I emailed Isao about custom PP OPT that is rated higher than their 100 watt unit. He told me it may be possible. I haven't talked to him about the PT, though. I like the Hashi's.
Last edited:
The designer used custom PTs made for his amp. I suppose I could try to use that, but my power supply is beefier than his standard amp. He really doesn't think the PT make to much of a difference as long as it is within spec.
Well, given this site is DIY Audio AND the query is posted in the Tube section are these two factors irrelevant?
Come on... how can doctrinaire responses such as those I summarised in my last post be helpful? After all none of the people posting such comments give any clue how to achieve optimal power supplies nor how they do it.
Well secondaries screened from the other secondary windings (they do exist) will help prevent this happening. An easier and cheaper way (likely) is to give the supplies separate transformers.
The winding resistance of the secondaries of the transformer will also have an impact on the noise generated by rectification. Often toroidal transformers have lower secondary resistance for a given voltage/current rating, the harmonics generated by the rectifier switching will therefore be greater. Some designers choose to use higher secondary R windings for this reason, I seem to recall Thorsten Loesch revealing this and other reasons why his company uses custom transformers.
So if building a tube amplifier from scratch , my suggestion is to use separate transformers for DC heaters, and electrostatic screens or chambered transformer bobbins in the event the builder can't build perfect supplies and needs some mains isolation.
Last edited:
I mentioned grounding as an issue in post 2. That sounds like a clue to me! I believe (I have not checked) that in a more recent post I mentioned the issue of magnetic coupling via large circuit loops - another clue.Robert F said:
These issues are dealt with by various people in various threads, so the details do not have to be repeated here. The clues should suffice.
I find this an interesting subject, so how could one go about testing for differences?
From a practical standpoint, it might be difficult, or at least expensive (several types of transformers).
Once the power supply is "well engineered" different transformer types of the same ratings could be substituted. What would then be tested? What would be listened for?
From a practical standpoint, it might be difficult, or at least expensive (several types of transformers).
Once the power supply is "well engineered" different transformer types of the same ratings could be substituted. What would then be tested? What would be listened for?
IME dynamics, bass quality, soundstage, general quietness. I would say in some cases differences exceed by far those between powercords; in others there is hardly any difference.
With the proviso at the end of your post, I suppose so... However, if the guy had researched those threads he wouldn't have needed to post the query.
Not that I need to have the last word you understand🙂
I have difficulties with the assumption of "well engineered". It is so open to interpretation and so often I think describes a near unobtainable ideal.
In the real world most tube circuits have negligible power supply rejection. Some such as cascodes have effectively none. So what is well engineered for one circuit may not cut the mustard for another.
No regulators I am aware of will filter at the frequency necessary to filter rectification artifacts. Whether those are audible is a different question, though in my experience snubbing the transformer secondary with an RC snubber is audibly worthwhile.
As DF mentions it is important to reduce magnetic coupling, but rectifier switching can generate rf radiation in wire or PCB links between transformer and rectifier, in practical builds these wires are often longish, particularly where efforts are made to distance the transformer from the audio circuit, and the noise will couple to audio circuits.
If the power supply is passively filtered then there are tradeoffs between filtering and power supply impedance.
My own experience with actual comparisons between transformers is not in the tube field. Many years ago I owned a Burmester phono stage. This had dual mono transformers outboard in a separate box. In my infancy as a fiddler I replaced the stock monolithic rectifier diodes with discrete schottky diodes. A year or so later one of the diodes failed and took a transformer out with it. I tried but failed to find replacement transformers of the same va and secondary voltage as the 14 va original. I bought two 50 va transformers of the same type as the encapsulated IE type originals. The sound was significantly better, mainly in the areas of apparent bass drive and power, and soundstage size. The pre amp was regulated and the current draw of the circuit was well within the original transformer ratings.
I also worked part time in hi fi retail back then and recall hearing the comparison between a PS Audio pre amp with standard power supply and with a larger one they had as an option. The larger one was considerably better. I have since read the difference was in the size of the power supply transformer only, though I can't verify if there were other differences, I do understand the transformers were both of the IE type.
different power transformers will have different characteristics in terms of regulation so that when used in an amp the operating points, B+ will not be that same, voltage sag under load will likewise be not the same so that there is a perceived sonic signatures difference in either cases...
Yes, I agree - and that makes actually testing the premise all that much harder.
Try to lay out a test in your mind or on paper.
Try to lay out a test in your mind or on paper.
- What amp topologies?
- What constitutes "good engineering"?
- Where would you get the different transformers?
- What would you measure?
- How quickly could you swap transformers in a listening test?
Yes it is tough to do a fair test alright.
However, there may be a few good rules of thumb to at least minimise technical and potential sonic pitfalls that are of universal use, say:
Ideally use transformers with electrostatic screens or chambered bobbins this will provide some mains isolation;
If space is tight so that mains transformers need to be mounted close to audio circuits, then low radiation transformers i.e. toroids or R cores should be used;
Use power transformers with good regulation in power amps, this may entail using larger than simply adequate va ratings.
Don't buy transformers that vibrate.
Use separate transformers for B+ and DC heater supplies.
Then there are implementation tips like orient the transformer primaries for lowest chassis potential. Orient the transformer to minimise magnetic coupling (particularly with IE core transformers, but be careful with where the leads exit in Toroids too since they tend to "leak" at the exit points) make sure any screens have a low impedance path to ground. Place phono stage mains transformers in a box remote from the audio circuit.
There must be many more.
A man got on a bus. He wasn't familiar with the town so he asked someone to tell him where he should get off the bus for his destination. "See where I get off; you get off two stops earlier".Robert F said:
If the man asked he would find recommendations for reliable transformer brands that don't tend to vibrate, on this very site, unless he asks you of course...🙂
I guess he would get a parable instead.
Unless space is tight i always do this and find the result very audible. Separate regulator per tube as well.
Sorry, I couldn't resist as at face value you had offered a two-pass algorithm. Yes, of course, you can ask for recommendations but where is the fun in that?Robert F said:
Interestingly, when I bought a Sowter transformer (generally regarded as being at the upper end of quality, and certainly the upper end of price) they warned me that it might vibrate so suggested rubber mounting. They were right.
good regulation requires low dc resistances in windings, since i build all my traffos, i found that current densities of 500cm/per ampere are adequate, using 700cm required larger gage wires....implication is that the EI cores will be bigger since lower turns count is mandatory if you want to use bigger sized wires in the windings.....then i operate at less than 1T flux densities...
of course, but if you already have one, there are remedial measures that can be taken...
this I like, since i can make transformers any which way i like...
you can bring the horse to the water well, but if the horse do not want to drink, there is little you can do....
you must be reffering to a torroid, torroids are prone to buzz since they are operated at higher flux densities over 1T and have no air gap unlike an EI....
EI's are easy to fix in terms of vibration, there are things you can do like dipping in varnish, shimming the bobbin with the center leg, and tightening bolts...
EI's are easy to fix in terms of vibration, there are things you can do like dipping in varnish, shimming the bobbin with the center leg, and tightening bolts...
- Status
- Not open for further replies.