The Menace

[n_maher]

seems like a perfect situation for tv damper diodes... Would you mind sharing the choke part number?

Agreed, that was one of the 3 options that we came up with for the power supply, the reason that it wasn't used is that it would have required yet another transformer (for the heaters). When I post the schematics I'll post that power supply option as well. As with anything DIY this is just one way of doing it and as usual each had their own compromises/sacrifices.

The large choke is a Hammond 193Q, the two smaller chokes are Hammond 1193H's.

And I just noticed while looking at the schematics that I have the caps mounted in the wrong positions, d'oh!

[luvdunhill]

seems like a perfect situation for tv damper diodes... Would you mind sharing the choke part number?

I'm assuming the 193Q is the first choke.. no way you're trying to use choke input, so I'm assuming there's a small cap before that. even so, I would be curious if you can actually get 300mA... I went down a similar road and pretty much gave up after simulating with psud2. needless to say, I'm interested to see how this works

[luvdunhill]

sounds great! I'd be interested to see the actual current capabilities once it gets up and running

[n_maher]

I'm assuming the 193Q is the first choke.. no way you're trying to use choke input, so I'm assuming there's a small cap before that.

Yes, the 193Q is the first choke. Choke input was looked into but given the current requirement (again, this is my understanding) this wasn't possible unless I wanted to get a custom choke made.

even so, I would be curious if you can actually get 300mA... I went down a similar road and pretty much gave up after simulating with psud2. needless to say, I'm interested to see how this works

300mA? You mean on the downhill side of the first choke? If so I sure hope I can get more out of it than that since the design was intended to produce ~380mA. And while this is a new design a similar one has been used successfully by the designer in the past so we both have a pretty high level of confidence in what has been sketched out so far.

[luvdunhill]

the reason that it wasn't used is that it would have required yet another transformer (for the heaters).

check out the EY500. I was going to use this due to the 6.3v heaters and avoiding the additional transformer.

[luvdunhill]

Yes, the 193Q is the first choke. <snip> If so I sure hope I can get more out of it than that since the design was intended to produce ~380mA.

yes. I looked at the 193Q and found some people saying that 220mA or so was more reasonable to expect. Here's one such post:

http://www.audioasylum.com/cgi/t.mpl?f=tubediy&m=139326

I suppose I lack the experience in this area to weed through random crap on the web and pretty much believe everything I read

[n_maher]

yes. I looked at the 193Q and found some people saying that 220mA or so was more reasonable to expect. Here's one such post:

http://www.audioasylum.com/cgi/t.mpl?f=tubediy&m=139326

I suppose I lack the experience in this area to weed through random crap on the web and pretty much believe everything I read

All of those posts are related to a choke input, correct?

[fierce_freak]

Looks like it.

[luvdunhill]

All of those posts are related to a choke input, correct?

"OTOH, if a pseudo choke I/P filter, with its "fudge factor" cap. is employed, you can safely draw a bit more current"

there were more posts I found.. anyways, if there is an open Q&A at some point, I'd be interested to learn more.

[luvdunhill]

"OTOH, if a pseudo choke I/P filter, with its "fudge factor" cap. is employed, you can safely draw a bit more current"

there were more posts I found.. anyways, if there is an open Q&A at some point, I'd be interested to learn more.

edit: seems to me you're going to have arcing issues with the 5AR4 from exceeding it's PIV and in order to get that choke to work, you'll probably exceed the recommended input cap for the 5AR4 as well. If at all possible, I'd like to learn why this will work

[n_maher]

check out the EY500. I was going to use this due to the 6.3v heaters and avoiding the additional transformer.

Sorry, missed this earlier, the diodes that we were considering were 6.3V heaters and that still would have meant running an additional transformer. It's my understand that running the rectifier and say the input tube off of the same heater supply is not a good idea at all. Not to mention that in this amp both the input and output tube heater supplies are referenced to B+ and floated above ground.

edit: seems to me you're going to have arcing issues with the 5AR4 from exceeding it's PIV and in order to get that choke to work, you'll probably exceed the recommended input cap for the 5AR4 as well. If at all possible, I'd like to learn why this will work

I've run your initial question by the designer and I'll add this information to the query and see what he says. Let's hope there aren't major issues with the PS design as that would put a serious hurting on my hopes of getting this thing running any time soon.

