DIY mini T2 Build Thread

[jamesmking]

3 hours ago, JoaMat said:

You don’t want to use same winding for small and output tubes. Separate windings for channels. So total four filament windings.

 

Thank you JoaMat, I thought seperate heater supplies was the case looking at how the heater pins are connected up via resistors to other parts of the circuit....

I think your are right about going seperate LV and HV transformers. I'm going golden reference HV and LV boards and have made a golden reference HV 220V dedicated board. I want to use synchronous rectifiers for the +-15V  on the golden reference LV board so I will go 2x15V  0.3A rather than centre tapped - Kevins sychchronous rectifiers dont like centre tapped transformers. Thank you for the specifications. Time to order some transformers....

[JoaMat]

Regarding GRHV regulators. What’s the minimum voltage differential/headroom for the GRHV to work properly?

[mwl168]

5 hours ago, JoaMat said:

Regarding GRHV regulators. What’s the minimum voltage differential/headroom for the GRHV to work properly?

 

This reminded me of an exercise/experiment I went through with mypasswordis (we also consulted with Kerry) two years ago. I was hoping to figure out what's the highest, stable regulated voltage I can get from the GRHV with the Antek transformer I was using (300VA/325v rated).

The short answer is that it's very load-dependent. See the measured data I copied from my PMs with mypasswordis.  The GRHV was setup to deliver 426VDC rails when these measurements were taken. We never got to the bottom of this whole thing but I ultimately return my GRHV to regulate to 400VDC rails. 

Notice the pre-reg voltage is quite low when the power supply is loaded. This is the part that still confuses me.

Maybe mypasswordis, Kerry and others will chime in.  

 

 

 

                         secondary ac (Antek 300VA/325v)     pre-reg V at bleeder resistor        regulated output          current draw 

No load                         332v/332v                                       +461V/-461V                                 +426V/-426V            

HV Carbon                    326v/326v                                       +431V/-430V                                 +426V/-425V              97mA/101mA 

Blue Hawaii                  318v/316v                                       +422V/-414V                                 +416V/-408V              81mA/147mA

 

Edited January 11, 2020 by mwl168

[jamesmking]

For me, with the golden reference hv connected to my modern blue hawaii via a variac the golden ref hv needs an *absolute* minimum of 315.5VAC at its AC input to regulate at 404.3VDC output.  In normal use I have 330VAC input to the golden ref for 404.3V output. 

 

Edited January 11, 2020 by jamesmking

[jose]

Am I the only one who uses 450Vdc output?

[JoseA]

Hi JoaMat,

You gave the power of the HV transformer (135VA or 143VA) what will be the power of the LV transformer?

Thanks

[JoaMat]

84,3 VA. 

Voltage * current of each winding. Then add VA of all windings. Done. 

Reservation for errors!

[JoaMat]

3 hours ago, mwl168 said:

 

This reminded me of an exercise/experiment I went through with mypasswordis (we also consulted with Kerry) two years ago. I was hoping to figure out what's the highest, stable regulated voltage I can get from the GRHV with the Antek transformer I was using (300VA/325v rated).

The short answer is that it's very load-dependent. See the measured data I copied from my PMs with mypasswordis.  The GRHV was setup to deliver 426VDC rails when these measurements were taken. We never got to the bottom of this whole thing but I ultimately return my GRHV to regulate to 400VDC rails. 

Notice the pre-reg voltage is quite low when the power supply is loaded. This is the part that still confuses me.

Maybe mypasswordis, Kerry and others will chime in.  

 

 

 

                         secondary ac (Antek 300VA/325v)     pre-reg V at bleeder resistor        regulated output                current draw 

No load                         332v/332v                                       +461V/-461V                                 +426V/-426V            

HV Carbon                    326v/326v                                       +431V/-430V                                 +426V/-425V                     97mA/101m 

Blue Hawaii                  318v/316v                                       +422V/-414V                                 +416V/-408V                     81mA/147mA

 

 

3 hours ago, jamesmking said:

For me, with the golden reference hv connected to my modern blue hawaii via a variac the golden ref hv needs an *absolute* minimum of 315.5VAC at its AC input to regulate at 404.3VDC output.  In normal use I have 330VAC input to the golden ref for 404.3V output. 

 

Thanks a lot, mwl168 and jamesmking.
 

Your information indicates – as I interpreter it - the voltage differential is in the region 30 – 40 volts. Good to know.

[mwl168]

1 hour ago, JoaMat said:

 Your information indicates – as I interpreter it - the voltage differential is in the region 30 – 40 volts. Good to know.

