Page 9 of 14

#121 Re: GM70 PP

Posted: Thu Nov 14, 2019 10:20 am
by izzy wizzy
RhythMick wrote: Thu Nov 14, 2019 9:54 am Why not think about a slow start option?
I don't think I need one. Nothing goes over voltage, the MVs can handle the current, the wire in the choke will be fine.

I don't mean that to sound dismissive of the idea. Slow start and sequential start gizmos are just not my thing and no one else switches these things on. I have had sequential start things in the past and they just drive me nuts in the end.

#122 Re: GM70 PP

Posted: Thu Nov 14, 2019 11:34 am
by Nick
I don't think I need one. Nothing goes over voltage, the MVs can handle the current, the wire in the choke will be fine.
Would get my vote, I use slow start to protect the input fuse, better to have components that are happy with worst case situations.

BTW, one of the things that makes winding a choke input choke different from a cap input one is the voltage it has to withstand from one terminal to the other, so the way the windings are laid down and insulation between layers becomes important.

#123 Re: GM70 PP

Posted: Thu Nov 14, 2019 11:42 am
by RhythMick
Nick wrote: Thu Nov 14, 2019 11:34 am BTW, one of the things that makes winding a choke input choke different from a cap input one is the voltage it has to withstand from one terminal to the other, so the way the windings are laid down and insulation between layers becomes important.
The Lundahl datasheet quotes "insulation between windings and core" as 4kv ... The max signal voltage AC RMS is also quoted, 450V for the 2743.

#124 Re: GM70 PP

Posted: Thu Nov 14, 2019 11:51 am
by izzy wizzy
Nick wrote: Thu Nov 14, 2019 11:34 am I use slow start to protect the input fuse, better to have components that are happy with worst case situations.
From memory, it's a while ago now, I used a varistor/MOV doodad to give things an easier start and then relay bypass it with the start sequencer.

What slow start gizmo are you using Nick?

#125 Re: GM70 PP

Posted: Thu Nov 14, 2019 12:10 pm
by Nick
I use a couple of relays, and some paralleled resistors to start up with the resistors in series, then bipass them. But I only do that for a few tenths of seconds while the torrid inrush goes away. This is with solid state BTW.

#126 Re: GM70 PP

Posted: Thu Nov 14, 2019 12:11 pm
by izzy wizzy
pre65 wrote: Thu Nov 14, 2019 10:04 am One reason why I like 6D22s damper diodes. :)
A very good option and if I wasn't addicted to mercury, would be the way to go. I even have them in the shed.

#127 Re: GM70 PP

Posted: Thu Nov 14, 2019 12:12 pm
by RhythMick
I went with thermistors in the mains primary, bypassed by relay after a few tenths of a second.

#128 Re: GM70 PP

Posted: Thu Nov 14, 2019 1:10 pm
by izzy wizzy
RhythMick wrote: Thu Nov 14, 2019 12:12 pm I went with thermistors in the mains primary, bypassed by relay after a few tenths of a second.
That's what I meant. One of them, not varistor/MOV as I said earlier.

#129 Re: GM70 PP

Posted: Thu Nov 14, 2019 3:51 pm
by Mike H
izzy wizzy wrote: Wed Nov 13, 2019 7:03 pm So dragged MJ of the shelf and he is talking about peak currents so your example Nick would seem to follow what he says and the PSUD sim is useful in that regard. For the current draw of 60mA and that choke, it should be rated higher than 78mA.
As you're using PSUD, have a look at the rectifier currents. You should find each one is conducting almost the entire half cycle. This is different to cap input, when they conduct only on the peaks. This is because after the input peak the choke field collapses and induces a current in its winding, in the same direction. The input end of the choke then becomes the negative going end of what is now a current source, and that keeps the rectifier(s) conducting to quite a low Voltage.

#130 Re: GM70 PP

Posted: Thu Nov 14, 2019 5:26 pm
by Nick
Mike H wrote: Thu Nov 14, 2019 3:51 pm
izzy wizzy wrote: Wed Nov 13, 2019 7:03 pm So dragged MJ of the shelf and he is talking about peak currents so your example Nick would seem to follow what he says and the PSUD sim is useful in that regard. For the current draw of 60mA and that choke, it should be rated higher than 78mA.
As you're using PSUD, have a look at the rectifier currents. You should find each one is conducting almost the entire half cycle. This is different to cap input, when they conduct only on the peaks. This is because after the input peak the choke field collapses and induces a current in its winding, in the same direction. The input end of the choke then becomes the negative going end of what is now a current source, and that keeps the rectifier(s) conducting to quite a low Voltage.
And that is why you get such a large voltage across the choke. And why you don't get switching spikes.

