GM70 PP

What people are working on at the moment
User avatar
izzy wizzy
Old Hand
Posts: 1496
Joined: Fri Nov 02, 2007 7:02 pm
Location: Auckland NZ
Contact:

#166 Re: GM70 PP

Post by izzy wizzy »

Finally replaced the 866 MVs as there's no going forward with them as too hard to get and expensive now. Popped in 6CL3 damper diodes and like the result.

When getting the amp ready for Owston, I lowered it a lot which brought the MVs a lot closer to the audio and it sounded like I got some RF interference. I moved the MVs to the back and that reduced the effect.

Hindsight being a wonderful thing; putting the 6CL3 in seems to have banished that issue and maybe I've been in denial about it but in the back of my mind, having something that emits RF in the centre of the amp was not a great idea.

That now means I can repackage it to its smaller size. Would be great to have all the valves on top but have not been able to get all I need into such a small footprint so the rectifiers will need to stay inside hidden.

The amp is now easier to use as the dampers have a nice slow turn on. No more blue flash at start up which makes it less mad science project.
User avatar
IslandPink
Amstrad Tower of Power
Posts: 10041
Joined: Tue May 29, 2007 7:01 pm
Location: Denbigh, N.Wales

#167 Re: GM70 PP

Post by IslandPink »

You'll miss the visuals, but it does sound like a better engineering solution. I've been sold on damper diodes for years since I first tried them. Reminds me I ought to get a few more 6DL3's for spares for my power supply.
"Once you find out ... the Circumstances ; then you can go out"
User avatar
izzy wizzy
Old Hand
Posts: 1496
Joined: Fri Nov 02, 2007 7:02 pm
Location: Auckland NZ
Contact:

#168 Re: GM70 PP

Post by izzy wizzy »

True, they're not as much fun as the MVs to look at. The only issue I came across with these, like the 6DL3, is the Novar base which is now hard to find easily. I scrunched the sockets on cheap chinese Magnoval base which seems to be OK; not quite as tight as I'd like but they're not on the top plate. The RS 6.3V transformer is a cheap way of heating them.

I've used 6CJ3/6DW4 not 6CL3 as mentioned earlier.
User avatar
izzy wizzy
Old Hand
Posts: 1496
Joined: Fri Nov 02, 2007 7:02 pm
Location: Auckland NZ
Contact:

#169 Re: GM70 PP

Post by izzy wizzy »

The amplifier nearly got to being put in a proper case. It's been very enjoyable listening to it over the last 18 months. However recently I started to measure it as I got a signal generator. Also inspired by Mr Idler's thread and measurement of the input transformer http://audio-talk.co.uk/phpBB3/viewtopi ... 73#p183073, I measured mine and it didn't look good.

Initially I had it as 1:2+2 as the datasheet but changed it to 1+1:2+2 preferring the sound. However when measured it was all over the place. I tried it as this thread suggested https://db.audioasylum.com/mhtml/m.html ... t%3DLL1676 which makes it a 2:1+1. Measured performance was as that thread suggests; very good. But gain is of course lower and now too low for my system. Previously, it required 7V in for full power out and now the input sensitivity is half that but I don't need anything like full power. The sound was significantly better so back I went to the 1:2+2. Measurements seemed to show it's performance to be just as good but the sound was awful so back I went to the last hookup.

And listened for days to just how glorious it was albiet at a lower volume than I would like. Something had to change but what?
Last edited by izzy wizzy on Mon Jan 04, 2021 9:30 am, edited 1 time in total.
User avatar
izzy wizzy
Old Hand
Posts: 1496
Joined: Fri Nov 02, 2007 7:02 pm
Location: Auckland NZ
Contact:

#170 Re: GM70 PP

Post by izzy wizzy »

As gain was marginal and now too low, plans had been to take the DHT thing further back in the chain to the line stage and so had proposed to use 2P29L in there but that would only double the gain and now, not get me where I needed to be. Max's thread http://www.audio-talk.co.uk/phpBB3/view ... =11&t=8049 has also helped in discussing ideas relating to transformer coupling.

As this journey was inspired by Lynn Olson's Amity/Karna, I've been following what he would do now. This thread in DIY Audiohttps://www.diyaudio.com/forums/tubes-v ... -pull.html has been addressing the input stage of a 3 stage amp. Back in the beginning of the 813PP amp, this was going to be a 3 stage amp with the input being a 12B4 but I didn't need the gain and they ended up in the line stage. This chap also has 12B4 in a very similar line stage.

It looks now the 2P29L could be the input stage. Reading the DIY Audio thread and pulling together thoughts from others leaves me with idea I could cap couple the 2P29L to the 4P1L. Just leaves the question of how to load them; choke, CCS or resistor. I could even start by transformer coupling as I have a spare Lundahl 1635 in the shed to get going.

