The ultimate DIY? A Stax SRM-T2!

[bui501]

6 hours ago, GeorgeP said:

hmmm, I thought that that boardset was a work in progress and never completed... maybe Kevin will recall.

George, do you know if t2shrunkv2 is also in the same state (of incompleteness)? From the early days of the thread, I thought Kevin had laid out that particular board for Lil’knight back then.

[Kerry]

11 hours ago, JoaMat said:

Above post show changes to the balance servo I’ve made.

Below schematic shows changes made to get the KGOOS (Opto Offset Servo)  -  the only offset servo you need.

Q33 and R73 replaced by a small daughter board and the resistor strings changed accordingly.

For reference here is t2schemop.pdf, schematic of the STAX T2 output section. Drawn by Kevin Gilmore and redrawn by Linear.

Nice Joachim

I've been playing with this servo in LTSpice as well.

Here's what I came up with...

I added a 1K resistor across Q5.  I didn't have R17, but I like it and it seems to model better.  D2 is probably not needed but wanted to protect Q5, just in case.

Have you tried it without the 2.2pF cap across the 2 Meg resistors.  I'd prefer not to use this.

I'm enjoying the show  

[GeorgeP]

2 hours ago, bui501 said:

George, do you know if t2shrunkv2 is also in the same state (of incompleteness)? From the early days of the thread, I thought Kevin had laid out that particular board for Lil’knight back then.

The lil knight shrunk board has one or two small board errors that can be easily fixed, but the amp board really isn't much smaller than the original (about 1/2" smaller), and the psu board is basically the same as the original. Also they use the same parts as the original diy boards.

[chinsettawong]

Lil Knight board's quality is really low though.

[JoaMat]

19 hours ago, bui501 said:

(1) One of Kevin’s later boards (t2shrinkedv10) has stn9360’s shown in the layout as TO220 parts, but when I did a search for stn9360 on mouser (and digikey), the part shows up as an smd component... does the stn9360 still come in non-SMD format? If so where is a reliable source? If not, what is a suitable replacement?  Can I mount the smd version on a smd-to-TO220 adapter board and use the smd parts, or are they not equivalent components?  I can’t find any data sheet for a non-smd version of the stn9360, so I do not know what the specs are.

(2) the board has a number of parts with leads rather close together with no outlines and with just a simple number by each pair of pads.  Are those resistors to be mounted vertically?

thanks for any helpful guidance.

 

 

 

 

Below is a picture of my modified DIY T2. Zoom in and you can finds several resistors standing. There are also  standing STN9360. You won’t find other types of STN9360. Those homemade boards have been in service for four years now. I’ve made several modifications and the boards have been very kind to me. Kevin’s t2shrinkedv10 is based on my home cooked boards. I’m confident his board is OK. But be aware if something goes wrong it could be a tricky thing working on that board as some areas are pretty crowded. Today I would make a different layout of the board as I find the board a bit obsolete.

Edited April 6, 2018 by JoaMat

[JoaMat]

19 hours ago, luvdunhill said:

Interesting, I added a simple trimmer here instead. Would these modifications work with the default batterys? I would think so. I am tempted to try these changes out

 

What working voltage is needed for the 10uF cap and the one in the same position for the previous schematic (feedback/integrator cap)?

Yep, I remember you used a trimmer – that’s why I came to use trimmer on my modified T2. I think working voltage 50 V will do. Try  the servo - it's easily done.

 

9 hours ago, Kerry said:

Nice Joachim

I've been playing with this servo in LTSpice as well.

Here's what I came up with...

I added a 1K resistor across Q5.  I didn't have R17, but I like it and it seems to model better.  D2 is probably not needed but wanted to protect Q5, just in case.

Have you tried it without the 2.2pF cap across the 2 Meg resistors.  I'd prefer not to use this.

I'm enjoying the show  

Thanks Kerry,

LTspice is really a very useful tool.

I’ve deliberate skipped the D2 zener (Kevin uses the Zener) to see what happens and so far no failures. The R29 resistor is probably a good idea but I’ve skipped that too as the servo seems to work all right without it.

As for C1 I use 1uF/50V in my Carbon and GG with tube input (same as Kevin later designs). But on the T2 I had some problem with 1uF so after some testing I ended up with 6.8uF (have just opened the amp to check).

Regarding the 2.2pF – I actually added more capacitance to reduce some oscillation at outputs seen on the oscilloscope. I probably better go back and have look at it again.

DIY is fun

[luvdunhill]

Is the opto CTR important at all? I have some CNY17 in a few grades on hand. I use these to bias some output devices in power amp design.

 

Ok, you have convinced me. I will get it queued up in the project list [emoji4]

 

[bui501]

5 hours ago, JoaMat said:

 

Below is a picture of my modified DIY T2. Zoom in and you can finds several resistors standing. There are also  standing STN9360. You won’t find other types of STN9360.  

 

Thanks, JoaMat. The stn9360’s are what still puzzles me—when i searched for STN9360, this part (data sheet below) is the only one that shows up...  Is this the same transistor that you used just in a vertical form factor back then?

 

[GeorgeP]

12 hours ago, chinsettawong said:

Lil Knight board's quality is really low though.

Yes, his first run full-size boards were not the greatest if you had to do any desoldering.

[JoaMat]

5 hours ago, bui501 said:

Thanks, JoaMat. The stn9360’s are what still puzzles me—when i searched for STN9360, this part (data sheet below) is the only one that shows up...  Is this the same transistor that you used just in a vertical form factor back then?