[luvdunhill]

I've run your initial question by the designer and I'll add this information to the query and see what he says. Let's hope there aren't major issues with the PS design as that would put a serious hurting on my hopes of getting this thing running any time soon.

thanks for being so understanding Nate

[n_maher]

thanks for being so understanding Nate

Hey, my only vested interest in this is making the design work. If there are issues I'd rather know about them now rather than when I flip the switch!

[luvdunhill]

Hey, my only vested interest in this is making the design work. If there are issues I'd rather know about them now rather than when I flip the switch!

here I thought you were interested in helping me learn ... oh well, you're still a decent guy in my book.

[n_maher]

here I thought you were interested in helping me learn ... oh well, you're still a decent guy in my book.

Oh, I'm interested in the learning part too, all I meant was that if there's a mistake/error/whatever that I'm not interested in defending the design and ignoring the fault.

[luvdunhill]

cool. I just went down an almost identical path 6 months ago and wussed out buying the parts for something, when in my mind there were red flags. I didn't really know anyone who was willing to help me with it, so I ditched it.

[n_maher]

cool. I just went down an almost identical path 6 months ago and wussed out buying the parts for something, when in my mind there were red flags. I didn't really know anyone who was willing to help me with it, so I ditched it.

I guess I'm lucky enough to have a friend who when I approached him with the idea ran with it. I guess it's time to explain this whole mess, but I'll create a separate post for that...

[n_maher]

A Brief History of The Meance

So really this started years ago, but if I tell the story that way it won't make a bit of sense to anyone other than myself. And I did say brief history and I swear I'll try to keep it that way.

This really got going in late December of this past year when I got a message from a friend (who shall remain nameless) that they had an abandoned project that I might be interested in picking up. The basic gist of the project was that they had taken the Van Waarde project along with the Wheatfield HA-2 design and started to improve upon the general concept. Most notably this involved paralleling the output tubes and increasing the output coupling capacitance to the point where driving low-impedance headphones would be possible. The original HA-2 was technically able to do this but not all that well in my opinion. So me being the unabashed Pete Millett fan that I am and a former HA-2 owner jumped all over it. Quick note - both the Van Waarde and the HA-2 schematic are out there if you know where to look, the Waarde is at Headwize (linked above) and owners of the HA-2 know where to find its schematic. I believe that Headroom still owns the HA-2 design and as such I won't be publishing that schematic, only my revised version. Now the revised design that started this mess had been taken to the proof of concept level but in such a way that things like the power supply still needed figuring. So while I won't be using much of what was done as a part of that project I do owe my friend a huge debt of gratitude in terms of motivation and inspiration, without that initial push I'm sure that this would have never happened.

Now I'm no designer, I never have been and I never will be since there just aren't enough hours in the day and honestly I'm completely comfortable leaving that up to other folks. I love building this stuff but typically my interest dies off quickly after that point. To a degree that had to change for this project since there was still much work to be done. So I started doing some basic math along with consulting with a few folks about the power supply. It was initially my hope to simply re-use the existing HA-2 tube-rectified design and tweak it as needed. This was quickly abandoned after seeing that the existing PS design was pretty much tapped in terms of current capacity. It was at this point that I decided to call in the big guns and see if I couldn't enlist the help of the guy who was really responsible for all this. I've communicated off and on with Pete Millett for the past few years, mostly as a result of my dealings with the Low Voltage Hybrid project as well as dealing with the issues that my original HA-2 had. Like I said, I'm a hug Pete fan. But Pete's a busy guy and I didn't really expect to have him launch into it the way that he did but it'd be safe to say that this project may well have died without his help. He continues to be a integral part of the design process and I can't say thanks enough so, thanks Pete! I started talking power supply design with Pete and we quickly came up with a couple of concepts and I elected to go with what seemed to be the simplest from a parts perspective (fewer transformers and bits that I was used to working with). I also elected to have the amplification side of the equation be essentially a tweaked HA-2 that will run parallel output tubes, larger output coupling capacitors and modified heater supplies. This should help some of the issues that the original HA-2 suffered from which were some PS hum and output impedance that wasn't really all that suited for driving Grados.

I think I'll stop here and leave the details to later on in the process but hopefully that paints at least a little bit of a picture as to where this came from and where it's headed.

