Ok I took the measurements. After about30-60 of warm up the primary winding of the right channel draws 740mA, setting the bias to absolute zero I get still 410mA of current. The primary of the left channel draws about 600mA with 275mA of bias, and 280mA with bias to zero. This is consistent with the fact the the right transformer hums a bit louder than the left, and that with bias to zero I can still hear a faint hum from the right transformer, while the left one is silent.
Any chance that the problem is not due to faulty transformers? Would this cause problems if using the amp? The amp sounds just fine and since I have a small listening room I think I might use no more than 20-30W of power when playing music, however I never have listened to a completely flawless ML23.5, might sound even better I don't know.
Yes the test was relatively straightforward. If the transformers are faulty, it may not be so terrible. There is a well known transformer company here in Rome that could rebuild no the transformers for a reasonable price. Obviously they could perform precise tests before to confirm the transformers have issues. But if they are not too difficult, risky or require additional equipment I could do the tests as well.
Stefan,
How exactly did you measure, because it's a bit complicate since both transformers are connected in parallel according to the circuit diagram.
It's not possible to separate the four paralleled 20V windings, but only the 100Volt windings can be removed, P301 / P323 for one side and P302 / P 324 for the other side. Is that what you did ?
But even then, both amps are fully powered so removing the bias from one side is exactly the same as removing it from the other side and that's what your measurement confirmed.
How did you set the bias to zero, with the pot or by disconnecting the rectifier. When disconnecting the rectifier you remove 275mA plus 50mA for the drivers so in total 325mA.
When driving the pot to zero bias, you go from 325mA to 25mA, current still flowing into the drivers.
But one thing can be concluded, removing 300mA (275mA bias + 25mA drivers) from 2x88Volt results in ca. 325mA less from 220Volt.
That would mean an efficiency of only 74%, so I guess you disconnected the rectifier which results in 80% efficiency.
Still not great because I would expect at least 90%.
There is still a power-on relay consuming only some 2.5 Watt power, but this is taken from the 20 Volt windings, so in either case it is powered.
In total it gives me the impression that one of the two transformers drawing 740mA is having issues and should be replaced.
The answer to that question can only be confirmed by completely removing that transformer and measure again.
Hans
How exactly did you measure, because it's a bit complicate since both transformers are connected in parallel according to the circuit diagram.
It's not possible to separate the four paralleled 20V windings, but only the 100Volt windings can be removed, P301 / P323 for one side and P302 / P 324 for the other side. Is that what you did ?
But even then, both amps are fully powered so removing the bias from one side is exactly the same as removing it from the other side and that's what your measurement confirmed.
How did you set the bias to zero, with the pot or by disconnecting the rectifier. When disconnecting the rectifier you remove 275mA plus 50mA for the drivers so in total 325mA.
When driving the pot to zero bias, you go from 325mA to 25mA, current still flowing into the drivers.
But one thing can be concluded, removing 300mA (275mA bias + 25mA drivers) from 2x88Volt results in ca. 325mA less from 220Volt.
That would mean an efficiency of only 74%, so I guess you disconnected the rectifier which results in 80% efficiency.
Still not great because I would expect at least 90%.
There is still a power-on relay consuming only some 2.5 Watt power, but this is taken from the 20 Volt windings, so in either case it is powered.
In total it gives me the impression that one of the two transformers drawing 740mA is having issues and should be replaced.
The answer to that question can only be confirmed by completely removing that transformer and measure again.
Hans
Hello Hans, thanks for the reply. To take the measurements on the right transformer I disconnected only P301 and connected the meter in series. To test the left transformer I disconnected P302 and put the meter in series. Is this correct? To set the bias to zero I just turned the pot, I didn’t disconnect the rectifier.
So you suggest taking these same measurements but disconnecting all the primaries from the other transformer (except the 20V winding for the relay)?
As I said there’s a company I think is trusty here in Rome who could repair the transformers, but I guess it would be better to do both transformers anyway for simmetry between the channels? Because I think they will need to completely disassemble them and redo all the windings.
That would indeed be the best option to take them out both. As you say, they have to be exactly the same.
I don’t think further experimenting at this stage brings extra info. It’s already obvious that at least one trafo is not O.K.