 JoaMat:

I want to be clear. In my case, a calculated rectified pre-reg voltage of 460VDC was not able to sustain the desired 425VDC regulated output with my GRHV when the load is more than 100mA per rail. For my Megatron build, I used a trafo with 350v secondaries to get regulated 425VDC rails from the GRHV under load. 

So, depending on the load, a calculated 30 - 40 VDC drop out voltage may or may not be enough with the GRHV.   

[jamesmking]

Unfortunately my electronic dc load only goes up to 100V, otherwise I would gladly do some measurements and plot some curves.  

[Kerry]

On 1/11/2020 at 12:37 PM, JoaMat said:

Regarding GRHV regulators. What’s the minimum voltage differential/headroom for the GRHV to work properly?

I’ve been able to run the regulators with a 12-15v differential. Yes load can come into play. To be safe you should add another 10% on top of the desired HV plus the differential. That’s to accommodate voltage variations of the mains, which may not be applicable for you. 

[jamesmking]

My mini T2 amp boards. Note the screw terminals on the top to make troubleshooting easier  I know JoaMat does not believe in screw terminals but I do. 😉. If you are careful its even possible to orientate the terminals so the silkscreen voltage markings are still visible 🙂 . I take a 2 terminal and cut it in half and put the original smooth side parallel to the silkscreen markings. There is not space for screw terminals for the heaters of the el34s so I (possibly temporarily) have soldered in a header pin at 45 dgrees 

 

 

I have used nylon m3 standoffs for the first layer so that there is no chance of the standoff shorting against anything and brass standoffs for the remaining layers to get to the required height.

SMD soldering was done in a three step process.

1 apply solder paste to the pads.

2. melt the solder using a hakko 888D station at 330 degree C using a fine tip until the paste has migrated to only be on the pads.

3 Hold the component with twisers onto the pads and with my other hand use a Quick 861DA hot air station using the smallest supplied nozzle with airflow setting all the way down to 5 (even then its possible to send one of the small diodes into orbit with that air flow)...  to reflow the solder onto the component.

the reason I soldered this way was,

1. I could collect any excess solder with the hakko, since I dont have an electronic solder paste dispenser... 😞 

2. I could be reasonably certain there was no solder paste left under a component.

I use the same method for soldering Kevins' synchronous rectifiers.

 

 

this is the schematic of the v0.2 pcb with valves from the silkscreen. The red components have changed since the last published schematic. 

Edited February 15, 2020 by jamesmking

[JoaMat]

Looks fine. You found smd LED with color mark?

One remark according the terminal block. The pad is C120-50 meaning pad diameter 120th and hole is 50th wide. Clearance to ground plane is 40th. When designing layout with terminal blocks I keep the entirely area under the block free from ground plane. It will certainly be alright to solder in a terminal block as done above but be sure to solder properly. You don’t want solder to flow out under the block.

I always let wires come in from the other (non tube) side.

[jamesmking]

I added  the red anode mark myself when I tested the leds with a multimeter.

 

[Kerry]

Regarding the Offset servo, I use OPA197 in my T2 build and have not had any failures.

[Mirko]

@jamesmking is the heater from pin 7 on el34 connected correctly on your schematic? Seems that it's connected to the junction of r28/29 on the original schematic?

[jamesmking]

Well spotted Mirko, you are correct. I will update the schematic.

[mwl168]

Thanks to JoaMat - I just posted updated schematic and BOM in the first post of this thread.

[JoaMat]

Intention is to have updated schematic and BOM at just on place to avoid different versions floating around. This place is the first post of this thread. The advantage is that I, as the disorganized person I’m, cant publish schematic or BOM whenever I feel for it. I need Michaels (mwl168) approval and he is the only one that can make changes public – in the first post of this thread.

Anyhow, please take look at the new files for errors or other issues.

Below are 3D prints of the new layouts (them I'm authorized to publish myself)

Please check for errors. I appreciate feedback especially regarding silkscreen, if anything seems unclear, let me know.

Changes:

• Some resistors value.
• Offset balance op amp: OPA197, hopefully more reliable.
• Balance servo op amp: LF356, is more similar to LF353 in the DIY T2. OP27 seems to be reliable in this position though.

Edited February 15, 2020 by JoaMat

[mwl168]

Another updated BOM from JoaMat in the first post.

There is a URL link in the BOM spreadsheet.Copy and paste this URL onto a web browser will take you into Mouser website with the BOM loaded and ready to be ordered (click on it in the spreadsheet won't work). Very slick! But, as always, do check for correctness and stocking status of each part.   

[mwl168]

I was inquired by PM about the board size so I'll post the information here for all.

Do keep in mind the dimensions are of the current boards and, though not very likely, there is no guarantee that the dimensions will not change when final testings are all completed. 

The mini T2 boards are 160mm x 100mm. The PSU main board is 140mm x 89mm.