[fierce_freak]

Great post, thanks Nate! I was thinking it was going to be along the lines of the Waarde or Wheatfield. I'm certainly interested in seeing the completion of your project...looks flippin' sweet.

[n_maher]

Ok, so I think part of the confusion is due to an error in something I posted earlier, no surprise there.

Each rectifier will be fed 260VAC, not 385 as previously mentioned. Serves me right for posting without the schematic in front of me and after a long day. Hopefully that clears at least part of this mess up, the rest will come from Pete. He's sent me a few simulations of the power supply and we're taking a few more measurements but the long and the short of it is that everything is perfectly within limits with the exception of peak current, which is borderline. We're going to run a few more simulations with updated data and I'll report the full results as soon as I have them.

[luvdunhill]

Each rectifier will be fed 260VAC, not 385 as previously mentioned. Serves me right for posting without the schematic in front of me and after a long day.

5AR4 seems much more sane now. If you're using psud2, then checking for current saturation in the choke should be easy. High current is why dual core chokes are often used, like LL1673. I personally think a dual core choke would be more appropriate for the current you're talking about.

Concerning the first capacitor, be wary of heavy AC charge current if you're using something larger than 33uF or so. If you push this, both the lifetime of the tube and the cap can suffer. The circuit I was experimenting had problems with hum when the input capacitor go too large. I found that using a series limiting resistor to control peak in-rush current helped. This is usually calculated from the coil resistance of the transformer and then adding a small resistance inline to reach some target value for the rectifier used. Check the datasheet to determine this. Also, the sound was heavily influence by the grounding arrangement. Most important seemed to be the diode connections. Remember, anything connected to them will radiate magnetic field, so keep the wires short and close to the chassis. This really helped the sound, IMHO.

This is about the extent of my experience on the matter I'm definitely not an expert, just a tinkerer.

[n_maher]

5AR4 seems much more sane now. If you're using psud2, then checking for current saturation in the choke should be easy. High current is why dual core chokes are often used, like LL1673. I personally think a dual core choke would be more appropriate for the current you're talking about.

You can't use PSUD to model parallel rectifiers but I know that Pete has been using it to semi-simulate the situation here. And your suggestion of the dual core choke is good and noted for future use. Like I said, I may not stop with this initial build and may tear the whole thing back down depending on what my ears tell me when it's done.

Concerning the first capacitor, be wary of heavy AC charge current if you're using something larger than 33uF or so. If you push this, both the lifetime of the tube and the cap can suffer.

Yup, this is one of the things that we'll be watching. The cap is currently sized at 40uF and the option exists to add a series resistor if problems arise or change out the cap.

The circuit I was experimenting had problems with hum when the input capacitor go too large. I found that using a series limiting resistor to control peak in-rush current helped. This is usually calculated from the coil resistance of the transformer and then adding a small resistance inline to reach some target value for the rectifier used. Check the datasheet to determine this.

Pete is re-running some of the simulations based on actual data that I've given him from the trafo and we'll adjust things as required.

Also, the sound was heavily influence by the grounding arrangement. Most important seemed to be the diode connections. Remember, anything connected to them will radiate magnetic field, so keep the wires short and close to the chassis. This really helped the sound, IMHO.

I appreciate all the advice and sanity checks, it's always better to have multiple sets of eyes on things. I'll be doing my best to keep things inline as far as wire routing goes but unfortunately some things about this amp are dictated by the physical size of the parts involved. I'm sure that I could have made things a little bit smaller in the PSU chassis I've designed so far but not by much without turning the top panel into worthless swiss cheese that wouldn't have been capable of supporting

This is about the extent of my experience on the matter I'm definitely not an expert, just a tinkerer.

You're way ahead of me in regards to this so don't sell yourself short.

[dc]

[Dusty Chalk]

Can I come and visit you when that is done?

[n_maher]

Can I come and visit you when that is done?

Heh, we'll have to see about that.

In real news I got about 60% of the power supply wired tonight. It's going much better than expected really and I hope to wrap up the majority of it tomorrow night. The amp chassis will arrive Wednesday and I've spent a bit of time working with its layout in CAD. It's pretty hard for me to believe that the amp, which largely consists of a bunch of resistors, is proving to be the bigger layout headache. And that's not even taking into consideration the wiring aspect of the amp, just the physical layout of the bits and pieces. I see some late nights in my future...