Hans
Ok, I will remove them and take them to repair. Since I cannot find the specs, aside from the primary voltages and rectified voltages, and the fact that they should be 1258VA each, do you have other information to rebuild them as close to the original as possible? Or they should be able to reverse engineer them at the laboratory?
The tranformer shop should be able to reconstruct the amount of winding, most likely by counting them.
Hans
To be honest, they look different from what I have seen so far.
But ML may have used products from different suppliers.
Hans
Yes I have only seen two different types, the black ones non-encased that look identical to the ones of the ML23, and in later models I see the plitron silver-coloured. Mine are in this plastic case and are potted. I don’t think they have been replaced, but who knows maybe it was a different (inferior?) supplier they used for some time during production.
I also noticed something weird when disassembling them, the bracket holding the posterior transformer is held to the chassis with three screws, the middle one is longer towards the inside and a rubber foot is placed loose upside down on the screw thread, the transformer then resting on the foot. It looks original from production, but definitely a bit messy. I also see on pictures of the bottom of the chassis from the internet that this screw look different from the others, so it may be how they used to assemble the amps.
Having both transformers separated from the amp and from each other, you could connect the four 100Volt and the 20Volt windings in series (carefully in the right direction), connect it in this 240Volt setup to mains and measure how much current they draw from mains in this unloaded situation.
Then measure the voltage over the four primary windings and over the secondary windings.
This will most likely tell you what’s wrong and which transformer is not o.k.
Hans
Thanks Hans, but I already brought them to the shop. They will test them and get back to me tomorrow.
That is an extremely weak test. If any winding is shorted, failure of the series +-83v windings to deliver 166 vac into an 8 ohm resistor would occur. It doesn't matter which winding has a shorted turn. Removal of the transformer was not necessary. Taking of the load current off the AC input of the bridge rectifier would have been quicker and provide a more difficult trial for the transformer.
That ~300 watt resistor would be necessary to test the speaker output for full wattage after any repair, too. Quick and easy test for condition of rail capacitors in tired amps - lack of rated wattage on speaker terminal.
I heard back from the transformer shop. They tested thoroughly both transformers and they say they measure equally good. They exclude internal shorts and/or other malfunctioning, and discouraged me from undergoing the rebuild. The problem must be somewhere else then.
They suggested me to check the transformer wiring since a few had the insulation damaged by the soldering iron during prior repair. I did notice that but concluded there was no risk of shorts, I will check more carefully and insulate with heat shrink if necessary. Other things to check? Should I check the current draw of all 8 rails?
They suggested me to check the transformer wiring since a few had the insulation damaged by the soldering iron during prior repair. I did notice that but concluded there was no risk of shorts, I will check more carefully and insulate with heat shrink if necessary. Other things to check? Should I check the current draw of all 8 rails?
Well congrats, at least that´s very good news.
For a start, your suggestion to measure the current flowing into the 2 x 4 primary sections is a good idea with Bias in both channels set a 275mA.
Also check whether the four 20Volt windings are correctly connected, if not they will shorten each other, but you will also notice that when measuring their currents.
But before doing that, you could also do as a first step the test described in #273 and #274.
Depending on the outcome, I´m looking forward to see the results.
Hans
Hello Hans,
To be honest no I didn’t have the time yet. I was only able to reassemble the transformers yesterday (I replaced the split lock washers that were all flattened out and reinstalling the nut was almost impossible, there wasn’t enough bolt for that… so I had to use a slightly longer bolt, everything’s fine and nothing shorted out ). The transformers still buzz, not sure if a bit less than before. I rechecked all the wiring and all windings are connected correctly with correct polarity, according to the ML23 schematics I have. So I think I will go ahead and measure the current draw of all rails and see if there’s anything unusual.
To be honest no I didn’t have the time yet. I was only able to reassemble the transformers yesterday (I replaced the split lock washers that were all flattened out and reinstalling the nut was almost impossible, there wasn’t enough bolt for that… so I had to use a slightly longer bolt, everything’s fine and nothing shorted out ). The transformers still buzz, not sure if a bit less than before. I rechecked all the wiring and all windings are connected correctly with correct polarity, according to the ML23 schematics I have. So I think I will go ahead and measure the current draw of all rails and see if there’s anything unusual.