as a parts unit for around $30 on EBAY... maybe it has a usable transformer?

JR

 

[TVC15]

Thanks, I see it... the auction ends in 3 days, so it will probably go for at least twice that though. But I'll add it to the watchlist.

 

[TVC15]

I got the new meanwell power supply installed. It fits well, and didn't require any drastic alterations with a dremel.
No joy yet, however. I have all my rails again, but I had to replace the 1st input opamp right away which revealed itself to be dead, which got me to the second opamp which was the one I had already replaced awhile ago, but it was dead again - don't know how that happened. So replacing that opamp (again) got me to the comparator U5, which now has input but no output. It's a LM311, I don't have any of those on hand so I'll have to get a few, and do some more troubleshooting.
All these bad components in the input stage utilize the +12 and -12 as a dual supply, so maybe there was a cascading failure related to whatever had caused the short in the second opamp between -12 and ground that was present when I first got the device.
Here's a schematic for a very similar or maybe the same(?) unit. And a photo of the new power supply in the device.