Keep in mind the way the PSU main board is laid out means the footprint needed will depend on how you decide to mount the GRHV and GRLV boards.

Edited February 25, 2020 by mwl168

[rayofsi]

curious if any builders are selling these.. I know how a soldering iron works.. but something like this is beyond my scope.

[JoaMat]

Nope!

[jamesmking]

My transformers have arrived! and I have begun initial amp board testing.

A few more mini T2 build notes/observations.

Once the amp boards are completed using diode test on a multimeter across the leds will not make them light. 

The 2 red leds on the underside of the amp boards will not light if just heaters and + and -15V are applied... you need the high voltages too.

The el34s heaters run at about -400V DC with respect to ground plane on the amp board, the 6922 heaters run just about 0V with respect to ground.

The attachment holes in the pcb for the heatsink L bracket are plated and connected to the amp boards ground plane. So the L bracket, and therefore the heatsink and ultimately the entire case is connected to the pcb ground plane.

I have decideed to turn the amp boards around so the 6922 are at the front... to me it looks more visually pleasing... 🙃

 

 

 

 

[jamesmking]

My Initial impressions of the mini T2. Listening on my stax 007a mk2, normal caveats apply. There is no point comparing it to the Stax srm007 its like comparing a Bentley to a slug. Compared to my hi-amp alpha centauri there is also no competition. The mini T2 out does it in every department as does my Blue Hawaii. I will compare the mini T2 to the Blue Hawaii

(mini t2 (golden ref HV and LV, kevin gilmore synchronous rectifiers in the LV tkd volume, cardas hookup wire, groove tubes el34s, electro harmonix gold pin 6922s, simulated single box - transformers piled on top of each other 😞)

(blue hawaii, (golden ref HV and LV, kevin gilmore synchronous rectifiers in the LV tkd volume, cardas hookup wire, groove tubes el34s single box) 

The t2 has more smoothness, and yet more detailed and faster treble with a little more sparkle than the BH. The treble speed is not done via a bright hardness but rather its just fast and articulate and is very musical and "valve like". Mid range is equally strong with more smoothness and more detail than the hawaii. Bass is good, not necessarily much better than the blue hawaii, perhaps the BH goes deeper but the mini T2 is again more rounded and smooth. I think the upper bass is better on the mini t2  which partly makes up for the slight lack of depth. The mini T2 does not shout or dazzle, its just fast and natural. It makes you want to turn the volume up, not because its missing something but because it sounds like its not straining and just wants to sing.

Micro dynamics are much better than my BH, Macro dynamics are perhaps a bit more restrained than the BH but the compromise works for me because the micro is so much better then the BH. The mini T2 loves strings and acoustic music. I can’t comment much beyond this since I only listen to Jazz and classical. Image placement is like a blanket around the listener, the Mini has the ability to present the sound stage close to the listener when the recording is close. With the BH everything seems about the same distance regardless of the recording. When the recording has the sound stage further back the mini T2 presents a very wide, detailed soundstage but does not hyper focus on individual instruments, there is detail and yet the acoustic of the venue comes through. I feel the mini T 2is less analytical than the BH, interplay between instruments is better presented, its more musical and makes you want to play at whatever volume you want. there is much less of a volume sweet spot than the BH. 

Some suggested listening for what this amp can do

Rodrigo guitar concertos academy of st martin in the fields philips - beautiful guitar sounds with detail and fantastic plucked strings.

Hank Thompson live at the golden nugget. - the mini t2 pulls off the orange blossom special without brightness and the level of detail shows you it’s a packed venue with gambling noises everywhere.

j c Bach trio sonatas on chandos - mini t2 shows the mid range warmth that the blue hawaii missing, the recording is on original instruments and can sound a little thin... no so with the mini t2

paul desmond bossa antigua, tape hiss is separate from the instruments, centre drums are articulate, it’s a warm sound but not dull or muddy. The sound is beautiful there is detail but it’s not thrown at you.

I am very glad I started this project - even though I had very little smd soldering experience. Everything worked first time no smoke no drama. Now all I need is some cnc machining for the case. the mini T2 consumes about 25W less than the BH and runs cooler too.

Conclusion, both the BH and mini T2 are very very good. Building a blue hawaii is easier - there is almost no smd parts and will be cheaper: one less psu board, less transformer windings, less valves and can be built in one box. But the mini T2 is better sounding..... If you have the time, money, skill and equipment give the mini T2 some serious consideration. Now I have to build a full T2 with modern components....

and finally a very large thank you to JoaMat for providing me the mini t2 boards and for his help and support during this build.

Best regards and happy building

James

 

 

 

Edited March 5, 2020 by jamesmking