Jump to content

Group Build: Dynafet


luvdunhill

Recommended Posts

here are the 4 board layouts suitable for production

http://gilmore.chem.northwestern.edu/dynahifeti.zip

http://gilmore.chem.northwestern.edu/dynahifetihs.zip

http://gilmore.chem.northwestern.edu/dynahifetr.zip

http://gilmore.chem.northwestern.edu/dynahifetrhs.zip

schematic above is correct except that i doubled the number of output transistors.

Link to comment
Share on other sites

here are the 4 board layouts suitable for production

http://gilmore.chem.northwestern.edu/dynahifeti.zip

http://gilmore.chem.northwestern.edu/dynahifetihs.zip

http://gilmore.chem.northwestern.edu/dynahifetr.zip

http://gilmore.chem.northwestern.edu/dynahifetrhs.zip

schematic above is correct except that i doubled the number of output transistors.

Remember that all the board files i posted use the dual jfets and the dual

cascode parts. All of which are almost impossible to get.

I'll need to take a look at the board files (I assume Gerbers), and think from what I have read that the Toshiba FETs, external heatsinks (dynahifetrhs.zip) is the one I want to look at?

As for the schematic, there is a pair of devices in it... these double to 4 total, the additional pair in parallel with the existing pair I assume (including addl drain resistors (is that the right term))?

Dual parts: thru matching, I think you can get better matches than the duals give you... one of my duals has to be 20% off in matching. For the dual BJTs, would something like 2SA970/2SC2240 be a good substitute?

A couple of requests on the boards: I always felt the pad sizes on the Dynalo/Dynahi group buy boards were too small, making soldering connections to the ground plane or power planes difficult, and making desoldering from these same areas really hard. Secondly, the footprint used for the dual parts on the Dynahi boards did not allow the use of Zip sockets to socket the devices. The Zip sockets I feel are a must to allow, even though we will be using pairs of single devices here. Like I said, I'll look at the gerbers.

Link to comment
Share on other sites

So for a complete ignorant on the Dynafet's topology, please let me know: To make a 2ch SE amp you need two or three boards like in the B22? To make a "proper" balanced and SE thing you need 4 or 6 boards?

Moreover, you need 1 S22 for every 2 or 3 boards, or would it be better matching one S22 with a single Dynafet board?

Link to comment
Share on other sites

Dual parts: thru matching, I think you can get better matches than the duals give you... one of my duals has to be 20% off in matching. For the dual BJTs, would something like 2SA970/2SC2240 be a good substitute?

yup, this is the idea and what the heat sinks referenced in this thread bring to the table. I've talked to the manu. and he is willing to do another run. If we're talking about 50 boards, then this comes to 200 heat sinks. I need to work out cost with him, as this run would be smaller than the previous.

Also, the cost for the matched JFETs won't be insignificant. 1000 of each type N and P was not a insignificant investment (look at pricing at B&D for example), plus the average of 3 minutes per device to match (including calibrating the test jig every 10 devices). This plus the heat sinks and you're not really looking at much of a "savings" over the traditional 7-zip packages. However, still cheaper than unobtanium.

Then there are the MOSFETs.... I'll work on that today.

Link to comment
Share on other sites

Wish I could offer help more.

One thing that is quite cool with my balanced amp is the ability for it to take a single-ended input and use transformers to turn it into balanced output.

The transformers used are Sowters which are a UK company. If I get an indication of how feasible this is and how many people are interested, I could contact them and provide pricing and options. Added bonus is the pound is currently very low compared to the dollar....

Cheers!

Link to comment
Share on other sites

A couple of requests on the boards: I always felt the pad sizes on the Dynalo/Dynahi group buy boards were too small, making soldering connections to the ground plane or power planes difficult, and making desoldering from these same areas really hard. Secondly, the footprint used for the dual parts on the Dynahi boards did not allow the use of Zip sockets to socket the devices. The Zip sockets I feel are a must to allow, even though we will be using pairs of single devices here. Like I said, I'll look at the gerbers.

This appears to be the same case here, as the duals were implemented using individual TO-92s rather than the 7-pin dual socket. Spacing between the pin rows is not high enough to support sockets of any type. Pad sizes and shapes are also on the small side. I prefer elongated pads on transistors, for example.