 

Yes, it’s the same SMD transistor. I just soldered them in standing to save space. As a DIYer you are entitled to do whatever you want.

If going "through hole parts" I like that layout a lot. Then one should change 2sa79/2sk216 to ksa1229a/ksc2690c and make output stages cascode current sources. That would be a great amplifier with available parts at reasonable prices.

Edited April 6, 2018 by JoaMat

[bui501]

8 hours ago, JoaMat said:

Yes, it’s the same SMD transistor. I just soldered them in standing to save space. As a DIYer you are entitled to do whatever you want.

If going "through hole parts" I like that layout a lot. Then one should change 2sa79/2sk216 to ksa1229a/ksc2690c and make output stages cascode current sources. That would be a great amplifier with available parts at reasonable prices.

Awesome!  Thanks so much for the clarification on the smd transistor and the advice re. the ksa1229a/ksc2690c.

(I had kind of guessed it was your name along with Kevin's on the shrunkedv10 board, but wasn't sure -- sorry, I'm a bit dense on these things. )

[bui501]

22 hours ago, chinsettawong said:

Lil Knight board's quality is really low though.

Thanks for the warning.  I'll stay away from those boards. 

I've been looking at the gerber files for the shrunkv2 board and the shrinkedv10 board with the thought of either making my own from the gerbers or sending it off for a small batch run.

23 hours ago, GeorgeP said:

The lil knight shrunk board has one or two small board errors that can be easily fixed, but the amp board really isn't much smaller than the original (about 1/2" smaller), and the psu board is basically the same as the original. Also they use the same parts as the original diy boards.

Thanks, George -- 1/2" doesn't seem like much, but it's a big savings in case work that I'd rather just put toward the electronics, which is why I was looking at the shrunk version and JoaMat's shrinked10 version of the boards.

Do you happen to know if the board errors were just in the particular run of boards that Lil Knight did back then or if they were inherent in the gerber files?  I was planning to either etch my own board (ala JoaMat style) or do a small batch run with some pcb prototyping service.

[GeorgeP]

13 hours ago, bui501 said:

Thanks for the warning.  I'll stay away from those boards. 

I've been looking at the gerber files for the shrunkv2 board and the shrinkedv10 board with the thought of either making my own from the gerbers or sending it off for a small batch run.

Thanks, George -- 1/2" doesn't seem like much, but it's a big savings in case work that I'd rather just put toward the electronics, which is why I was looking at the shrunk version and JoaMat's shrinked10 version of the boards.

Do you happen to know if the board errors were just in the particular run of boards that Lil Knight did back then or if they were inherent in the gerber files?  I was planning to either etch my own board (ala JoaMat style) or do a small batch run with some pcb prototyping service.

They were in the gerbers, but Kevin might have corrected them. That said the error was simple to correct (and is documented in this thread). I do have one of the blue amp boards before the correction - almost too high quality with 4oz copper and very thick boards (you need a good iron to work on these boards).

[Kerry]

@JoaMat

Here are some boards I drew up...

    

[JoaMat]

Wow, that’s wonderful! U1 - op amp in sot23-5 package, any suggestion?

[Kerry]

I used OPA197IDBVR in the GRLV micro (same here).

Edited April 8, 2018 by Kerry

[bui501]

10 hours ago, GeorgeP said:

They were in the gerbers, but Kevin might have corrected them. That said the error was simple to correct (and is documented in this thread). I do have one of the blue amp boards before the correction - almost too high quality with 4oz copper and very thick boards (you need a good iron to work on these boards).

Found it!!—(The missing trace in the original shrunk board).  Thank you, George.  Looks like Kevin fixed it in shrunkedv2.  Problem with reading this thread all the way thru at once is that everything is running together.  Either that or I’m just getting old.

[socalguy]

If anybody have spare boards or plan on a group buy, I'd be interested in 2 sets. 

Thanks.

[bui501]

Would Sanyo 2SK2160 be a suitable substitute for the 2SK216? It looks more readily available. (Data sheet is attached.)

 

2SK2160.pdf

[kevin gilmore]

10 times the input capacitance, 20 times the output capacitance. Yes it will work as the gate follower, but in the gain stage the frequency response is definitely going to change

[bui501]

16 hours ago, kevin gilmore said:

10 times the input capacitance, 20 times the output capacitance. Yes it will work as the gate follower, but in the gain stage the frequency response is definitely going to change

Thanks, Kevin.  Was hoping I had stumbled on a cheap drop-in replacement. :-(

 

[8street]

On 15.04.2018 at 4:11 PM, kevin gilmore said:

10 times the input capacitance, 20 times the output capacitance.

Also the problem that this transistor is even more enhancement than SK216 and also more current. Which gives more nonlinear distortion.

[bui501]

JoaMat, would either of these two part be suitable as a substitute for stn9360 on your shrunken board? (Stn9360 is currently out of stock at all the major components vendors until 2019.)

PBHV9560ZX can handle up to 500mA but has higher capacitance; PBHV3160ZX has input/output capacitance closer to that of STN9360 but can only handle up to 100mA. Both are 600V devices like the STN9360.

Datasheets attached.

Even the more recent sands are starting to be difficult to source.

 

PBHV9560Z-465958.pdf

PBHV3160Z-466472.pdf

[Whitigir]

This is the reason why I buy spare parts and tuck it away :D.  Who knows when something started to stop working lol! A sure thing is that 10 years from now parts would be unobtainium 

[jose]

yes, I have the same habit. You never know when we'll need spare parts