Group Build: Dynafet

[DigiPete]

So, a bit of an update. Things are slowly coming together. I have 2 payments so far using "e-check", which there seems to be a flat $5 fee for using. I'd recommend using this. However, one thing to note is it can take a while. I really need a bulk of the money in my bank account (which means I then have to transfer from PayPal) on Monday. This is looking like this may be hard to do. So, if you're using an "e-check" I'd really like you to start the process by COB tomorrow. Same with mailing a paper check, or money order. If you're using other PayPal methods, you have a bit of time, but don't wait until the last second, please. Also, try and use a PayPal calculator to figure in the fees. Let's not drag this out too long guys.

Team Spain: YGPM

Icarium: Feel free to chime in if I missed any salient details.

So I missed the post with the details, what do you need Luvdunhill?

[Icarium]

He needs the money for the transistors (This covers all transistors needed for the project) which is 60 bucks per board (240 for you) + any sort of fees.

If you want to pay with echeck for the flat 5 dollar fee or paper check/money order you will have to get it out today. He is more flexible with normal paypal, but he needs the money prior to monday so he can do his orders. If you need his paypal info or mailing address ill send that to you now.

[DigiPete]

payment sent.

[Pars]

Short might happen? Vibration? In my ignorance of technical matters and considering case heat sinking was an option, I thought it could be a good thing Would you please would be so kind to explain why to me

I'm not vvs, but I'll take a guess. The heatsinks are probably tied to board ground (looks like it on the sigma22). Normally you would use a ground loop breaker, which is a low value resistor and x-rated cap (AC line) in parallel, from the board ground to the case (i.e., wall ground). The heatsink touching could potentially short this ground loop breaker out, rendering it useless.

[n_maher]

I believe the heat sinks on all of amb's boards are 100% isolated. Yes, they have solder pads and thru-hole plating but they are not tied to the ground plane that I can see.

[Pars]

In the board pics on amb's site, it looked like the bottom ground plane extended under the sinks. I couldn't tell for sure whether the holes were actually tied to the ground plane or not, and being at work, can't measure one.

[Torpedo]

I believe the heat sinks on all of amb's boards are 100% isolated. Yes, they have solder pads and thru-hole plating but they are not tied to the ground plane that I can see.

So then, what's the problem in extending the sinking area to the case, by letting the heat sinks to touch the box?

[naamanf]

So then, what's the problem in extending the sinking area to the case, by letting the heat sinks to touch the box?

Nothing as long as you electrically isolate the MOSFET from the heat sink. No different than having a off board heat sink that is part of the chassis.

[n_maher]

Nothing as long as you electrically isolate the MOSFET from the heat sink. No different than having a off board heat sink that is part of the chassis.

What he said, there's no reason not to provided the proper precautions are taken.

[Torpedo]

Thanks, I thought so, but since previously it was stated as a very bad idea, I just wanted to know why

Now we just need to try how demanding are 2 Dynafet boards for one S22 and see if the big heat sinks that will come in the kit are enough to keep it cold.

[Icarium]

My understanding is it depends a lot on how you bias it and that's somewhat a personal choice.

[vvs_75]

Nothing as long as you electrically isolate the MOSFET from the heat sink. No different than having a off board heat sink that is part of the chassis.

X2

Onboard heat sinks can touch the case (or top panel) if you mount MOSFETs on isolating pads as you would do with offboard heat sinks. In this case should be no problem.

Since onboard heat sink electrically do not connected to anything then you can mount transistor without using isolating TO-220 mounting kit.

MOSFET mounting tab internally connected to the base pin, so when you short it you may and probably will fry the transistor. Never happened to me though.

Since I didn

[DigiPete]

Thanks, I thought so, but since previously it was stated as a very bad idea, I just wanted to know why

Now we just need to try how demanding are 2 Dynafet boards for one S22 and see if the big heat sinks that will come in the kit are enough to keep it cold.

It won't be cold, you just want to make sure the chosen heat sink keeps the transistor within reasonable temperature limits.

[Torpedo]

It won't be cold, you just want to make sure the chosen heat sink keeps the transistor within reasonable temperature limits.