Just have to find some room in the existing amps for the extra stage.
User avatar
izzy wizzy
Old Hand
Posts: 1496
Joined: Fri Nov 02, 2007 7:02 pm
Location: Auckland NZ
Contact:

#171 Re: GM70 PP

Post by izzy wizzy »

These few posts are more of a brain dump; it helps thinking out my next move. Sorry if I may be repeating some of the thoughts.

What I do for the first stage isn't really any of my own work. Zig Zag Flux and Lynn Olson on DIYAudio forum have been there before me and I'm following on in my own way from what they've done/struggled with. They and me are very happy with a decent driver doing it's thing to the finals via a transformer and have documented all the good reasons why. Main difference is they have used 45 or 46 valves which requires deep pockets if you can get them at all or foresight to have stocked up. I have neither so opted for the 4P1L which compared to the 46 I tried is a choice I'm happy with. ZZF had similar driver transformer to mine but in the end went for Monolith bifilar but that's out of reach for now.

Both have lived with an interstage transformer coupled to the driver type amplifier and feel transformer coupling the input to the driver to be problematic for a bunch of reasons like distortion and phase issues that can be much lower with other coupling/loading methods. Plus the desire to use a lower distortion/more linear valve like a 6SN7 or variant which are no good for transformer coupling. My version will be a variant of Lynn's Karna Kay proposal - 6SN7 cap 45 IT 300B

The driver is an easy load on the input valve so suitable for capacitor coupling. I could also use SN7 or variant but I want to continue with direct heated valves and have a cheap supply of 2P29L. Choice now would be to either load it with a choke or CCS. The HT is far to low for resistive loading. I have a transformer and could use just the primary as the choke load.

But before anything happens, I need to get parts for the Lehane VCCS boards to light the 2P29Ls.
RhythMick
Old Hand
Posts: 747
Joined: Fri Jun 23, 2017 12:24 pm

#172 Re: GM70 PP

Post by RhythMick »

As a result of my switch from PP to SE I have a number of spare Lehane boards and transformers. I'm quite happy to keep them, you never know what future projects may require, but equally if those boards are in short supply there's no point my hoarding.

Do feel free to PM me if needed.
User avatar
izzy wizzy
Old Hand
Posts: 1496
Joined: Fri Nov 02, 2007 7:02 pm
Location: Auckland NZ
Contact:

#173 Re: GM70 PP

Post by izzy wizzy »

Having done a lot more reading, I thought it might be an idea to try 6SN7 up front with CCS loads as I had some but after a shed rummage, could only find 1 6SN7 which is odd as I know I have at least 2. Got 4 10M45 for rudimentary CSS loads. An IDHT would get me going sooner and it might be fun to hear a 6SN7; see what all the fuss is about.

If I can't find another, thought I could try ECC99 or 6N6. Looking at a load line calculator http://www.trioda.com/tools/triode.html proved to be very confusing especially when relating it to known designs like the Reichert Flesh and Blood fer instance. For the first stage op point, H2 I thought quite high but H3 super low. Maybe that's the point?

I only have about 220V HT so tried to find low distortion op point say around 160V leaving about 30V swing up and 20V or so still across the CSS. Weirdly higher anode voltages produced more distortion in the calculator esp H2 while H3 remained very low. ECC99 looked very bad. 6N6 reasonably low for 100V Va and 15mA. Nothing as low as the 6SN7 though which seemed to work well at about 155V @ 10mA. I don't know what to think about this calculator's accuracy though it is fun to play about with it.
User avatar
izzy wizzy
Old Hand
Posts: 1496
Joined: Fri Nov 02, 2007 7:02 pm
Location: Auckland NZ
Contact:

#174 Re: GM70 PP

Post by izzy wizzy »

RhythMick wrote: Thu Jan 07, 2021 12:39 pm As a result of my switch from PP to SE I have a number of spare Lehane boards and transformers. I'm quite happy to keep them, you never know what future projects may require, but equally if those boards are in short supply there's no point my hoarding.

Do feel free to PM me if needed.
Thanks for the offer however I'm sorted for boards and transformers. You never know when you might have an idea and need them :)
User avatar
IslandPink
Amstrad Tower of Power
Posts: 10041
Joined: Tue May 29, 2007 7:01 pm
Location: Denbigh, N.Wales

#175 Re: GM70 PP

Post by IslandPink »

izzy wizzy wrote: Thu Jan 07, 2021 1:21 pm 1. confusing especially when relating it to known designs like the Reichert Flesh and Blood fer instance. For the first stage op point, H2 I thought quite high but H3 super low. Maybe that's the point?