#131 Re: GM70 PP

Posted: Sat Nov 16, 2019 12:06 am
by Mike H
Image and wot he said. :D

Took me ages to get my head round how choke input works. Must be threads around here from about 10 years ago, when Nick was trying to explain it.

#132 Re: GM70 PP

Posted: Sun Dec 01, 2019 7:41 pm
by izzy wizzy
After living with the Lehane regs on the GM70 fils, I got to thinking I was not liking things as much, even turning off a few times as the sound was quite sterile and not for me so got it out to have a look. Noted the common connection was the negative end while the 4P1L was the positive fil connection. I had reasoned that the positive connection should be the common one yet connected the GM70 differently. Altered the GM70 to the same config as the 4P1L and bingo, back in business and the gap between this connection and the SMPS is far greater now.

One thing the Lehane filament regs do is open up the top end end I thought things up top had got a bit much; a bit tilted up. So replaced the Lundahl 1676 on the input wired as 1:4 with a Sowter 3575 which is 1:1. I don't need the gain of the 1:4 any more. Oooh dear, that kinda sucked the life out of things tonally and dynamically. Googled around a bit and found some info from Lynn Olson who had measured the 1676 finding it to have an incredible measured frequency response and very impressive phase splitting performance at 1:1 when loaded correctly, slightly worse at 1:2. So bunged it back in at 1:1 and most of my reservations about the top have been much reduced so will see over time. I may have to experiment with loading to see what happens.

Planning for a front end supply is ongoing. I need to get an order off to Majestic for some more iron to do this properly. Anyone I've read about having a separate driver supply rates the result as significant so got me thinking, I should do this properly.

Cheers,
Stephen

#133 Re: GM70 PP

Posted: Sat Feb 01, 2020 10:58 am
by izzy wizzy
Naively I thought making a new supply for the front end would be easy. And yes, getting it going was and I hear the benefit. It is a significant improvement.

However I'm still having to fiddle with things as I think a parasitic has been uncovered. I hear this as a pressure on my eyes and sinuses and it wasn't there before.

I did start with more top end than I thought right and removing the small snubbers on the MVs got rid and made it sound more like the RC decoupled circuit but then I noticed the paracitic.

So a bit more to do but heading the right way.

IMG_20200201_104550212_compress15.jpg

#134 Re: GM70 PP

Posted: Sun Feb 02, 2020 11:04 am
by izzy wizzy
It's very helpful to post thoughts along the way to focus the mind on where this is going and the last post did that. It's already changed from 46 813 as there was no future in those valves as to hard too get and costly. And now I find a similar position using mercury for the front end. I don't have the spares for the future and the little ones are unobtanium so time to move on and that will affect the phono as well.

Choices are swap out the MVs for other 4V DH rectifiers like RV120/DW4/U14. Never used these, don't have any but they seem easy to get and not too silly in price. But do they sound any good, work well, are there better devices? Advantage is they would just slot in without any changes.

Then there's the 5V DH rectifiers. The one I hear most about is the 5R4GY. I'd need a new filament tx, not a huge deal, and by all accounts, a lovely rectifier. Not too hard to get and not totally stupid money yet.

Do you get an overly valve sound (in the wrong way) or is it a win in a typical DH valve kinda thing? Bearing in mind Steve's epxeriemce in his OTL for the driver; it being most sensitive or was that HT differences. The driver can usually be a small power amp in its own right hence its effect on the chahracter of the amp.

And finally damper diodes like the 6AX4GT. I use these in the line/headphone amp. Seem OK but haven't compared, relatively cheap and plentiful and will also need another heater transformer but again, no biggy.

6AX4 seems an easy choice but as it's a DH amp so like the idea of DH rectifier. Choices, choices, choices.

Any war stories, experience of these things welcome.

#135 Re: GM70 PP

Posted: Sun Feb 02, 2020 11:37 am
by simon
If it helps any the GM70 amp I brought to Owston last year used 6AX4GTs for the GM70 PS. The 12GN7 drivers used SS rectifiers but I can't for the life of me remember if they were simple 1Ns or something faster.