I'm looking at the dynafetr files using external heatsinks.

Link to comment
Share on other sites

This appears to be the same case here, as the duals were implemented using individual TO-92s rather than the 7-pin dual socket. Spacing between the pin rows is not high enough to support sockets of any type. Pad sizes and shapes are also on the small side. I prefer elongated pads on transistors, for example.

I'm looking at the dynafetr files using external heatsinks.

yup, those are the correct boards.

well, if we did want to rework the boards, I'd love to see the JFET comp. pairs back-to-back, as referenced at the start of this thread.

Link to comment
Share on other sites

Wish I could offer help more.

One thing that is quite cool with my balanced amp is the ability for it to take a single-ended input and use transformers to turn it into balanced output.

The transformers used are Sowters which are a UK company. If I get an indication of how feasible this is and how many people are interested, I could contact them and provide pricing and options. Added bonus is the pound is currently very low compared to the dollar....

Cheers!

I might be interested.. but I think it'd have to be worked out with whoever helps out with the building. So let's check in again when its further down the line :)

Link to comment
Share on other sites

okay, I'm taking some liberties here to firm up numbers and make the math work:

  • luvdunhill - 8 - 4
  • Naamanf - 6 - 4
  • looser101 - 4 - 2
  • n_maher - 4 - 0
  • Digipete - 8 - 4
  • Pars - 6 - 2
  • Fing - 4 - 4
  • Asr - 0 - 4
  • Stretch - 0 - 4
  • PFKMan23 - 0 - 4
  • Icarium - 0 - 4
  • Dreadhead - 0 - 4

laxx, guzziguy, pabbi1 any interest? This isn't a commitment (yet) just need to firm up some numbers for the quantity discounts and what not.

Edited by grawk
Link to comment
Share on other sites

here are the 4 board layouts suitable for production

http://gilmore.chem.northwestern.edu/dynahifeti.zip

http://gilmore.chem.northwestern.edu/dynahifetihs.zip

http://gilmore.chem.northwestern.edu/dynahifetr.zip

http://gilmore.chem.northwestern.edu/dynahifetrhs.zip

schematic above is correct except that i doubled the number of output transistors.

Kevin,

Does the doubling number of output transistors affect the sound quality, say if driving a HD650 or K701? Just trying to get a handle on what are the upsides/downsides of 2 vs. 4 output transistors.

Cheers,

Pete

Link to comment
Share on other sites

Kevin,

Does the doubling number of output transistors affect the sound quality, say if driving a HD650 or K701? Just trying to get a handle on what are the upsides/downsides of 2 vs. 4 output transistors.

Cheers,

Pete

I don't remember. If you make a board that can take all of them, you can populate

only half if you want to. Probably makes absolutely no difference unless you are

driving 8 ohm speakers.

Link to comment
Share on other sites

I'll need to take a look at the board files (I assume Gerbers), and think from what I have read that the Toshiba FETs, external heatsinks (dynahifetrhs.zip) is the one I want to look at?

As for the schematic, there is a pair of devices in it... these double to 4 total, the additional pair in parallel with the existing pair I assume (including addl drain resistors (is that the right term))?

Dual parts: thru matching, I think you can get better matches than the duals give you... one of my duals has to be 20% off in matching. For the dual BJTs, would something like 2SA970/2SC2240 be a good substitute?

A couple of requests on the boards: I always felt the pad sizes on the Dynalo/Dynahi group buy boards were too small, making soldering connections to the ground plane or power planes difficult, and making desoldering from these same areas really hard. Secondly, the footprint used for the dual parts on the Dynahi boards did not allow the use of Zip sockets to socket the devices. The Zip sockets I feel are a must to allow, even though we will be using pairs of single devices here. Like I said, I'll look at the gerbers.

I wonder if Kevin or someone would work with this group to make design changes to the current board design to take into account single vs dual transistors, and other useability considerations?? ...since this seems to be the best shot to get boards built for the DynaFet.

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
×
  • Create New...

Important Information

By using this site, you agree to our Terms of Use.