2. I only have about 220V HT so tried to find low distortion op point say around 160V leaving about 30V swing up and 20V or so still across the CSS. Weirdly higher anode voltages produced more distortion in the calculator esp H2 while H3 remained very low. ECC99 looked very bad. 6N6 reasonably low for 100V Va and 15mA. Nothing as low as the 6SN7 though which seemed to work well at about 155V @ 10mA. I don't know what to think about this calculator's accuracy though it is fun to play about with it.
Interesting stuff, I haven't tried it myself
1. I would say that this sounds likely - that was my understanding about how the Flesh and Blood is set up - it's not known as a low-distortion amp, but the spectrum is supposed to be ear-like.

2. Interesting comment on ECC99. I didn't like them in the Aurora. What products was it showing ?
In my Aurora, I went ECC99>5687>6N6P . I felt that the 6N6P was best, Nick suggested that because of their poorer section-to-section matching than the 5687, it may be that I was hearing some extra 2nd-harmonic coming in, with those.

Reminds me I really need to get LTSpice working again, I never got the valve models going again after changing my computer.
"Once you find out ... the Circumstances ; then you can go out"
User avatar
izzy wizzy
Old Hand
Posts: 1496
Joined: Fri Nov 02, 2007 7:02 pm
Location: Auckland NZ
Contact:

#176 Re: GM70 PP

Post by izzy wizzy »

I was just fiddling with the load line calculator ref'd above not spice. The ECC99 had a very stubborn high H3 until very low currents like a few mA. H2 was quite high until very low Va. The few times I've tried them, we haven't really got on at op points I thought looked OK. The load line tool seems to direct towards lowish Va for the types I tried. 813 was low distortion compared to GM70 yet wasn't my experience when listening.

Maybe that load line tool is like that Sakuma quote "The tube manual is like a telephone book. It gives perfect numbers. It is useful to make it possible to speak with a girl but we cannot see her beautiful face from the mere telephone number. "
User avatar
izzy wizzy
Old Hand
Posts: 1496
Joined: Fri Nov 02, 2007 7:02 pm
Location: Auckland NZ
Contact:

#177 Re: GM70 PP

Post by izzy wizzy »

IslandPink wrote: Fri Jan 08, 2021 1:22 pm Reminds me I really need to get LTSpice working again, I never got the valve models going again after changing my computer.
I found on mine I was updating the model files yet they weren't appearing in the sims. Took me ages before I figured out I had two installs; one in the right place (program files) and one I'd stuck on the NAS with the sim files :roll: The one in program files was installed after a hard disk crash but the one with the files was the one that started when I double clicked the sim file.
User avatar
izzy wizzy
Old Hand
Posts: 1496
Joined: Fri Nov 02, 2007 7:02 pm
Location: Auckland NZ
Contact:

#178 Re: GM70 PP

Post by izzy wizzy »

I was hoping to get the output transformers looking smart by painting them. Is this something that's easy enough to DIY? Or is it better to get someone else to do it professionally at reasonable cost?

Is it better to disassemble to get the end bells off. Is there a risk to the transformer in removing all the bolts to do this?

The idea is to try and make the amps look good which is something I've never done before. I could of course just stick a box over the top but the GM70s give off a lot of heat and although I cen get them a decent distance, they might be still too close.
User avatar
ed
retired
Posts: 5384
Joined: Thu Jun 21, 2007 4:01 pm
Location: yorkshire
Contact:

#179 Re: GM70 PP

Post by ed »

I did ask this question way back in the mists of time...before dismantling transformer ends for spraying..

I've used 2 methods:
I've taken the bellends off and given to my favourite painter to do them along with the chassis
and
Ive done them the lazy way by wrapping masking tape around the stack and just spraying them as a whole....

I discovered in my travels and dealings with BLS when I took the bellends off one of their power Tx that there were no top hat washers on the bolts. These transformers were pants!...they were entirely different when I installed some after sale washers. I would say best to make sure you put the washers back when re-assembling.
There's nowhere you can be that isn't where you're meant to be
User avatar
izzy wizzy
Old Hand
Posts: 1496
Joined: Fri Nov 02, 2007 7:02 pm
Location: Auckland NZ
Contact:

#180 Re: GM70 PP

Post by izzy wizzy »

Progress being made on repackaging the amps. They're considerably narrower, same depth and will be a bit higher though probably not as much as in the photos.

I'm very happy with the BB ply/pcb sandwich and how it fits in the top plate recess.

So far the layout work in sketch up is proving to be very helpful.

IMG_20220409_133846691_compress52.jpg
IMG_20220409_133846691_compress52.jpg (174.32 KiB) Viewed 5750 times
IMG_20220409_134122900_compress48.jpg
IMG_20220409_134122900_compress48.jpg (173.15 KiB) Viewed 5750 times
IMG_20220409_134153551_compress40.jpg
IMG_20220409_134153551_compress40.jpg (158.71 KiB) Viewed 5750 times
Post Reply