I know, I tried to mean it were "cold enough" to be in a safe working range I wonder if someone here has tried already the Dynafet using S22 PS and if it really needed case heat sinking or whether it run smooth with onboard heat sinks. If the big on board heat sinks we're planning to use aren't enough we'll be in trouble.

[Icarium]

I don't think anyone has tried it.

[luvdunhill]

ok, we are getting close to being set on initial payments. I have made arrangements or received payments from all but 1 member.

At this point, Icarium is going to PM you with details about the sigma22 orders. We can go ahead and move forward on those.

I am working on the input transformer side of things. I will be taking orders for two different transformers and two different options on each. Details here:

diyAudio Forums - GB: Sowter 10K:10K Input Transformer - Page 1

... "current info" here

diyAudio Forums - GB: Sowter 10K:10K Input Transformer - Page 2

... and sign-up and more concise information is here:

diyAudio.com Wiki - projects by fanatics, for fanatics

I'm committed to getting this working, as I will be using this on my own Dynafet. I will buy a pair of each model to experiment with both options. As I've said before, these transformers really prefer a constant impedance volume control, like a stepped attenuator. At worse case, we would have to use a relay based attenuator that mimics a stepped attenuator (lots of relays) or a actual stepped attenuator, if we couldn't get some sort of hybrid arrangement working. Just a small disclaimer.

From my calculations, this group buy will probably put us right where we need to be (50 Cinemags and 12 Sowters), and once this happens, I'll close the group buy shortly thereafter and begin invoicing. Please read the first thread posted above.

*** If you want to, please PM me or Icarium your preference and we will add you to the DiyA wiki ***

[Icarium]

Okay kinda slacked on getting these out since 1. they are kind of confusing and unwieldy as excel spread sheets and 2. been real busy and sick

But I need email addresses from:

naamanf

torpedo

looser101

nate

To forward you your excel spreadsheet. I think it's mostlyyy self explanatory the only confusing parts are how to deal with paypal and shipping. I thinkkk you just 0 out paypal if you aren't going pay with that and that will adjust the total and shipping is basically based on international and domestic. The larger orders (Mine) and the Spanish one is a little tricky since he had to do some guess work to determine if it'll fit so that may be off. He suggests that Looser101 go with first class instead of priority as that will equate to more like 10 bucks rather than 24.

[Pars]

I know, I tried to mean it were "cold enough" to be in a safe working range I wonder if someone here has tried already the Dynafet using S22 PS and if it really needed case heat sinking or whether it run smooth with onboard heat sinks. If the big on board heat sinks we're planning to use aren't enough we'll be in trouble.

I've never seen pics of the one(s) that KG built, and I don't recall him saying what he used for a PSU in them. He's the only one I know of who has built a DynaFET.

On another note, have board pics been posted yet?

[n_maher]

But I need email addresses from:

nate

To forward you your excel spreadsheet.

What do I need a spreadsheet for?

[luvdunhill]

What do I need a spreadsheet for?

have you purchased from Glass Jar Audio before?

[n_maher]

Yes, several times.

[Smeggy]

Nate, or other experienced B22 maker, how hard do you think switching between B22 and Dynafet boards would be so they can both be run from the same Sigma PSU via either a switch or cable plug? I'm not asking you to do it, just seeing if you can think of a neat and practical way of switching amps in the same case.

Maybe sigma> umbilical> dual inputs (beta, Dyna) each having their own inputs/outputs and vol control in one case.

Whadda you guys think?

[luvdunhill]

Yes, several times.

well, the spreadsheet is just the same spreadsheet that Jeff sends as an invoice.

[n_maher]

Sounds insanely complicated to me if you want a physical switch to be able to do it on the fly. Even if you leave the switching of the power to a passive device (say, different umbilical inlets for the different amps) you're still stuck with how to switch the inputs/pot so that they feed the right amp. At least I don't think you'd just want to wire them in parallel with each other. But I freely admit I've only given this about 2 minutes of thought and it deserves a lot more.

[grawk]

SPOT switch

(single pole oct throw)